diff --git "a/146m14b100mdedup/3328731.err" "b/146m14b100mdedup/3328731.err"
new file mode 100644--- /dev/null
+++ "b/146m14b100mdedup/3328731.err"
@@ -0,0 +1,5708 @@
+6: 2023-03-17 10:25:12.714285: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714314: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714330: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: 2023-03-17 10:25:12.714714: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714725: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714723: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714731: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714733: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714330: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714347: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714346: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: 2023-03-17 10:25:12.714741: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714349: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714740: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+2: 2023-03-17 10:25:12.714761: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+2: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+6: 2023-03-17 10:25:12.714358: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+6: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715151: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715257: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715267: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715291: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715292: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715390: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715395: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+1: 2023-03-17 10:25:12.715402: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+1: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715647: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715670: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715704: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715740: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715745: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715786: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: 2023-03-17 10:25:12.715836: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715847: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715781: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715876: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+7: 2023-03-17 10:25:12.715827: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+7: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715887: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715873: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715896: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715900: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+0: 2023-03-17 10:25:12.715908: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+0: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716440: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716499: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716529: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716553: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716565: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716527: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716572: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+3: 2023-03-17 10:25:12.716585: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+3: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716898: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716912: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716897: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716949: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716944: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716935: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716955: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716942: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: 2023-03-17 10:25:12.716958: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716933: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716944: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716951: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+5: 2023-03-17 10:25:12.716973: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+5: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716938: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716972: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:12.716990: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA
+4: To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
+4: 2023-03-17 10:25:25.512782: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512813: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512828: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512878: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512869: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512863: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512880: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.512900: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.531885: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+4: 2023-03-17 10:25:25.531925: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+4: 2023-03-17 10:25:25.531963: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+4: 2023-03-17 10:25:25.531961: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+4: 2023-03-17 10:25:25.531984: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.513222: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 2023-03-17 10:25:25.531990: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.513455: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+4: 2023-03-17 10:25:25.531995: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:25.532005: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 2023-03-17 10:25:25.513366: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513481: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513250: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513516: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 2023-03-17 10:25:25.513396: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 2023-03-17 10:25:25.513750: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513297: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532153: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513534: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 2023-03-17 10:25:25.513419: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513475: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513777: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513307: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.513427: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532179: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.513782: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513518: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.513443: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513504: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513743: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513322: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.513461: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:25.513818: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513556: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.532197: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 2023-03-17 10:25:25.513455: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513539: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513843: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513287: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.513483: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:25.513857: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:25.532240: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513544: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.532223: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 2023-03-17 10:25:25.513459: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513581: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513819: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513322: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.513513: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532202: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.513881: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.513561: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.513488: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513573: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513862: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.513329: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+5: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.513529: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532225: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.513887: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:25.532245: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.532559: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532232: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 2023-03-17 10:25:25.532241: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 2023-03-17 10:25:25.532260: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513525: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513879: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.532385: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532590: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.513542: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+1: 2023-03-17 10:25:25.513488: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 2023-03-17 10:25:25.513874: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:25.532417: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.532455: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.532470: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532610: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532625: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532639: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532642: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.532246: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+1: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:25.532324: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513558: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513870: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+6: 2023-03-17 10:25:25.532519: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+5: 2023-03-17 10:25:25.532665: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.513554: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 2023-03-17 10:25:25.513888: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:25.532727: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:25.513587: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+3: 2023-03-17 10:25:25.513888: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+7: 2023-03-17 10:25:25.513591: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_46200
+0: 2023-03-17 10:25:25.532793: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:25.532266: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:25.532269: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 0125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:25.532271: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.532287: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.532284: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.532302: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.532830: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.532851: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532580: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:25.532270: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:25.532283: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:25.532285: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:25.532289: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+7: 2023-03-17 10:25:25.532312: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.532858: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.532860: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532618: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.532603: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.532613: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.532871: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532633: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532675: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+6: 2023-03-17 10:25:25.532630: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532688: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532699: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532707: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+2: 2023-03-17 10:25:25.532889: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.533009: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+0: 2023-03-17 10:25:25.533025: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
+3: 2023-03-17 10:25:56.065401: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065431: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065453: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065466: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065498: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065507: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065521: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.065728: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+3: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068312: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068328: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068325: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068323: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068323: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068324: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068327: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068330: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068329: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+3: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+3: 2023-03-17 10:25:56.068342: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068348: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068347: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068351: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068351: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068352: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+3: 2023-03-17 10:25:56.068354: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.073822: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073848: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073872: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073883: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073892: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073906: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.073995: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.074001: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+4: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.074978: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075008: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075028: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075038: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075049: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075053: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075068: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.075268: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+0: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075643: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075687: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075695: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075719: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075732: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075735: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075739: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.075816: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+2: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076166: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076168: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076168: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076171: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076170: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076176: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076182: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076182: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076177: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076191: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076192: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076195: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076195: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076198: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+4: 2023-03-17 10:25:56.076238: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+4: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+4: 2023-03-17 10:25:56.076254: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.076772: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076803: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076836: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076850: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076861: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076872: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076875: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.076959: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+5: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077798: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077837: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077850: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077878: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: 2023-03-17 10:25:56.077956: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077887: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.077894: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.077974: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: 2023-03-17 10:25:56.077889: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.077996: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+6: 2023-03-17 10:25:56.077911: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.078004: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.078017: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.078019: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.078035: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.078038: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+7: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078201: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078203: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078205: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078207: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078205: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078220: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078220: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078222: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078208: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078223: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078230: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078236: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078272: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078293: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+2: 2023-03-17 10:25:56.078298: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+2: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+2: 2023-03-17 10:25:56.078312: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079845: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079847: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079848: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079849: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079851: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079856: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079851: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079855: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+5: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+5: 2023-03-17 10:25:56.079865: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079867: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079868: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079869: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079871: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079871: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079873: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+5: 2023-03-17 10:25:56.079874: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+7: 2023-03-17 10:25:56.080151: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080153: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080214: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: 2023-03-17 10:25:56.080156: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080158: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080166: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080160: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080215: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: 2023-03-17 10:25:56.080160: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080218: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: 2023-03-17 10:25:56.080168: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+7: 2023-03-17 10:25:56.080175: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+7: 2023-03-17 10:25:56.080178: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080182: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+7: 2023-03-17 10:25:56.080183: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080220: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: 2023-03-17 10:25:56.080223: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080239: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080223: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+7: 2023-03-17 10:25:56.080243: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+7: 2023-03-17 10:25:56.080259: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080226: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080234: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080234: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080235: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080238: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080231: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080239: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080235: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+6: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+6: 2023-03-17 10:25:56.080252: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080257: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+6: 2023-03-17 10:25:56.080260: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.091327: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091363: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091376: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091413: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091421: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091431: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091457: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.091490: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/project_462000125
+1: /samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093495: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093501: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093502: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093511: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093507: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093508: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093506: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093524: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093525: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093529: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093532: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093532: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093586: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093590: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+1: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+1: 2023-03-17 10:25:56.093606: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+1: 2023-03-17 10:25:56.093610: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077519: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077521: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077523: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077524: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077526: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077527: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077536: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077537: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077536: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077535: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077545: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077546: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077546: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077549: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /pfs/lustrep2/projappl/project_462000125/samantao-public/apps/aws-ofi-rccl:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/rccl/rccl-develop-release/rccl/lib:/pfs/lustrep4/projappl/project_462000075/samantao-public/rocm/glibc/selected:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hip/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/hsa/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/rocm/rocm-5.2.3/llvm:/pfs/lustrep2/projappl/pro
+0: ject_462000125/samantao-public/apps/suse-repo-deps/lib64:/pfs/lustrep2/projappl/project_462000125/samantao-public/apps/suse-repo-deps/usr/lib64:/opt/cray/pe/python/3.9.12.1/lib:/opt/cray/pe/gcc-libs:/opt/cray/libfabric/1.15.0.0/lib64
+0: 2023-03-17 10:25:56.077556: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: 2023-03-17 10:25:56.077565: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
+0: Successfully preprocessed all matching files.
+0: Detected CUDA files, patching ldflags
+0: Emitting ninja build file /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/fused_kernels/build/build.ninja...
+0: Building extension module scaled_upper_triang_masked_softmax_cuda...
+0: Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
+0: Loading extension module scaled_upper_triang_masked_softmax_cuda...
+0: Successfully preprocessed all matching files.
+0: Detected CUDA files, patching ldflags
+0: Emitting ninja build file /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/fused_kernels/build/build.ninja...
+0: Building extension module scaled_masked_softmax_cuda...
+0: Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
+0: Loading extension module scaled_masked_softmax_cuda...
+0: Successfully preprocessed all matching files.
+0: Detected CUDA files, patching ldflags
+0: Emitting ninja build file /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/fused_kernels/build/build.ninja...
+0: Building extension module fused_mix_prec_layer_norm_cuda...
+0: Allowing ninja to set a default number of workers... (overridable by setting the environment variable MAX_JOBS=N)
+0: Loading extension module fused_mix_prec_layer_norm_cuda...
+0: Successfully preprocessed all matching files.
+0: Successfully preprocessed all matching files.
+0: Successfully preprocessed all matching files.
+0: Successfully preprocessed all matching files.
+0: Successfully preprocessed all matching files.
+0: Successfully preprocessed all matching files.
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+1: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+1:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+6: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+6:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+4: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+4:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+2:   warnings.warn(
+5: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+5:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+0: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+0:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+3: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+3:   warnings.warn(
+7: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py:429: UserWarning: torch.distributed.distributed_c10d._get_global_rank is deprecated please use torch.distributed.distributed_c10d.get_global_rank instead
+7:   warnings.warn(
+7: [E ProcessGroupNCCL.cpp:821] [Rank 60] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805943 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 56] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805950 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 59] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 57] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 58] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 63] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 62] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:821] [Rank 61] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 5] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805964 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 52] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805958 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 33] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 38] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1806000 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 45] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 46] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 42] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 41] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 43] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 47] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 40] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: [E ProcessGroupNCCL.cpp:821] [Rank 44] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 20] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 18] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 19] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 22] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 23] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 21] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 50] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+2: [E ProcessGroupNCCL.cpp:821] [Rank 17] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 13] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 9] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 10] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 6] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805997 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 55] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 51] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 53] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 54] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 48] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6: [E ProcessGroupNCCL.cpp:821] [Rank 49] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 32] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 35] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 34] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 39] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 36] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+4: [E ProcessGroupNCCL.cpp:821] [Rank 37] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 24] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 26] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805982 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 25] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 29] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 11] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 8] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805989 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 12] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 15] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:821] [Rank 14] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 28] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 27] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 31] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: [E ProcessGroupNCCL.cpp:821] [Rank 30] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=50, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 2] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 7] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 1] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 3] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: [E ProcessGroupNCCL.cpp:821] [Rank 4] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+7: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0: Traceback (most recent call last):
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0: Traceback (most recent call last):
+7:     return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3: Traceback (most recent call last):
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,    
+7: return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0: Traceback (most recent call last):
+3: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3: Traceback (most recent call last):
+3: Traceback (most recent call last):
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7: Traceback (most recent call last):
+7: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3: Traceback (most recent call last):
+4: Traceback (most recent call last):
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7: Traceback (most recent call last):
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3: Traceback (most recent call last):
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5: Traceback (most recent call last):
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4: Traceback (most recent call last):
+0:     main()
+3:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3: Traceback (most recent call last):
+3:     main()
+7: Traceback (most recent call last):
+0:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0: main()
+2: Traceback (most recent call last):
+2: Traceback (most recent call last):
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2: Traceback (most recent call last):
+2: Traceback (most recent call last):
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+2: Traceback (most recent call last):
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5: Traceback (most recent call last):
+5: Traceback (most recent call last):
+5: Traceback (most recent call last):
+4: Traceback (most recent call last):
+4:     main()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:         main()
+7: main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6: Traceback (most recent call last):
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4: Traceback (most recent call last):
+7:     main()
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:     main()    
+3: main()
+5: Traceback (most recent call last):
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7: Traceback (most recent call last):
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3: main()
+5: Traceback (most recent call last):
+4: Traceback (most recent call last):
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0: Traceback (most recent call last):
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4: Traceback (most recent call last):
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:     return f(*args, **kwargs)
+5: Traceback (most recent call last):
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     return f(*args, **kwargs)
+2:     main()
+2: Traceback (most recent call last):
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6: Traceback (most recent call last):
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4:     Traceback (most recent call last):
+4: return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:     return f(*args, **kwargs)
+5:     main()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     return f(*args, **kwargs)
+0: Traceback (most recent call last):
+2:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     main()
+6:     main()
+5:             main()main()main()
+5: 
+5: 
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0: Traceback (most recent call last):
+0:     return f(*args, **kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:         main()main()    
+2: 
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:     main()
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:     main()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     return f(*args, **kwargs)
+0:     return f(*args, **kwargs)
+2: main()
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     return f(*args, **kwargs)
+6: Traceback (most recent call last):
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:     main()
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:     Traceback (most recent call last):
+0: pretrain(train_valid_test_datasets_provider, model_provider, forward_step,    
+0: main()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2: Traceback (most recent call last):
+2:     return f(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     main()
+4:     main()
+7:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:     main()    pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:         pretrain(train_valid_test_datasets_provider, model_provider, forward_step,return f(*args, **kwargs)
+3: 
+4:     main()
+7: return f(*args, **kwargs)
+7:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7: model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1: Traceback (most recent call last):
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:     return f(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:         return f(*args, **kwargs)    return f(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     main()
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2: return f(*args, **kwargs)
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:     return f(*args, **kwargs)    
+3: return f(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:     main()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     return f(*args, **kwargs)    
+0: main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:     return f(*args, **kwargs)
+5:             return f(*args, **kwargs)return f(*args, **kwargs)return f(*args, **kwargs)
+5: 
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     main()
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1: Traceback (most recent call last):
+1: Traceback (most recent call last):
+1: Traceback (most recent call last):
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     return f(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:     return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:     return f(*args, **kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     return f(*args, **kwargs)
+7:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     return f(*args, **kwargs)
+2:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6: Traceback (most recent call last):
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:     return f(*args, **kwargs)
+4:     return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1: Traceback (most recent call last):
+1: Traceback (most recent call last):
+0:     return f(*args, **kwargs)
+2:         pretrain(train_valid_test_datasets_provider, model_provider, forward_step,      File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2: pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2: pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6: Traceback (most recent call last):
+6: Traceback (most recent call last):
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:     return f(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     return f(*args, **kwargs)
+7:     engine = PipelineEngine(args=args,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1: Traceback (most recent call last):
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     main()
+5:     return f(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4: 
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6: Traceback (most recent call last):
+5:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:         model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)    
+2: 
+2: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7: return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:     return f(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     return f(*args, **kwargs)
+1:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     return f(*args, **kwargs)
+1:         main()main()
+1: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6: Traceback (most recent call last):
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:     main()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:     return f(*args, **kwargs)
+4:     return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:         model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0: 
+2:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:     main()
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     main()Traceback (most recent call last):
+1: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2: model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:           File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2: model, optimizer, _, lr_scheduler = deepspeed.initialize(model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:         main()
+6: main()
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     return f(*args, **kwargs)        main()
+1: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1: main()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1: 
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:     main()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:         return f(*args, **kwargs)return f(*args, **kwargs)
+1: 
+0:     engine = PipelineEngine(args=args,
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     main()
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     engine = PipelineEngine(args=args,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:         engine = PipelineEngine(args=args,
+2: engine = PipelineEngine(args=args,    
+2: engine = PipelineEngine(args=args,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     return f(*args, **kwargs)
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0: model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1: return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     engine = PipelineEngine(args=args,
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:         engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,
+0: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:     engine = PipelineEngine(args=args,
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     engine = PipelineEngine(args=args,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:     super().__init__(*super_args, **super_kwargs)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:     main()
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     return f(*args, **kwargs)
+4: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     return f(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     super().__init__(*super_args, **super_kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     return f(*args, **kwargs)    
+4:     engine = PipelineEngine(args=args,
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:         super().__init__(*super_args, **super_kwargs)super().__init__(*super_args, **super_kwargs)
+0: 
+6: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)    
+6: return f(*args, **kwargs)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     return f(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     super().__init__(*super_args, **super_kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:         engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,
+7: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     super().__init__(*super_args, **super_kwargs)
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     engine = PipelineEngine(args=args,
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     super().__init__(*super_args, **super_kwargs)
+0:     super().__init__(*super_args, **super_kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     super().__init__(*super_args, **super_kwargs)
+7:         engine = PipelineEngine(args=args,    engine = PipelineEngine(args=args,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7: super().__init__(*super_args, **super_kwargs)
+7:     
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     super().__init__(*super_args, **super_kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:         engine = PipelineEngine(args=args,model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7: self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:     super().__init__(*super_args, **super_kwargs)
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     engine = PipelineEngine(args=args,
+6: pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     engine = PipelineEngine(args=args,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6: pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4: 
+7:     super().__init__(*super_args, **super_kwargs)
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     self._configure_distributed_model(model)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     super().__init__(*super_args, **super_kwargs)
+7:     engine = PipelineEngine(args=args,
+1:     return f(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:     self._configure_distributed_model(model)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:     super().__init__(*super_args, **super_kwargs)
+0:     self._configure_distributed_model(model)
+2:     engine = PipelineEngine(args=args,
+6:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     super().__init__(*super_args, **super_kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:         super().__init__(*super_args, **super_kwargs)super().__init__(*super_args, **super_kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:     engine = PipelineEngine(args=args,
+7:     super().__init__(*super_args, **super_kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:     self._configure_distributed_model(model)
+0:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0: self._configure_distributed_model(model)
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(    
+1: model, optimizer, _, lr_scheduler = deepspeed.initialize(  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     super().__init__(*super_args, **super_kwargs)
+7:     self._configure_distributed_model(model)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     self._configure_distributed_model(model)
+2:     super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     engine = PipelineEngine(args=args,
+7:         super().__init__(*super_args, **super_kwargs)self._configure_distributed_model(model)
+7: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     self._configure_distributed_model(model)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4: engine = PipelineEngine(args=args,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:     super().__init__(*super_args, **super_kwargs)
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:     self._configure_distributed_model(model)
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1: 
+0:     self._configure_distributed_model(model)
+2:     super().__init__(*super_args, **super_kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     super().__init__(*super_args, **super_kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:     super().__init__(*super_args, **super_kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:     self._configure_distributed_model(model)
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:     self._broadcast_model()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     self._broadcast_model()
+3:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     engine = PipelineEngine(args=args,
+4:         super().__init__(*super_args, **super_kwargs)super().__init__(*super_args, **super_kwargs)
+4: 
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:     self._configure_distributed_model(model)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     self._broadcast_model()
+0:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0: self._broadcast_model()
+3:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)    
+3: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     self._configure_distributed_model(model)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)    
+3: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:         self._broadcast_model()self._broadcast_model()
+0: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:         model, optimizer, _, lr_scheduler = deepspeed.initialize(model, optimizer, _, lr_scheduler = deepspeed.initialize(    
+6: 
+4:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:     self._broadcast_model()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     self._configure_distributed_model(model)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6: engine = PipelineEngine(args=args,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6: model, optimizer, _, lr_scheduler = deepspeed.initialize(  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._configure_distributed_model(model)
+2:     self._configure_distributed_model(model)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     self._broadcast_model()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     self._configure_distributed_model(model)
+3:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     dist.broadcast(p,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:         self._configure_distributed_model(model)self._configure_distributed_model(model)
+2: 
+3:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     engine = PipelineEngine(args=args,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:             model, optimizer, _, lr_scheduler = deepspeed.initialize(model, optimizer, _, lr_scheduler = deepspeed.initialize(model, optimizer, _, lr_scheduler = deepspeed.initialize(    
+6:             engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,
+6: 
+6: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()
+1:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     dist.broadcast(p,
+3: model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3: 
+3: 
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     super().__init__(*super_args, **super_kwargs)
+4:     self._configure_distributed_model(model)    
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     dist.broadcast(p,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4: self._configure_distributed_model(model)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()    
+7: dist.broadcast(p,
+1:     engine = PipelineEngine(args=args,
+0:     dist.broadcast(p,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:     engine = PipelineEngine(args=args,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:     self._broadcast_model()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     dist.broadcast(p,
+2:     return func(*args, **kwargs)
+3:     engine = PipelineEngine(args=args,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:     dist.broadcast(p,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:     self._broadcast_model()
+1:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:     engine = PipelineEngine(args=args,
+0:     dist.broadcast(p,
+0:     dist.broadcast(p,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0: 
+3:     engine = PipelineEngine(args=args,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:     super().__init__(*super_args, **super_kwargs)
+4:     self._broadcast_model()
+7:     self._broadcast_model()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:         engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,
+1: 
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:         self._broadcast_model()self._broadcast_model()
+2: 
+3:         self._configure_distributed_model(model)engine = PipelineEngine(args=args,
+3: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     super().__init__(*super_args, **super_kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     return func(*args, **kwargs)
+0:     return func(*args, **kwargs)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0: 
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     engine = PipelineEngine(args=args,    
+4:     self._broadcast_model()
+7:     dist.broadcast(p,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:     engine = PipelineEngine(args=args,
+6: super().__init__(*super_args, **super_kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6: super().__init__(*super_args, **super_kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     return func(*args, **kwargs)
+2:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:     self._broadcast_model()
+7:     dist.broadcast(p,
+1:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:     self._broadcast_model()
+3:         engine = PipelineEngine(args=args,engine = PipelineEngine(args=args,
+3: 
+6:     super().__init__(*super_args, **super_kwargs)    
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     return func(*args, **kwargs)
+0:     dist.broadcast(p,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6: super().__init__(*super_args, **super_kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:     self._broadcast_model()
+7:     dist.broadcast(p,
+1:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:     super().__init__(*super_args, **super_kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4: return func(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:     self._broadcast_model()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     super().__init__(*super_args, **super_kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     dist.broadcast(p,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:     engine = PipelineEngine(args=args,
+2:     self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2: 
+3:     engine = PipelineEngine(args=args,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:     self._broadcast_model()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     self._broadcast_model()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:     self._configure_distributed_model(model)
+4:     self._broadcast_model()
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:     super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:         return func(*args, **kwargs)return func(*args, **kwargs)
+7: 
+3:     super().__init__(*super_args, **super_kwargs)    
+3: super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:         self._configure_distributed_model(model)self._configure_distributed_model(model)
+6: 
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:     return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:     dist.broadcast(p,
+7:     dist.broadcast(p,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     super().__init__(*super_args, **super_kwargs)
+6:         self._configure_distributed_model(model)self._configure_distributed_model(model)
+6: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4: dist.broadcast(p,
+7:     dist.broadcast(p,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:         self._configure_distributed_model(model)engine = PipelineEngine(args=args,
+3: 
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3: super().__init__(*super_args, **super_kwargs)
+6:     self._configure_distributed_model(model)
+4:     dist.broadcast(p,
+7:     return func(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     self._configure_distributed_model(model)
+4:     return torch.distributed.broadcast(tensor=tensor,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:     return func(*args, **kwargs)
+1:     super().__init__(*super_args, **super_kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     self._configure_distributed_model(model)    
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:     return func(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     return func(*args, **kwargs)
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:     dist.broadcast(p,
+3: self._broadcast_model()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:     self._configure_distributed_model(model)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:     return func(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:     return func(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:     dist.broadcast(p,
+3:         self._configure_distributed_model(model)self._configure_distributed_model(model)
+3: 
+4:         return func(*args, **kwargs)dist.broadcast(p,
+4: 
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:     return func(*args, **kwargs)
+0:     return func(*args, **kwargs)
+2:     dist.broadcast(p,
+3:     super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:     self._broadcast_model()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     return func(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:         self._configure_distributed_model(model)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:     dist.broadcast(p,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3: self._configure_distributed_model(model)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:     self._broadcast_model()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:     return func(*args, **kwargs)
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     dist.broadcast(p,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6: self._broadcast_model()
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2: dist.broadcast(p,
+3:     self._configure_distributed_model(model)
+6:     self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:     
+4:     dist.broadcast(p,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6: self._broadcast_model()
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     dist.broadcast(p,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:     self._broadcast_model()
+6:     self._broadcast_model()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:     return func(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:     return func(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:     self._configure_distributed_model(model)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     dist.broadcast(p,
+6:     self._broadcast_model()
+4:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)    
+7: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:     return func(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4: return func(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:         return torch.distributed.broadcast(tensor=tensor,return torch.distributed.broadcast(tensor=tensor,
+0: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     self._broadcast_model()
+6:     self._broadcast_model()
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:     return func(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:     return torch.distributed.broadcast(tensor=tensor,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     self._broadcast_model()
+4:     return torch.distributed.broadcast(tensor=tensor,
+7:     return torch.distributed.broadcast(tensor=tensor,
+1:     super().__init__(*super_args, **super_kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:     dist.broadcast(p,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:     return torch.distributed.broadcast(tensor=tensor,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:     return func(*args, **kwargs)    
+2: return func(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     self._broadcast_model()
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:     return func(*args, **kwargs)
+7:     return torch.distributed.broadcast(tensor=tensor,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3: 
+6:     dist.broadcast(p,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1: super().__init__(*super_args, **super_kwargs)
+2:     return func(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3: self._broadcast_model()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6: dist.broadcast(p,
+4:     return torch.distributed.broadcast(tensor=tensor,
+7:         return torch.distributed.broadcast(tensor=tensor,return torch.distributed.broadcast(tensor=tensor,
+7: 
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1: super().__init__(*super_args, **super_kwargs)
+2:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:         dist.broadcast(p,self._broadcast_model()
+3: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:     super().__init__(*super_args, **super_kwargs)
+2:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:     dist.broadcast(p,
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:             return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)work = group.broadcast([tensor], opts)return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4: 
+4: 
+7:     return torch.distributed.broadcast(tensor=tensor,
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     dist.broadcast(p,
+6: dist.broadcast(p,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4: RuntimeError: NCCL communicator was aborted on rank 32.  Original reason for failure was: [Rank 32] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: super().__init__(*super_args, **super_kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:     dist.broadcast(p,
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     return torch.distributed.broadcast(tensor=tensor,
+1:     super().__init__(*super_args, **super_kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     return func(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:             pretrain(train_valid_test_datasets_provider, model_provider, forward_step,pretrain(train_valid_test_datasets_provider, model_provider, forward_step,pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5: 
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:         return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2: 
+3:         return func(*args, **kwargs)dist.broadcast(p,
+3: 
+6:     dist.broadcast(p,
+5:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:         return torch.distributed.broadcast(tensor=tensor,return torch.distributed.broadcast(tensor=tensor,
+4: 
+7:         work = group.broadcast([tensor], opts)work = group.broadcast([tensor], opts)
+7: 
+2:     work = group.broadcast([tensor], opts)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:     
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5: 
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7: RuntimeErrorRuntimeError: : NCCL communicator was aborted on rank 59.  Original reason for failure was: [Rank 59] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.NCCL communicator was aborted on rank 56.  Original reason for failure was: [Rank 56] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805950 milliseconds before timing out.
+7: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     dist.broadcast(p,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:     dist.broadcast(p,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     return torch.distributed.broadcast(tensor=tensor,
+7:     work = group.broadcast([tensor], opts)
+1:     super().__init__(*super_args, **super_kwargs)
+3:     dist.broadcast(p,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7: RuntimeError: NCCL communicator was aborted on rank 63.  Original reason for failure was: [Rank 63] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2: RuntimeError    : NCCL communicator was aborted on rank 22.  Original reason for failure was: [Rank 22] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.return torch.distributed.broadcast(tensor=tensor,
+2: 
+3:     dist.broadcast(p,
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:             work = group.broadcast([tensor], opts)work = group.broadcast([tensor], opts)work = group.broadcast([tensor], opts)
+7: 
+7: 
+1:     self._configure_distributed_model(model)
+0:     return torch.distributed.broadcast(tensor=tensor,
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:     return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)    
+5: model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:     work = group.broadcast([tensor], opts)
+7: RuntimeErrorRuntimeErrorRuntimeError: : : NCCL communicator was aborted on rank 62.  Original reason for failure was: [Rank 62] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.NCCL communicator was aborted on rank 58.  Original reason for failure was: [Rank 58] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.NCCL communicator was aborted on rank 57.  Original reason for failure was: [Rank 57] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7: 
+7: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     self._broadcast_model()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4: RuntimeError: NCCL communicator was aborted on rank 33.  Original reason for failure was: [Rank 33] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+7:     work = group.broadcast([tensor], opts)
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1: self._configure_distributed_model(model)
+0:         return torch.distributed.broadcast(tensor=tensor,    return torch.distributed.broadcast(tensor=tensor,
+0: return torch.distributed.broadcast(tensor=tensor,
+0: 
+2:     return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     work = group.broadcast([tensor], opts)
+7: RuntimeError: NCCL communicator was aborted on rank 60.  Original reason for failure was: [Rank 60] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805943 milliseconds before timing out.
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     dist.broadcast(p,    
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4: RuntimeError: NCCL communicator was aborted on rank 34.  Original reason for failure was: [Rank 34] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+1:     self._configure_distributed_model(model)
+0:     return torch.distributed.broadcast(tensor=tensor,
+2:         return torch.distributed.broadcast(tensor=tensor,return torch.distributed.broadcast(tensor=tensor,
+2: 
+3: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     work = group.broadcast([tensor], opts)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4: RuntimeError: NCCL communicator was aborted on rank 36.  Original reason for failure was: [Rank 36] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+1:     self._configure_distributed_model(model)
+0:     work = group.broadcast([tensor], opts)
+2:     return torch.distributed.broadcast(tensor=tensor,
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     work = group.broadcast([tensor], opts)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0: RuntimeError: NCCL communicator was aborted on rank 0.  Original reason for failure was: [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=50, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     return func(*args, **kwargs)return func(*args, **kwargs)
+3: 
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:     RuntimeErrorwork = group.broadcast([tensor], opts): 
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4: NCCL communicator was aborted on rank 39.  Original reason for failure was: [Rank 39] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: RuntimeError: NCCL communicator was aborted on rank 35.  Original reason for failure was: [Rank 35] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:         self._configure_distributed_model(model)self._configure_distributed_model(model)
+1: 
+3:     return func(*args, **kwargs)
+6:     return func(*args, **kwargs)
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     work = group.broadcast([tensor], opts)
+1:     self._configure_distributed_model(model)
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4: RuntimeError: NCCL communicator was aborted on rank 38.  Original reason for failure was: [Rank 38] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1806000 milliseconds before timing out.
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     dist.broadcast(p,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     work = group.broadcast([tensor], opts)
+3:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0: RuntimeError: NCCL communicator was aborted on rank 2.  Original reason for failure was: [Rank 2] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:     return func(*args, **kwargs)
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     work = group.broadcast([tensor], opts)
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     self._broadcast_model()
+0: RuntimeError: NCCL communicator was aborted on rank 6.  Original reason for failure was: [Rank 6] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805997 milliseconds before timing out.
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:         work = group.broadcast([tensor], opts)work = group.broadcast([tensor], opts)
+0: 
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     self._broadcast_model()
+1:     self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0: RuntimeError: RuntimeErrorNCCL communicator was aborted on rank 4.  Original reason for failure was: [Rank 4] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1: 
+1:     self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1: 
+0: : NCCL communicator was aborted on rank 1.  Original reason for failure was: [Rank 1] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:     work = group.broadcast([tensor], opts)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:         engine = PipelineEngine(args=args,model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0: RuntimeError: NCCL communicator was aborted on rank 7.  Original reason for failure was: [Rank 7] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+3:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:     self._broadcast_model()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     engine = PipelineEngine(args=args,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:     engine = PipelineEngine(args=args,
+1:     self._broadcast_model()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:     work = group.broadcast([tensor], opts)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0: RuntimeError: NCCL communicator was aborted on rank 3.  Original reason for failure was: [Rank 3] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+3:     return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:     return func(*args, **kwargs)
+5:     engine = PipelineEngine(args=args,
+2:     work = group.broadcast([tensor], opts)
+3:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:     engine = PipelineEngine(args=args,
+1:         return func(*args, **kwargs)self._broadcast_model()
+1: 
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:     engine = PipelineEngine(args=args,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2: RuntimeError: NCCL communicator was aborted on rank 20.  Original reason for failure was: [Rank 20] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+3:     work = group.broadcast([tensor], opts)
+5:     engine = PipelineEngine(args=args,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     RuntimeErrorreturn torch.distributed.broadcast(tensor=tensor,: 
+3: NCCL communicator was aborted on rank 25.  Original reason for failure was: [Rank 25] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:     return func(*args, **kwargs)
+6: return func(*args, **kwargs)
+5:     super().__init__(*super_args, **super_kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:     return torch.distributed.broadcast(tensor=tensor,    
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     self._configure_distributed_model(model)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3: return torch.distributed.broadcast(tensor=tensor,    
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:     return func(*args, **kwargs)
+5:     super().__init__(*super_args, **super_kwargs)
+2:     work = group.broadcast([tensor], opts)
+3: return torch.distributed.broadcast(tensor=tensor,
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5: super().__init__(*super_args, **super_kwargs)
+3:     return torch.distributed.broadcast(tensor=tensor,
+5:     super().__init__(*super_args, **super_kwargs)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:     dist.broadcast(p,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2: RuntimeError: NCCL communicator was aborted on rank 23.  Original reason for failure was: [Rank 23] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:         super().__init__(*super_args, **super_kwargs)super().__init__(*super_args, **super_kwargs)
+5: 
+1:     dist.broadcast(p,
+3:     work = group.broadcast([tensor], opts)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3: RuntimeError: NCCL communicator was aborted on rank 31.  Original reason for failure was: [Rank 31] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+6:     return func(*args, **kwargs)
+5:     super().__init__(*super_args, **super_kwargs)
+1:     dist.broadcast(p,
+3:     work = group.broadcast([tensor], opts)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3: RuntimeError: NCCL communicator was aborted on rank 28.  Original reason for failure was: [Rank 28] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+6:         return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)    
+6: 
+5:     self._configure_distributed_model(model)
+1:     dist.broadcast(p,
+3:     work = group.broadcast([tensor], opts)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:     work = group.broadcast([tensor], opts)
+3:     RuntimeErrorwork = group.broadcast([tensor], opts): 
+5:     self._configure_distributed_model(model)    
+5: self._configure_distributed_model(model)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3: NCCL communicator was aborted on rank 29.  Original reason for failure was: [Rank 29] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3: RuntimeError: NCCL communicator was aborted on rank 26.  Original reason for failure was: [Rank 26] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805982 milliseconds before timing out.
+6: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)    
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6: 
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     self._configure_distributed_model(model)
+2: RuntimeError: NCCL communicator was aborted on rank 18.  Original reason for failure was: [Rank 18] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+3:     work = group.broadcast([tensor], opts)
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     dist.broadcast(p,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3: RuntimeError    : work = group.broadcast([tensor], opts)NCCL communicator was aborted on rank 30.  Original reason for failure was: [Rank 30] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: 
+6:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:     work = group.broadcast([tensor], opts)
+3: RuntimeError: NCCL communicator was aborted on rank 24.  Original reason for failure was: [Rank 24] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:     dist.broadcast(p,
+3:     work = group.broadcast([tensor], opts)
+6:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     self._configure_distributed_model(model)
+5:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2: RuntimeError: NCCL communicator was aborted on rank 16.  Original reason for failure was: [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.
+3: RuntimeError: NCCL communicator was aborted on rank 27.  Original reason for failure was: [Rank 27] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5: self._configure_distributed_model(model)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:     return func(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     self._broadcast_model()
+1:     return func(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:     work = group.broadcast([tensor], opts)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:     return torch.distributed.broadcast(tensor=tensor,
+5:     self._broadcast_model()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:         self._broadcast_model()  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5: self._broadcast_model()
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:     return func(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2: RuntimeError: NCCL communicator was aborted on rank 17.  Original reason for failure was: [Rank 17] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:         return func(*args, **kwargs)return func(*args, **kwargs)
+1: 
+5:     self._broadcast_model()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:         return func(*args, **kwargs)    dist.broadcast(p,
+1: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     self._broadcast_model()
+1: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:     work = group.broadcast([tensor], opts)
+5:     self._broadcast_model()
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2: RuntimeError: NCCL communicator was aborted on rank 19.  Original reason for failure was: [Rank 19] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+5:     dist.broadcast(p,
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     dist.broadcast(p,
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:         return func(*args, **kwargs)    dist.broadcast(p,
+5: dist.broadcast(p,
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)        
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1: return torch.distributed.broadcast(tensor=tensor,return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     dist.broadcast(p,
+1:         return torch.distributed.broadcast(tensor=tensor,return func(*args, **kwargs)
+1: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     dist.broadcast(p,
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)    
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1: return torch.distributed.broadcast(tensor=tensor,
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:     dist.broadcast(p,
+1:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:     return torch.distributed.broadcast(tensor=tensor,  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: 
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:         return func(*args, **kwargs)return func(*args, **kwargs)
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:     return torch.distributed.broadcast(tensor=tensor,
+5:     return func(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     return func(*args, **kwargs)  File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5: 
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:     return torch.distributed.broadcast(tensor=tensor,
+6:     return torch.distributed.broadcast(tensor=tensor,
+1:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: work = group.broadcast([tensor], opts)
+5:     return func(*args, **kwargs)
+1: RuntimeError: NCCL communicator was aborted on rank 8.  Original reason for failure was: [Rank 8] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805989 milliseconds before timing out.
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:     work = group.broadcast([tensor], opts)
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)    
+1: RuntimeError: NCCL communicator was aborted on rank 10.  Original reason for failure was: [Rank 10] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+5: return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     work = group.broadcast([tensor], opts)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1: RuntimeError    : work = group.broadcast([tensor], opts)NCCL communicator was aborted on rank 11.  Original reason for failure was: [Rank 11] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: 
+5:     return torch.distributed.broadcast(tensor=tensor,    
+5: return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:         work = group.broadcast([tensor], opts)work = group.broadcast([tensor], opts)
+1: RuntimeError
+5: return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: : NCCL communicator was aborted on rank 13.  Original reason for failure was: [Rank 13] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1: RuntimeErrorRuntimeError: : NCCL communicator was aborted on rank 12.  Original reason for failure was: [Rank 12] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.NCCL communicator was aborted on rank 15.  Original reason for failure was: [Rank 15] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: 
+5:     return torch.distributed.broadcast(tensor=tensor,
+1:     work = group.broadcast([tensor], opts)
+6:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: RuntimeError: NCCL communicator was aborted on rank 9.  Original reason for failure was: [Rank 9] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+5:     return torch.distributed.broadcast(tensor=tensor,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:     return torch.distributed.broadcast(tensor=tensor,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:     return torch.distributed.broadcast(tensor=tensor,
+1:     work = group.broadcast([tensor], opts)
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6: return torch.distributed.broadcast(tensor=tensor,
+6:       File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6: return torch.distributed.broadcast(tensor=tensor,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1: RuntimeError: NCCL communicator was aborted on rank 14.  Original reason for failure was: [Rank 14] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+6:     return torch.distributed.broadcast(tensor=tensor,
+5:     work = group.broadcast([tensor], opts)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5: RuntimeError: NCCL communicator was aborted on rank 41.  Original reason for failure was: [Rank 41] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+6:     return torch.distributed.broadcast(tensor=tensor,
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:     work = group.broadcast([tensor], opts)
+5:     work = group.broadcast([tensor], opts)
+6: RuntimeError: NCCL communicator was aborted on rank 50.  Original reason for failure was: [Rank 50] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: RuntimeError: NCCL communicator was aborted on rank 47.  Original reason for failure was: [Rank 47] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+6: RuntimeError: NCCL communicator was aborted on rank 54.  Original reason for failure was: [Rank 54] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+5:     work = group.broadcast([tensor], opts)
+5:     work = group.broadcast([tensor], opts)
+5: RuntimeError: NCCL communicator was aborted on rank 40.  Original reason for failure was: [Rank 40] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: RuntimeError: NCCL communicator was aborted on rank 46.  Original reason for failure was: [Rank 46] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+5:     work = group.broadcast([tensor], opts)
+5: RuntimeError:     NCCL communicator was aborted on rank 42.  Original reason for failure was: [Rank 42] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.work = group.broadcast([tensor], opts)
+5: 
+5: RuntimeError: NCCL communicator was aborted on rank 43.  Original reason for failure was: [Rank 43] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+6: RuntimeError: NCCL communicator was aborted on rank 48.  Original reason for failure was: [Rank 48] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+6: RuntimeError: NCCL communicator was aborted on rank 53.  Original reason for failure was: [Rank 53] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+6:     work = group.broadcast([tensor], opts)RuntimeError
+6: : NCCL communicator was aborted on rank 52.  Original reason for failure was: [Rank 52] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805958 milliseconds before timing out.
+5:     work = group.broadcast([tensor], opts)
+6: RuntimeError: NCCL communicator was aborted on rank 55.  Original reason for failure was: [Rank 55] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+5: RuntimeError: NCCL communicator was aborted on rank 44.  Original reason for failure was: [Rank 44] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+6: RuntimeError: NCCL communicator was aborted on rank 49.  Original reason for failure was: [Rank 49] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6:     work = group.broadcast([tensor], opts)
+6: RuntimeError: NCCL communicator was aborted on rank 51.  Original reason for failure was: [Rank 51] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+0: Traceback (most recent call last):
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+0:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:     engine = PipelineEngine(args=args,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:     super().__init__(*super_args, **super_kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:     self._configure_distributed_model(model)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:     self._broadcast_model()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:     dist.broadcast(p,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     return func(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:     return torch.distributed.broadcast(tensor=tensor,
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:     work = group.broadcast([tensor], opts)
+0: RuntimeError: NCCL communicator was aborted on rank 5.  Original reason for failure was: [Rank 5] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805964 milliseconds before timing out.
+4: Traceback (most recent call last):
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+4:     main()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:     engine = PipelineEngine(args=args,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:     super().__init__(*super_args, **super_kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:     self._configure_distributed_model(model)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:     self._broadcast_model()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:     dist.broadcast(p,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:     return func(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:     return torch.distributed.broadcast(tensor=tensor,
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:     work = group.broadcast([tensor], opts)
+4: RuntimeError: NCCL communicator was aborted on rank 37.  Original reason for failure was: [Rank 37] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: Traceback (most recent call last):
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+5:     main()
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:     return f(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:     engine = PipelineEngine(args=args,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:     super().__init__(*super_args, **super_kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:     self._configure_distributed_model(model)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:     self._broadcast_model()
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:     dist.broadcast(p,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:     return func(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:     return torch.distributed.broadcast(tensor=tensor,
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:     work = group.broadcast([tensor], opts)
+5: RuntimeError: NCCL communicator was aborted on rank 45.  Original reason for failure was: [Rank 45] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7: terminate called after throwing an instance of '  what():  std::runtime_error[Rank 62] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.'
+7: 
+7:   what():  Fatal Python error: [Rank 57] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.Aborted
+7: 
+7: 
+7: Thread 0x000014b57592b4c0 (most recent call first):
+7: <no Python frame>
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x0000145fc31964c0 (most recent call first):
+7: <no Python frame>
+4: terminate called after throwing an instance of 'std::runtime_error'
+4:   what():  [Rank 35] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: Fatal Python error: Aborted
+4: 
+4: Thread 0x000014f4a2c2e4c0 (most recent call first):
+4: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 5] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805964 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x0000153171c624c0 (most recent call first):
+0: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 40] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x0000153e2cc1b4c0 (most recent call first):
+5: <no Python frame>
+2: terminate called after throwing an instance of 'std::runtime_error'
+2: terminate called after throwing an instance of 'std::runtime_error'
+2:   what():  [Rank 20] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   what():  [Rank 19] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: Fatal Python error: Aborted
+2: 
+2: Thread 0x00001476424cb4c0Fatal Python error:  (most recent call first):
+2: Aborted<no Python frame>
+2: 
+2: 
+2: Thread 0x0000150a034324c0 (most recent call first):
+2: <no Python frame>
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():  [Rank 26] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805982 milliseconds before timing out.
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():  [Rank 24] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x0000153fe29044c0 (most recent call first):
+3: <no Python frame>
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():  [Rank 25] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():  [Rank 31] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x000014b6f5b024c0 (most recent call first):
+3: <no Python frame>
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x000014980f92a4c0 (most recent call first):
+3: <no Python frame>
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x000014b2acfc74c0 (most recent call first):
+3: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 60] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805943 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x00001488e68ea4c0 (most recent call first):
+7: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 56] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805950 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x000015042fdda4c0 (most recent call first):
+7: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 58] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x000014e55e5be4c0 (most recent call first):
+7: <no Python frame>
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: terminate called after throwing an instance of 'std::runtime_error'
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():  [Rank 28] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+4:   what():  [Rank 36] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+4: terminate called after throwing an instance of 'Fatal Python error: std::runtime_errorAbortedterminate called after throwing an instance of ''
+4: 
+4: 
+4: std::runtime_error'
+4: Thread 0x000014d060af24c0 (most recent call first):
+4: <no Python frame>
+3: Fatal Python error: Aborted
+3: 
+4:   what():    what():  [Rank 37] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.terminate called after throwing an instance of '[Rank 39] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+3: Thread 0x0000148c9c44b4c0 (most recent call first):
+3: <no Python frame>
+4: std::runtime_error
+4: '
+4: terminate called after throwing an instance of 'std::runtime_error'
+4:   what():  [Rank 34] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4: Fatal Python error: Fatal Python error: AbortedAborted
+4: 
+4: 
+4: 
+4: Thread 0x  what():  Thread 0x00001508f3f0c4c0[Rank 33] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out. (most recent call first):
+4: 00001483ab0c54c0
+4:  (most recent call first):
+4: <no Python frame>
+4: <no Python frame>
+4: Fatal Python error: Abortedterminate called after throwing an instance of '
+4: 
+4: std::runtime_error'
+4: Thread 0x000014763f1fc4c0 (most recent call first):
+4: <no Python frame>
+4:   what():  Fatal Python error: [Rank 32] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.Aborted
+4: 
+4: 
+4: Thread 0x00001517012504c0 (most recent call first):
+4: <no Python frame>
+4: Fatal Python error: Aborted
+4: 
+4: Thread 0x000014c0dfc254c0 (most recent call first):
+4: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 1] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x000014cf8120b4c0 (most recent call first):
+0: <no Python frame>
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: terminate called after throwing an instance of '[E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+2: std::runtime_error[E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: '
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+2:   what():  [Rank 22] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2: terminate called after throwing an instance of 'std::runtime_error'
+2: terminate called after throwing an instance of 'std::runtime_error'
+2: Fatal Python error: Aborted
+2: 
+2:   what():  Thread 0x[Rank 21] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.0000152d924604c0  what():  
+2:  (most recent call first):
+2: [Rank 18] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.<no Python frame>
+2: 
+2: Fatal Python error: Aborted
+2: 
+2: Fatal Python error: terminate called after throwing an instance of 'Thread 0xAbortedstd::runtime_error00001504c0ae74c0
+2: 
+2: '
+2:  (most recent call first):
+2: <no Python frame>
+2: Thread 0x0000145da05aa4c0 (most recent call first):
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py"terminate called after throwing an instance of ', line 1404 in std::runtime_errorbroadcast'
+2:   what():  
+2: [Rank 23] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.  File "
+2: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70 in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231 in broadcast
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line terminate called after throwing an instance of '126  what():  std::runtime_error in [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.'
+2: log_wrapper
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py"Fatal Python error: , line 1007Aborted in 
+2: 
+2: _broadcast_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092 in _configure_distributed_model
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py"  what():  , line 287 in [Rank 17] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.__init__
+2: 
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py"Fatal Python error: , line 59Aborted in 
+2: 
+2: __init__
+2:   File Thread 0x00001456785424c0 (most recent call first):
+2: <no Python frame>
+2: Fatal Python error: Aborted
+2: 
+2: Thread 0x0000145601c094c0 (most recent call first):
+2: <no Python frame>
+2: "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137 in initialize
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424 in setup_model_and_optimizer
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141 in pretrain
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231 in main
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346 in wrapper
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235 in <module>
+2: Thread 0x000014f7d63664c0 (most recent call first):
+2: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 43] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x0000145dabeac4c0 (most recent call first):
+5: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 59] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x000014fd53b624c0 (most recent call first):
+7: <no Python frame>
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1: terminate called after throwing an instance of 'std::runtime_error'
+1:   what():  [Rank 9] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1:   what():  [Rank 8] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805989 milliseconds before timing out.
+1: Fatal Python error: Aborted
+1: 
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x000014d44428b4c0 (most recent call first):
+1: <no Python frame>
+1: Thread 0x000014f31ef3b4c0 (most recent call first):
+1: <no Python frame>
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+6: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+6: terminate called after throwing an instance of 'std::runtime_error'
+6:   what():  [Rank 55] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6: terminate called after throwing an instance of 'terminate called after throwing an instance of 'std::runtime_error'
+6: std::runtime_error'
+6:   what():  [Rank 53] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.Fatal Python error: 
+6: Aborted  what():  
+6: 
+6: [Rank 54] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6: Thread 0x000014abdd3564c0 (most recent call first):
+6: <no Python frame>
+6: terminate called after throwing an instance of 'std::runtime_error'
+6: Fatal Python error: terminate called after throwing an instance of 'AbortedFatal Python error: 
+6: 
+6: std::runtime_errorAbortedterminate called after throwing an instance of '
+6: 
+6: '
+6: Thread 0xstd::runtime_error000014a14ba254c0Thread 0x'
+6:  (most recent call first):
+6: 00001456f51904c0<no Python frame>
+6:   what():   (most recent call first):
+6: [Rank 49] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.<no Python frame>
+6: 
+6:   what():  terminate called after throwing an instance of '[Rank 50] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+6: std::runtime_error'
+6:   what():  [Rank 52] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805958 milliseconds before timing out.
+6: Fatal Python error: Aborted
+6: 
+6: terminate called after throwing an instance of 'Thread 0xstd::runtime_error000014df375524c0 (most recent call first):
+6: '
+6: <no Python frame>
+6: Fatal Python error: Aborted
+6: 
+6: Fatal Python error: Aborted
+6: 
+6: Thread 0xThread 0x00001537f073a4c0 (most recent call first):
+6: 000014cde78454c0<no Python frame>
+6:   what():   (most recent call first):
+6: <no Python frame>
+6: [Rank 51] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6: Fatal Python error: Aborted
+6: 
+6: Thread 0x00001494b5a2d4c0 (most recent call first):
+6: <no Python frame>
+6:   what():  [Rank 48] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6: Fatal Python error: Aborted
+6: 
+6: Thread 0x000014ff3524c4c0 (most recent call first):
+6: <no Python frame>
+7: Traceback (most recent call last):
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 235, in <module>
+7:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:     pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:     model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:     model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:     engine = PipelineEngine(args=args,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:     super().__init__(*super_args, **super_kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:     self._configure_distributed_model(model)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:     self._broadcast_model()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:     dist.broadcast(p,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:     return func(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:     return torch.distributed.broadcast(tensor=tensor,
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:     work = group.broadcast([tensor], opts)
+7: RuntimeError: NCCL communicator was aborted on rank 61.  Original reason for failure was: [Rank 61] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1: terminate called after throwing an instance of 'std::runtime_error'
+1:   what():  [Rank 11] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1:   what():  [Rank 12] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x0000153b7f0784c0 (most recent call first):
+1: <no Python frame>
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x00001485f56e54c0 (most recent call first):
+1: <no Python frame>
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1:   what():  [Rank 15] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x000014ed56dab4c0 (most recent call first):
+1: <no Python frame>
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1:   what():  [Rank 13] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x00001492a35e04c0 (most recent call first):
+1: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 3] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x0000145c01ea34c0 (most recent call first):
+0: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 46] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x000014ef358c14c0 (most recent call first):
+5: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 42] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x0000151ecc4674c0 (most recent call first):
+5: <no Python frame>
+4: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+4: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+4: terminate called after throwing an instance of 'std::runtime_error'
+4:   what():  [Rank 38] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1806000 milliseconds before timing out.
+4: Fatal Python error: Aborted
+4: 
+4: Thread 0x000015024c6774c0 (most recent call first):
+4: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 44] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x00001526794e14c0 (most recent call first):
+5: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 47] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x00001491e460a4c0 (most recent call first):
+5: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 7] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x000014651dfae4c0 (most recent call first):
+0: <no Python frame>
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1:   what():  [Rank 14] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1: terminate called after throwing an instance of 'std::runtime_error'
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x000014fe16bc64c0 (most recent call first):
+1:   what():  <no Python frame>
+1: [Rank 10] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+1: Fatal Python error: Aborted
+1: 
+1: Thread 0x0000149bff33c4c0 (most recent call first):
+1: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 4] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x000014c0b43904c0 (most recent call first):
+0: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 2] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: terminate called after throwing an instance of 'Fatal Python error: std::runtime_errorAborted'
+0: 
+0: 
+0: Thread 0x000014c995eaf4c0 (most recent call first):
+0: <no Python frame>
+0:   what():  [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=50, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x000014cf151bf4c0 (most recent call first):
+0: <no Python frame>
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+5: terminate called after throwing an instance of 'std::runtime_error'
+5: terminate called after throwing an instance of 'std::runtime_error'
+5:   what():  [Rank 45] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   what():  [Rank 41] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x00001478a86444c0 (most recent call first):
+5: <no Python frame>
+5: Fatal Python error: Aborted
+5: 
+5: Thread 0x000014b96def24c0 (most recent call first):
+5: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 63] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x0000146d240d84c0 (most recent call first):
+7: <no Python frame>
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+3: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+3: terminate called after throwing an instance of 'std::runtime_error'
+3: terminate called after throwing an instance of 'std::runtime_error'
+3:   what():    what():  [Rank 29] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.[Rank 30] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3: 
+3: terminate called after throwing an instance of 'std::runtime_error'
+3: Fatal Python error: Aborted
+3: 
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x  what():  00001515b078b4c0 (most recent call first):
+3: [Rank 27] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.<no Python frame>
+3: Thread 0x
+3: 0000151c9b3a44c0 (most recent call first):
+3: <no Python frame>
+3: Fatal Python error: Aborted
+3: 
+3: Thread 0x000014b8f9da74c0 (most recent call first):
+3: <no Python frame>
+7: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+7: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+7: terminate called after throwing an instance of 'std::runtime_error'
+7:   what():  [Rank 61] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7: Fatal Python error: Aborted
+7: 
+7: Thread 0x000014dd6c3ca4c0 (most recent call first):
+7: <no Python frame>
+0: [E ProcessGroupNCCL.cpp:456] Some NCCL operations have failed or timed out. Due to the asynchronous nature of CUDA kernels, subsequent GPU operations might run on corrupted/incomplete data.
+0: [E ProcessGroupNCCL.cpp:461] To avoid data inconsistency, we are taking the entire process down.
+0: terminate called after throwing an instance of 'std::runtime_error'
+0:   what():  [Rank 6] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805997 milliseconds before timing out.
+0: Fatal Python error: Aborted
+0: 
+0: Thread 0x000014e69b7a84c0 (most recent call first):
+0: <no Python frame>
+2: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 79242) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+7: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 78891) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+5: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 82244) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+4: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 83777) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+3: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 79524) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+1: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 130831) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+0: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 39792) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+6: ERROR:torch.distributed.elastic.multiprocessing.api:failed (exitcode: -6) local_rank: 0 (pid: 80705) of binary: /pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/bin/python
+3: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/0/error.json)
+1: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/0/error.json)
+3: Traceback (most recent call last):
+3:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+1: Traceback (most recent call last):
+1:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+3:     return _run_code(code, main_globals, None,
+3:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+1:     return _run_code(code, main_globals, None,
+1:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+0: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/0/error.json)
+0: Traceback (most recent call last):
+1:     exec(code, run_globals)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+0:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+3:     exec(code, run_globals)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+0:     return _run_code(code, main_globals, None,
+0:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+0:     exec(code, run_globals)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+3:     main()
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/0/error.json)
+3:     return f(*args, **kwargs)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+7: Traceback (most recent call last):
+7:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+1:     main()
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     return _run_code(code, main_globals, None,
+7:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+0:     main()
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     exec(code, run_globals)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+1:     return f(*args, **kwargs)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+0:     return f(*args, **kwargs)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+3:     run(args)
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+1:     run(args)
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+3:     elastic_launch(
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+7:     main()
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:     run(args)
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+1:     elastic_launch(
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+7:     return f(*args, **kwargs)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+0:     elastic_launch(
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+3:     return launch_agent(self._config, self._entrypoint, list(args))
+3:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+1:     return launch_agent(self._config, self._entrypoint, list(args))
+1:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+3:     raise ChildFailedError(
+7:     run(args)
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+0:     return launch_agent(self._config, self._entrypoint, list(args))
+0:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+3: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+3: ============================================================
+3: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+3: ------------------------------------------------------------
+3: Failures:
+3: [1]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 25 (local_rank: 1)
+3:   exitcode  : -6 (pid: 79525)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/1/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:     elastic_launch(
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+1:     raise ChildFailedError(
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 25.  Original reason for failure was: [Rank 25] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3:   
+3: [2]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 26 (local_rank: 2)
+3:   exitcode  : -6 (pid: 79526)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/2/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+1: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+1: ============================================================
+1: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+1: ------------------------------------------------------------
+1: Failures:
+1: [1]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 9 (local_rank: 1)
+1:   exitcode  : -6 (pid: 130832)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/1/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     raise ChildFailedError(
+0: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+0: ============================================================
+0: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+0: ------------------------------------------------------------
+0: Failures:
+0: [1]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 1 (local_rank: 1)
+0:   exitcode  : -6 (pid: 39793)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/1/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 26.  Original reason for failure was: [Rank 26] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805982 milliseconds before timing out.
+3:   
+3: [3]:
+3:   time      : 2023-03-17_10:58:00
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+3:   host      : nid005361
+3:   rank      : 27 (local_rank: 3)
+3:   exitcode  : -6 (pid: 79527)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/3/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     return launch_agent(self._config, self._entrypoint, list(args))
+7:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 9.  Original reason for failure was: [Rank 9] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1:   
+1: [2]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 10 (local_rank: 2)
+1:   exitcode  : -6 (pid: 130833)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/2/error.json
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 1.  Original reason for failure was: [Rank 1] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: [2]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 2 (local_rank: 2)
+0:   exitcode  : -6 (pid: 39794)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/2/error.json
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 27.  Original reason for failure was: [Rank 27] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3:   
+3: [4]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 28 (local_rank: 4)
+3:   exitcode  : -6 (pid: 79528)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/4/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+7:     raise ChildFailedError(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 10.  Original reason for failure was: [Rank 10] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+1:   
+1: [3]:
+1:   time      : 2023-03-17_10:58:00
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 2.  Original reason for failure was: [Rank 2] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: [3]:
+0:   time      : 2023-03-17_10:58:00
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+7: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+7: ============================================================
+7: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+7: ------------------------------------------------------------
+7: Failures:
+7: [1]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+7:   rank      : 57 (local_rank: 1)
+7:   exitcode  : -6 (pid: 78892)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/1/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:   host      : nid005359
+1:   rank      : 11 (local_rank: 3)
+1:   exitcode  : -6 (pid: 130834)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/3/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   host      : nid005358
+0:   rank      : 3 (local_rank: 3)
+0:   exitcode  : -6 (pid: 39795)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/3/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 28.  Original reason for failure was: [Rank 28] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3:   
+3: [5]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 29 (local_rank: 5)
+3:   exitcode  : -6 (pid: 79529)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/5/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 57.  Original reason for failure was: [Rank 57] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7:   
+7: [2]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+7:   rank      : 58 (local_rank: 2)
+7:   exitcode  : -6 (pid: 78893)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/2/error.json
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 11.  Original reason for failure was: [Rank 11] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1:   
+1: [4]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 12 (local_rank: 4)
+1:   exitcode  : -6 (pid: 130835)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/4/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 3.  Original reason for failure was: [Rank 3] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: [4]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 4 (local_rank: 4)
+0:   exitcode  : -6 (pid: 39796)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/4/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+3:       return func(*args, **kwargs)
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 29.  Original reason for failure was: [Rank 29] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3:   
+3: [6]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+3:   rank      : 30 (local_rank: 6)
+3:   exitcode  : -6 (pid: 79530)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/6/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+4: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/0/error.json)
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 58.  Original reason for failure was: [Rank 58] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7:   
+7: [3]:
+7:   time      : 2023-03-17_10:58:00
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 12.  Original reason for failure was: [Rank 12] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1:   
+1: [5]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 13 (local_rank: 5)
+1:   exitcode  : -6 (pid: 130836)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/5/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 4.  Original reason for failure was: [Rank 4] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: [5]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 5 (local_rank: 5)
+0:   exitcode  : -6 (pid: 39797)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/5/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:   host      : nid005365
+7:   rank      : 59 (local_rank: 3)
+7:   exitcode  : -6 (pid: 78894)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/3/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:   RuntimeError: NCCL communicator was aborted on rank 30.  Original reason for failure was: [Rank 30] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3:   
+3: [7]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 31 (local_rank: 7)
+3:   exitcode  : -6 (pid: 79531)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/7/error.json
+3:   traceback : Traceback (most recent call last):
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+1:       return func(*args, **kwargs)
+0:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+4: Traceback (most recent call last):
+4:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 13.  Original reason for failure was: [Rank 13] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805973 milliseconds before timing out.
+1:   
+1: [6]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 5.  Original reason for failure was: [Rank 5] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805964 milliseconds before timing out.
+0:   
+0: [6]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:     return _run_code(code, main_globals, None,
+4:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 59.  Original reason for failure was: [Rank 59] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7:   
+7: [4]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+7:   rank      : 60 (local_rank: 4)
+7:   exitcode  : -6 (pid: 78895)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/4/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+1:   rank      : 14 (local_rank: 6)
+1:   exitcode  : -6 (pid: 130837)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/6/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:   rank      : 6 (local_rank: 6)
+0:   exitcode  : -6 (pid: 39798)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/6/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 31.  Original reason for failure was: [Rank 31] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805987 milliseconds before timing out.
+3:   
+3: ------------------------------------------------------------
+3: Root Cause (first observed failure):
+3: [0]:
+3:   time      : 2023-03-17_10:58:00
+3:   host      : nid005361
+3:   rank      : 24 (local_rank: 0)
+3:   exitcode  : -6 (pid: 79524)
+4:     exec(code, run_globals)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+3:   error_file: /tmp/torchelastic_1_4jdb1b/none_fx92baoh/attempt_0/0/error.json
+3:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+3:       return f(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+3:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+3:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+3:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+3:       engine = PipelineEngine(args=args,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+3:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+3:       self._configure_distributed_model(model)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+3:       self._broadcast_model()
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 60.  Original reason for failure was: [Rank 60] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805943 milliseconds before timing out.
+7:   
+7: [5]:
+7:   time      : 2023-03-17_10:58:01
+7:   host      : nid005365
+7:   rank      : 61 (local_rank: 5)
+7:   exitcode  : -6 (pid: 78896)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/5/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+1:   RuntimeError: NCCL communicator was aborted on rank 14.  Original reason for failure was: [Rank 14] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1:   
+1: [7]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 15 (local_rank: 7)
+1:   exitcode  : -6 (pid: 130838)
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/7/error.json
+1:   traceback : Traceback (most recent call last):
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:   RuntimeError: NCCL communicator was aborted on rank 6.  Original reason for failure was: [Rank 6] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805997 milliseconds before timing out.
+0:   
+0: [7]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 7 (local_rank: 7)
+0:   exitcode  : -6 (pid: 39799)
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/7/error.json
+0:   traceback : Traceback (most recent call last):
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+3:       dist.broadcast(p,
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+3:       return func(*args, **kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+3:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+3:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+3:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+3:       work = group.broadcast([tensor], opts)
+3:   RuntimeError: NCCL communicator was aborted on rank 24.  Original reason for failure was: [Rank 24] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805983 milliseconds before timing out.
+3:   
+3: ============================================================
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return func(*args, **kwargs)
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 15.  Original reason for failure was: [Rank 15] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805988 milliseconds before timing out.
+1:   
+1: ------------------------------------------------------------
+1: Root Cause (first observed failure):
+1: [0]:
+1:   time      : 2023-03-17_10:58:00
+1:   host      : nid005359
+1:   rank      : 8 (local_rank: 0)
+1:   exitcode  : -6 (pid: 130831)
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 7.  Original reason for failure was: [Rank 7] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: ------------------------------------------------------------
+0: Root Cause (first observed failure):
+0: [0]:
+0:   time      : 2023-03-17_10:58:00
+0:   host      : nid005358
+0:   rank      : 0 (local_rank: 0)
+0:   exitcode  : -6 (pid: 39792)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 61.  Original reason for failure was: [Rank 61] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7:   
+7: [6]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+1:   error_file: /tmp/torchelastic_9id5al_8/none_xu0ou91c/attempt_0/0/error.json
+1:   traceback : Traceback (most recent call last):
+0:   error_file: /tmp/torchelastic_y1xliq25/none_3q568_3e/attempt_0/0/error.json
+0:   traceback : Traceback (most recent call last):
+7:   rank      : 62 (local_rank: 6)
+7:   exitcode  : -6 (pid: 78897)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/6/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+1:       return f(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+1:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+1:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+1:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+0:       return f(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+0:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+0:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+0:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+1:       engine = PipelineEngine(args=args,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+1:       super().__init__(*super_args, **super_kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+1:       self._configure_distributed_model(model)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+1:       self._broadcast_model()
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+0:       engine = PipelineEngine(args=args,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+0:       super().__init__(*super_args, **super_kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+0:       self._configure_distributed_model(model)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+0:       self._broadcast_model()
+4:     main()
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+1:       dist.broadcast(p,
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+1:       return func(*args, **kwargs)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+1:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+1:       return torch.distributed.broadcast(tensor=tensor,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+0:       dist.broadcast(p,
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+0:       return func(*args, **kwargs)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+0:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+0:       return torch.distributed.broadcast(tensor=tensor,
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+1:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+1:       work = group.broadcast([tensor], opts)
+1:   RuntimeError: NCCL communicator was aborted on rank 8.  Original reason for failure was: [Rank 8] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=42, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805989 milliseconds before timing out.
+1:   
+1: ============================================================
+0:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+0:       work = group.broadcast([tensor], opts)
+0:   RuntimeError: NCCL communicator was aborted on rank 0.  Original reason for failure was: [Rank 0] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=50, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805990 milliseconds before timing out.
+0:   
+0: ============================================================
+7:   RuntimeError: NCCL communicator was aborted on rank 62.  Original reason for failure was: [Rank 62] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805953 milliseconds before timing out.
+7:   
+7: [7]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+7:   rank      : 63 (local_rank: 7)
+7:   exitcode  : -6 (pid: 78898)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/7/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     return f(*args, **kwargs)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 63.  Original reason for failure was: [Rank 63] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805952 milliseconds before timing out.
+7:   
+7: ------------------------------------------------------------
+7: Root Cause (first observed failure):
+7: [0]:
+7:   time      : 2023-03-17_10:58:00
+7:   host      : nid005365
+7:   rank      : 56 (local_rank: 0)
+7:   exitcode  : -6 (pid: 78891)
+7:   error_file: /tmp/torchelastic_rchtnbjh/none_hhbsa1ko/attempt_0/0/error.json
+7:   traceback : Traceback (most recent call last):
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+7:       return f(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+7:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+7:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+7:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+7:       engine = PipelineEngine(args=args,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+7:       super().__init__(*super_args, **super_kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+7:       self._configure_distributed_model(model)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+7:       self._broadcast_model()
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+7:       dist.broadcast(p,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+7:       return func(*args, **kwargs)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+7:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+7:       return torch.distributed.broadcast(tensor=tensor,
+7:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+7:       work = group.broadcast([tensor], opts)
+7:   RuntimeError: NCCL communicator was aborted on rank 56.  Original reason for failure was: [Rank 56] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805950 milliseconds before timing out.
+7:   
+7: ============================================================
+4:     run(args)
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+4:     elastic_launch(
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+4:     return launch_agent(self._config, self._entrypoint, list(args))
+4:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+4:     raise ChildFailedError(
+4: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+4: ============================================================
+4: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+4: ------------------------------------------------------------
+4: Failures:
+4: [1]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 33 (local_rank: 1)
+4:   exitcode  : -6 (pid: 83778)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/1/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 33.  Original reason for failure was: [Rank 33] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+4:   
+4: [2]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 34 (local_rank: 2)
+4:   exitcode  : -6 (pid: 83779)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/2/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 34.  Original reason for failure was: [Rank 34] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4:   
+4: [3]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 35 (local_rank: 3)
+4:   exitcode  : -6 (pid: 83780)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/3/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 35.  Original reason for failure was: [Rank 35] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4:   
+4: [4]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 36 (local_rank: 4)
+4:   exitcode  : -6 (pid: 83781)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/4/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 36.  Original reason for failure was: [Rank 36] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+4:   
+4: [5]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 37 (local_rank: 5)
+4:   exitcode  : -6 (pid: 83782)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/5/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 37.  Original reason for failure was: [Rank 37] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805975 milliseconds before timing out.
+4:   
+4: [6]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 38 (local_rank: 6)
+4:   exitcode  : -6 (pid: 83783)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/6/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 38.  Original reason for failure was: [Rank 38] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1806000 milliseconds before timing out.
+4:   
+4: [7]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 39 (local_rank: 7)
+4:   exitcode  : -6 (pid: 83784)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/7/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 39.  Original reason for failure was: [Rank 39] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4:   
+4: ------------------------------------------------------------
+4: Root Cause (first observed failure):
+4: [0]:
+4:   time      : 2023-03-17_10:58:00
+4:   host      : nid005362
+4:   rank      : 32 (local_rank: 0)
+4:   exitcode  : -6 (pid: 83777)
+4:   error_file: /tmp/torchelastic_0mkj7gbo/none_tfo6glno/attempt_0/0/error.json
+4:   traceback : Traceback (most recent call last):
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+4:       return f(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+4:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+4:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+4:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+4:       engine = PipelineEngine(args=args,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+4:       super().__init__(*super_args, **super_kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+4:       self._configure_distributed_model(model)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+4:       self._broadcast_model()
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+4:       dist.broadcast(p,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+4:       return func(*args, **kwargs)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+4:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+4:       return torch.distributed.broadcast(tensor=tensor,
+4:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+4:       work = group.broadcast([tensor], opts)
+4:   RuntimeError: NCCL communicator was aborted on rank 32.  Original reason for failure was: [Rank 32] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805974 milliseconds before timing out.
+4:   
+4: ============================================================
+6: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/0/error.json)
+6: Traceback (most recent call last):
+6:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+6:     return _run_code(code, main_globals, None,
+6:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+6:     exec(code, run_globals)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+6:     main()
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:     return f(*args, **kwargs)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+6:     run(args)
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+6:     elastic_launch(
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+6:     return launch_agent(self._config, self._entrypoint, list(args))
+6:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+6:     raise ChildFailedError(
+6: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+6: ============================================================
+6: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+6: ------------------------------------------------------------
+6: Failures:
+6: [1]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 49 (local_rank: 1)
+6:   exitcode  : -6 (pid: 80706)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/1/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 49.  Original reason for failure was: [Rank 49] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6:   
+6: [2]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 50 (local_rank: 2)
+6:   exitcode  : -6 (pid: 80707)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/2/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 50.  Original reason for failure was: [Rank 50] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+6:   
+6: [3]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 51 (local_rank: 3)
+6:   exitcode  : -6 (pid: 80708)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/3/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 51.  Original reason for failure was: [Rank 51] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6:   
+6: [4]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 52 (local_rank: 4)
+6:   exitcode  : -6 (pid: 80709)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/4/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 52.  Original reason for failure was: [Rank 52] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805958 milliseconds before timing out.
+6:   
+6: [5]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 53 (local_rank: 5)
+6:   exitcode  : -6 (pid: 80710)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/5/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 53.  Original reason for failure was: [Rank 53] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6:   
+6: [6]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 54 (local_rank: 6)
+6:   exitcode  : -6 (pid: 80711)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/6/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/0/error.json)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 54.  Original reason for failure was: [Rank 54] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805971 milliseconds before timing out.
+6:   
+6: [7]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 55 (local_rank: 7)
+6:   exitcode  : -6 (pid: 80712)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/7/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2: Traceback (most recent call last):
+2:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+2:     return _run_code(code, main_globals, None,
+2:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 55.  Original reason for failure was: [Rank 55] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+6:   
+6: ------------------------------------------------------------
+6: Root Cause (first observed failure):
+6: [0]:
+6:   time      : 2023-03-17_10:58:00
+6:   host      : nid005364
+6:   rank      : 48 (local_rank: 0)
+6:   exitcode  : -6 (pid: 80705)
+6:   error_file: /tmp/torchelastic_re932cu8/none_2y4skvxw/attempt_0/0/error.json
+6:   traceback : Traceback (most recent call last):
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+6:       return f(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+6:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+6:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+6:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     exec(code, run_globals)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+6:       engine = PipelineEngine(args=args,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+6:       super().__init__(*super_args, **super_kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+6:       self._configure_distributed_model(model)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+6:       self._broadcast_model()
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+6:       dist.broadcast(p,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+6:       return func(*args, **kwargs)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+6:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+6:       return torch.distributed.broadcast(tensor=tensor,
+6:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+6:       work = group.broadcast([tensor], opts)
+6:   RuntimeError: NCCL communicator was aborted on rank 48.  Original reason for failure was: [Rank 48] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805972 milliseconds before timing out.
+6:   
+6: ============================================================
+2:     main()
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:     return f(*args, **kwargs)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+2:     run(args)
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+2:     elastic_launch(
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+2:     return launch_agent(self._config, self._entrypoint, list(args))
+2:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+2:     raise ChildFailedError(
+2: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+2: ============================================================
+2: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+2: ------------------------------------------------------------
+2: Failures:
+2: [1]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 17 (local_rank: 1)
+2:   exitcode  : -6 (pid: 79243)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/1/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 17.  Original reason for failure was: [Rank 17] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805970 milliseconds before timing out.
+2:   
+2: [2]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 18 (local_rank: 2)
+2:   exitcode  : -6 (pid: 79244)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/2/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 18.  Original reason for failure was: [Rank 18] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   
+2: [3]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 19 (local_rank: 3)
+2:   exitcode  : -6 (pid: 79245)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/3/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 19.  Original reason for failure was: [Rank 19] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   
+2: [4]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 20 (local_rank: 4)
+2:   exitcode  : -6 (pid: 79246)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/4/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 20.  Original reason for failure was: [Rank 20] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   
+2: [5]:
+2:   time      : 2023-03-17_10:58:02
+2:   host      : nid005360
+2:   rank      : 21 (local_rank: 5)
+2:   exitcode  : -6 (pid: 79247)
+2:   error_file: <N/A>
+2:   traceback : Signal 6 (SIGABRT) received by PID 79247
+2: [6]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 22 (local_rank: 6)
+2:   exitcode  : -6 (pid: 79248)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/6/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 22.  Original reason for failure was: [Rank 22] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   
+2: [7]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 23 (local_rank: 7)
+2:   exitcode  : -6 (pid: 79249)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/7/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 23.  Original reason for failure was: [Rank 23] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805965 milliseconds before timing out.
+2:   
+2: ------------------------------------------------------------
+2: Root Cause (first observed failure):
+2: [0]:
+2:   time      : 2023-03-17_10:58:00
+2:   host      : nid005360
+2:   rank      : 16 (local_rank: 0)
+2:   exitcode  : -6 (pid: 79242)
+2:   error_file: /tmp/torchelastic_2hayt3oa/none_85zf9u3h/attempt_0/0/error.json
+2:   traceback : Traceback (most recent call last):
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+2:       return f(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+2:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+2:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+2:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+2:       engine = PipelineEngine(args=args,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+2:       super().__init__(*super_args, **super_kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+2:       self._configure_distributed_model(model)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+2:       self._broadcast_model()
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+2:       dist.broadcast(p,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+2:       return func(*args, **kwargs)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+2:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+2:       return torch.distributed.broadcast(tensor=tensor,
+2:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+2:       work = group.broadcast([tensor], opts)
+2:   RuntimeError: NCCL communicator was aborted on rank 16.  Original reason for failure was: [Rank 16] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805966 milliseconds before timing out.
+2:   
+2: ============================================================
+5: ERROR:torch.distributed.elastic.multiprocessing.errors.error_handler:no error file defined for parent, to copy child error file (/tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/0/error.json)
+5: Traceback (most recent call last):
+5:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 197, in _run_module_as_main
+5:     return _run_code(code, main_globals, None,
+5:   File "/opt/cray/pe/python/3.9.12.1/lib/python3.9/runpy.py", line 87, in _run_code
+5:     exec(code, run_globals)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 766, in <module>
+5:     main()
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:     return f(*args, **kwargs)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 762, in main
+5:     run(args)
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/run.py", line 753, in run
+5:     elastic_launch(
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 132, in __call__
+5:     return launch_agent(self._config, self._entrypoint, list(args))
+5:   File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/launcher/api.py", line 246, in launch_agent
+5:     raise ChildFailedError(
+5: torch.distributed.elastic.multiprocessing.errors.ChildFailedError: 
+5: ============================================================
+5: Megatron-DeepSpeed/pretrain_gpt.py FAILED
+5: ------------------------------------------------------------
+5: Failures:
+5: [1]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 41 (local_rank: 1)
+5:   exitcode  : -6 (pid: 82245)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/1/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 41.  Original reason for failure was: [Rank 41] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5:   
+5: [2]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 42 (local_rank: 2)
+5:   exitcode  : -6 (pid: 82246)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/2/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 42.  Original reason for failure was: [Rank 42] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   
+5: [3]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 43 (local_rank: 3)
+5:   exitcode  : -6 (pid: 82247)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/3/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 43.  Original reason for failure was: [Rank 43] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   
+5: [4]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 44 (local_rank: 4)
+5:   exitcode  : -6 (pid: 82248)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/4/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 44.  Original reason for failure was: [Rank 44] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5:   
+5: [5]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 45 (local_rank: 5)
+5:   exitcode  : -6 (pid: 82249)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/5/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 45.  Original reason for failure was: [Rank 45] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   
+5: [6]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 46 (local_rank: 6)
+5:   exitcode  : -6 (pid: 82250)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/6/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 46.  Original reason for failure was: [Rank 46] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   
+5: [7]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 47 (local_rank: 7)
+5:   exitcode  : -6 (pid: 82251)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/7/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 47.  Original reason for failure was: [Rank 47] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805960 milliseconds before timing out.
+5:   
+5: ------------------------------------------------------------
+5: Root Cause (first observed failure):
+5: [0]:
+5:   time      : 2023-03-17_10:58:00
+5:   host      : nid005363
+5:   rank      : 40 (local_rank: 0)
+5:   exitcode  : -6 (pid: 82244)
+5:   error_file: /tmp/torchelastic_cnbvbhpd/none_q_n5z7ff/attempt_0/0/error.json
+5:   traceback : Traceback (most recent call last):
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/elastic/multiprocessing/errors/__init__.py", line 346, in wrapper
+5:       return f(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/pretrain_gpt.py", line 231, in main
+5:       pretrain(train_valid_test_datasets_provider, model_provider, forward_step,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 141, in pretrain
+5:       model, optimizer, lr_scheduler = setup_model_and_optimizer(model_provider)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/Megatron-DeepSpeed/megatron/training.py", line 424, in setup_model_and_optimizer
+5:       model, optimizer, _, lr_scheduler = deepspeed.initialize(
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/__init__.py", line 137, in initialize
+5:       engine = PipelineEngine(args=args,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/pipe/engine.py", line 59, in __init__
+5:       super().__init__(*super_args, **super_kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 287, in __init__
+5:       self._configure_distributed_model(model)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1092, in _configure_distributed_model
+5:       self._broadcast_model()
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/runtime/engine.py", line 1007, in _broadcast_model
+5:       dist.broadcast(p,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 126, in log_wrapper
+5:       return func(*args, **kwargs)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/comm.py", line 231, in broadcast
+5:       return cdb.broadcast(tensor=tensor, src=src, group=group, async_op=async_op)
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/deepspeed/comm/torch.py", line 70, in broadcast
+5:       return torch.distributed.broadcast(tensor=tensor,
+5:     File "/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/venv/lib/python3.9/site-packages/torch/distributed/distributed_c10d.py", line 1404, in broadcast
+5:       work = group.broadcast([tensor], opts)
+5:   RuntimeError: NCCL communicator was aborted on rank 40.  Original reason for failure was: [Rank 40] Watchdog caught collective operation timeout: WorkNCCL(SeqNum=41, OpType=BROADCAST, Timeout(ms)=1800000) ran for 1805961 milliseconds before timing out.
+5:   
+5: ============================================================
+srun: error: nid005365: task 7: Exited with exit code 1
+srun: launch/slurm: _step_signal: Terminating StepId=3328731.0
+srun: error: nid005358: task 0: Exited with exit code 1
+srun: error: nid005359: task 1: Exited with exit code 1
+srun: error: nid005361: task 3: Exited with exit code 1
+srun: error: nid005362: task 4: Exited with exit code 1
+srun: error: nid005360: task 2: Exited with exit code 1
+srun: error: nid005364: task 6: Exited with exit code 1
+srun: error: nid005363: task 5: Exited with exit code 1