End of training

all_results.json

@@ -1,5 +1,10 @@
 {
     "epoch": 0.9931662870159453,
+    "eval_loss": 2.7200307846069336,
+    "eval_runtime": 0.4837,
+    "eval_samples": 25,
+    "eval_samples_per_second": 43.417,
+    "eval_steps_per_second": 2.067,
     "total_flos": 14975678545920.0,
     "train_loss": 3.682411685996099,
     "train_runtime": 758.7359,

config.json

@@ -24,6 +24,6 @@
   "rope_theta": 10000.0,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.46.3",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 256000
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 0.9931662870159453,
+    "eval_loss": 2.7200307846069336,
+    "eval_runtime": 0.4837,
+    "eval_samples": 25,
+    "eval_samples_per_second": 43.417,
+    "eval_steps_per_second": 2.067
+}

runs/Nov21_21-47-04_main-fft-mistral-coding-0-0/events.out.tfevents.1732245462.main-fft-mistral-coding-0-0.546.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:41e1ba0bfe14305908a6e6e1200e5054210b9120b9b9355712ec2322d622969a
+size 354