Training in progress, epoch 2, checkpoint
Browse files- last-checkpoint/global_step834/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step834/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step834/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step834/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step834/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step834/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/model-00001-of-00004.safetensors +1 -1
- last-checkpoint/model-00002-of-00004.safetensors +1 -1
- last-checkpoint/model-00003-of-00004.safetensors +1 -1
- last-checkpoint/model-00004-of-00004.safetensors +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +648 -2
last-checkpoint/global_step834/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f4d631a1b650315fa7f47d494d5dc90813cfa1a228bbc9829c7e772fc0ee414
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step834/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4bd981a3eca97cc976039dc6467c9712d528ec4aea67c25a590460f07981139e
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step834/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:763444484d0640da859eaac1055fd9670e227fdab7063816c1ab3f418674fd5f
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step834/zero_pp_rank_0_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:024ace79dd673f7972bcda314a22238789f9b2c7ef7d25eb76567e5d3f324fa1
|
| 3 |
+
size 168021
|
last-checkpoint/global_step834/zero_pp_rank_1_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:46717b755c898ee645bde65296e2b1b0172ede54f0775e47bb2278d05db803f8
|
| 3 |
+
size 168021
|
last-checkpoint/global_step834/zero_pp_rank_2_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8cc94c6743e0f17155029d5e411e13115bbbc756d14639250e19dcd6d3fcf17
|
| 3 |
+
size 168021
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step834
|
last-checkpoint/model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4877660776
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:503a0542de7434fd94cf507a1454667cc4db309bcb5768fe92cc7048e5cf709b
|
| 3 |
size 4877660776
|
last-checkpoint/model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4932751008
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5435bf5d9f48362da28a7f0483244712eab1583416c1d93ba4ff4f8c9571990f
|
| 3 |
size 4932751008
|
last-checkpoint/model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4330865200
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c64867fad1455680c0ccba908bdb82f5ef2f574e48845eb1662eec4b64a7838f
|
| 3 |
size 4330865200
|
last-checkpoint/model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1089994880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:473a3519933c9b22c048af564bfcbc7f216e05e95e6fd261ec5743f527a9163b
|
| 3 |
size 1089994880
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:122ac94678b6917d31dc059510d8313aa8631655cb9d780b749ba1ea53548879
|
| 3 |
size 14768
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:911b3814b631205d77beb4597874c7b3e54a28967f5a2afad57483b742364341
|
| 3 |
size 14768
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:132ff809344340444ac3d547eaa795a9ce6b4dbf6ac473db7e4ccf506c2d41e9
|
| 3 |
size 14768
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b827597f074228bbdf970b7a8c86f735e50cc45cc1d15e8574037a53c30f9d7
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -653,6 +653,652 @@
|
|
| 653 |
"eval_samples_per_second": 17.17,
|
| 654 |
"eval_steps_per_second": 0.773,
|
| 655 |
"step": 417
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 656 |
}
|
| 657 |
],
|
| 658 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 834,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 653 |
"eval_samples_per_second": 17.17,
|
| 654 |
"eval_steps_per_second": 0.773,
|
| 655 |
"step": 417
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.0071942446043165,
|
| 659 |
+
"grad_norm": 3.0018225466088073,
|
| 660 |
+
"learning_rate": 3.693333333333333e-07,
|
| 661 |
+
"logits/chosen": -1.28125,
|
| 662 |
+
"logits/rejected": -1.765625,
|
| 663 |
+
"logps/chosen": -235.0,
|
| 664 |
+
"logps/rejected": -342.0,
|
| 665 |
+
"loss": 0.0515,
|
| 666 |
+
"rewards/accuracies": 0.987500011920929,
|
| 667 |
+
"rewards/chosen": 1.4453125,
|
| 668 |
+
"rewards/margins": 9.0,
|
| 669 |
+
"rewards/rejected": -7.53125,
|
| 670 |
+
"step": 420
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 1.0311750599520384,
|
| 674 |
+
"grad_norm": 0.5745094685784482,
|
| 675 |
+
"learning_rate": 3.6488888888888884e-07,
|
| 676 |
+
"logits/chosen": -1.34375,
|
| 677 |
+
"logits/rejected": -1.78125,
|
| 678 |
+
"logps/chosen": -268.0,
|
| 679 |
+
"logps/rejected": -356.0,
|
| 680 |
+
"loss": 0.0099,
|
| 681 |
+
"rewards/accuracies": 0.987500011920929,
|
| 682 |
+
"rewards/chosen": 1.703125,
|
| 683 |
+
"rewards/margins": 9.75,
|
| 684 |
+
"rewards/rejected": -8.0,
|
| 685 |
+
"step": 430
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 1.0551558752997603,
|
| 689 |
+
"grad_norm": 0.23864223290524259,
|
| 690 |
+
"learning_rate": 3.604444444444444e-07,
|
| 691 |
+
"logits/chosen": -1.2734375,
|
| 692 |
+
"logits/rejected": -1.734375,
|
| 693 |
+
"logps/chosen": -237.0,
|
| 694 |
+
"logps/rejected": -372.0,
|
| 695 |
+
"loss": 0.0096,
|
| 696 |
+
"rewards/accuracies": 1.0,
|
| 697 |
+
"rewards/chosen": 1.34375,
|
| 698 |
+
"rewards/margins": 9.625,
|
| 699 |
+
"rewards/rejected": -8.25,
|
| 700 |
+
"step": 440
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 1.079136690647482,
|
| 704 |
+
"grad_norm": 7.163267004525079,
|
| 705 |
+
"learning_rate": 3.5599999999999996e-07,
|
| 706 |
+
"logits/chosen": -1.3359375,
|
| 707 |
+
"logits/rejected": -1.71875,
|
| 708 |
+
"logps/chosen": -247.0,
|
| 709 |
+
"logps/rejected": -372.0,
|
| 710 |
+
"loss": 0.0294,
|
| 711 |
+
"rewards/accuracies": 1.0,
|
| 712 |
+
"rewards/chosen": 1.296875,
|
| 713 |
+
"rewards/margins": 10.0,
|
| 714 |
+
"rewards/rejected": -8.6875,
|
| 715 |
+
"step": 450
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 1.1031175059952039,
|
| 719 |
+
"grad_norm": 5.519803193762622,
|
| 720 |
+
"learning_rate": 3.5155555555555554e-07,
|
| 721 |
+
"logits/chosen": -1.3125,
|
| 722 |
+
"logits/rejected": -1.8046875,
|
| 723 |
+
"logps/chosen": -235.0,
|
| 724 |
+
"logps/rejected": -354.0,
|
| 725 |
+
"loss": 0.0251,
|
| 726 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 727 |
+
"rewards/chosen": 1.75,
|
| 728 |
+
"rewards/margins": 9.125,
|
| 729 |
+
"rewards/rejected": -7.40625,
|
| 730 |
+
"step": 460
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1270983213429258,
|
| 734 |
+
"grad_norm": 0.5909316496529458,
|
| 735 |
+
"learning_rate": 3.471111111111111e-07,
|
| 736 |
+
"logits/chosen": -1.3125,
|
| 737 |
+
"logits/rejected": -1.703125,
|
| 738 |
+
"logps/chosen": -247.0,
|
| 739 |
+
"logps/rejected": -350.0,
|
| 740 |
+
"loss": 0.0093,
|
| 741 |
+
"rewards/accuracies": 1.0,
|
| 742 |
+
"rewards/chosen": 1.4609375,
|
| 743 |
+
"rewards/margins": 9.5,
|
| 744 |
+
"rewards/rejected": -8.0625,
|
| 745 |
+
"step": 470
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.1510791366906474,
|
| 749 |
+
"grad_norm": 3.011810128815269,
|
| 750 |
+
"learning_rate": 3.4266666666666666e-07,
|
| 751 |
+
"logits/chosen": -1.3359375,
|
| 752 |
+
"logits/rejected": -1.8046875,
|
| 753 |
+
"logps/chosen": -242.0,
|
| 754 |
+
"logps/rejected": -362.0,
|
| 755 |
+
"loss": 0.0096,
|
| 756 |
+
"rewards/accuracies": 1.0,
|
| 757 |
+
"rewards/chosen": 1.6484375,
|
| 758 |
+
"rewards/margins": 10.1875,
|
| 759 |
+
"rewards/rejected": -8.5,
|
| 760 |
+
"step": 480
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.1750599520383693,
|
| 764 |
+
"grad_norm": 0.17942480332235972,
|
| 765 |
+
"learning_rate": 3.382222222222222e-07,
|
| 766 |
+
"logits/chosen": -1.3828125,
|
| 767 |
+
"logits/rejected": -1.71875,
|
| 768 |
+
"logps/chosen": -247.0,
|
| 769 |
+
"logps/rejected": -364.0,
|
| 770 |
+
"loss": 0.0059,
|
| 771 |
+
"rewards/accuracies": 1.0,
|
| 772 |
+
"rewards/chosen": 1.1875,
|
| 773 |
+
"rewards/margins": 9.1875,
|
| 774 |
+
"rewards/rejected": -8.0,
|
| 775 |
+
"step": 490
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 1.1990407673860912,
|
| 779 |
+
"grad_norm": 0.1553054654263466,
|
| 780 |
+
"learning_rate": 3.337777777777778e-07,
|
| 781 |
+
"logits/chosen": -1.296875,
|
| 782 |
+
"logits/rejected": -1.7734375,
|
| 783 |
+
"logps/chosen": -260.0,
|
| 784 |
+
"logps/rejected": -358.0,
|
| 785 |
+
"loss": 0.0064,
|
| 786 |
+
"rewards/accuracies": 1.0,
|
| 787 |
+
"rewards/chosen": 1.671875,
|
| 788 |
+
"rewards/margins": 11.0,
|
| 789 |
+
"rewards/rejected": -9.375,
|
| 790 |
+
"step": 500
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.223021582733813,
|
| 794 |
+
"grad_norm": 1.1092919311137273,
|
| 795 |
+
"learning_rate": 3.293333333333333e-07,
|
| 796 |
+
"logits/chosen": -1.3046875,
|
| 797 |
+
"logits/rejected": -1.6953125,
|
| 798 |
+
"logps/chosen": -232.0,
|
| 799 |
+
"logps/rejected": -376.0,
|
| 800 |
+
"loss": 0.0198,
|
| 801 |
+
"rewards/accuracies": 1.0,
|
| 802 |
+
"rewards/chosen": 1.265625,
|
| 803 |
+
"rewards/margins": 9.9375,
|
| 804 |
+
"rewards/rejected": -8.625,
|
| 805 |
+
"step": 510
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 1.2470023980815348,
|
| 809 |
+
"grad_norm": 0.37821643712530456,
|
| 810 |
+
"learning_rate": 3.248888888888889e-07,
|
| 811 |
+
"logits/chosen": -1.25,
|
| 812 |
+
"logits/rejected": -1.671875,
|
| 813 |
+
"logps/chosen": -245.0,
|
| 814 |
+
"logps/rejected": -354.0,
|
| 815 |
+
"loss": 0.0078,
|
| 816 |
+
"rewards/accuracies": 1.0,
|
| 817 |
+
"rewards/chosen": 1.15625,
|
| 818 |
+
"rewards/margins": 9.8125,
|
| 819 |
+
"rewards/rejected": -8.625,
|
| 820 |
+
"step": 520
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 1.2709832134292567,
|
| 824 |
+
"grad_norm": 0.3096451149240373,
|
| 825 |
+
"learning_rate": 3.204444444444444e-07,
|
| 826 |
+
"logits/chosen": -1.328125,
|
| 827 |
+
"logits/rejected": -1.7578125,
|
| 828 |
+
"logps/chosen": -278.0,
|
| 829 |
+
"logps/rejected": -358.0,
|
| 830 |
+
"loss": 0.0213,
|
| 831 |
+
"rewards/accuracies": 1.0,
|
| 832 |
+
"rewards/chosen": 0.953125,
|
| 833 |
+
"rewards/margins": 9.625,
|
| 834 |
+
"rewards/rejected": -8.6875,
|
| 835 |
+
"step": 530
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.2949640287769784,
|
| 839 |
+
"grad_norm": 4.82912252555677,
|
| 840 |
+
"learning_rate": 3.1599999999999997e-07,
|
| 841 |
+
"logits/chosen": -1.2890625,
|
| 842 |
+
"logits/rejected": -1.734375,
|
| 843 |
+
"logps/chosen": -251.0,
|
| 844 |
+
"logps/rejected": -368.0,
|
| 845 |
+
"loss": 0.0034,
|
| 846 |
+
"rewards/accuracies": 1.0,
|
| 847 |
+
"rewards/chosen": 1.21875,
|
| 848 |
+
"rewards/margins": 10.5,
|
| 849 |
+
"rewards/rejected": -9.3125,
|
| 850 |
+
"step": 540
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.3189448441247003,
|
| 854 |
+
"grad_norm": 0.21054005672290058,
|
| 855 |
+
"learning_rate": 3.115555555555555e-07,
|
| 856 |
+
"logits/chosen": -1.3515625,
|
| 857 |
+
"logits/rejected": -1.8671875,
|
| 858 |
+
"logps/chosen": -230.0,
|
| 859 |
+
"logps/rejected": -364.0,
|
| 860 |
+
"loss": 0.003,
|
| 861 |
+
"rewards/accuracies": 1.0,
|
| 862 |
+
"rewards/chosen": 1.6953125,
|
| 863 |
+
"rewards/margins": 10.5,
|
| 864 |
+
"rewards/rejected": -8.8125,
|
| 865 |
+
"step": 550
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.3429256594724222,
|
| 869 |
+
"grad_norm": 0.3308230367772809,
|
| 870 |
+
"learning_rate": 3.071111111111111e-07,
|
| 871 |
+
"logits/chosen": -1.265625,
|
| 872 |
+
"logits/rejected": -1.7421875,
|
| 873 |
+
"logps/chosen": -244.0,
|
| 874 |
+
"logps/rejected": -368.0,
|
| 875 |
+
"loss": 0.0066,
|
| 876 |
+
"rewards/accuracies": 0.987500011920929,
|
| 877 |
+
"rewards/chosen": 1.421875,
|
| 878 |
+
"rewards/margins": 10.5625,
|
| 879 |
+
"rewards/rejected": -9.1875,
|
| 880 |
+
"step": 560
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 1.3669064748201438,
|
| 884 |
+
"grad_norm": 2.5398850541641953,
|
| 885 |
+
"learning_rate": 3.026666666666666e-07,
|
| 886 |
+
"logits/chosen": -1.25,
|
| 887 |
+
"logits/rejected": -1.796875,
|
| 888 |
+
"logps/chosen": -234.0,
|
| 889 |
+
"logps/rejected": -366.0,
|
| 890 |
+
"loss": 0.005,
|
| 891 |
+
"rewards/accuracies": 1.0,
|
| 892 |
+
"rewards/chosen": 1.296875,
|
| 893 |
+
"rewards/margins": 10.75,
|
| 894 |
+
"rewards/rejected": -9.5,
|
| 895 |
+
"step": 570
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 1.3908872901678657,
|
| 899 |
+
"grad_norm": 0.03226588556395318,
|
| 900 |
+
"learning_rate": 2.982222222222222e-07,
|
| 901 |
+
"logits/chosen": -1.2734375,
|
| 902 |
+
"logits/rejected": -1.71875,
|
| 903 |
+
"logps/chosen": -258.0,
|
| 904 |
+
"logps/rejected": -370.0,
|
| 905 |
+
"loss": 0.0039,
|
| 906 |
+
"rewards/accuracies": 1.0,
|
| 907 |
+
"rewards/chosen": 1.5546875,
|
| 908 |
+
"rewards/margins": 11.5625,
|
| 909 |
+
"rewards/rejected": -10.0,
|
| 910 |
+
"step": 580
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 1.4148681055155876,
|
| 914 |
+
"grad_norm": 3.5465151461901057,
|
| 915 |
+
"learning_rate": 2.937777777777778e-07,
|
| 916 |
+
"logits/chosen": -1.3203125,
|
| 917 |
+
"logits/rejected": -1.65625,
|
| 918 |
+
"logps/chosen": -238.0,
|
| 919 |
+
"logps/rejected": -366.0,
|
| 920 |
+
"loss": 0.0054,
|
| 921 |
+
"rewards/accuracies": 1.0,
|
| 922 |
+
"rewards/chosen": 1.421875,
|
| 923 |
+
"rewards/margins": 9.875,
|
| 924 |
+
"rewards/rejected": -8.4375,
|
| 925 |
+
"step": 590
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 1.4388489208633093,
|
| 929 |
+
"grad_norm": 0.9672041242288474,
|
| 930 |
+
"learning_rate": 2.8933333333333333e-07,
|
| 931 |
+
"logits/chosen": -1.2890625,
|
| 932 |
+
"logits/rejected": -1.765625,
|
| 933 |
+
"logps/chosen": -256.0,
|
| 934 |
+
"logps/rejected": -378.0,
|
| 935 |
+
"loss": 0.0091,
|
| 936 |
+
"rewards/accuracies": 1.0,
|
| 937 |
+
"rewards/chosen": 1.2421875,
|
| 938 |
+
"rewards/margins": 11.0625,
|
| 939 |
+
"rewards/rejected": -9.8125,
|
| 940 |
+
"step": 600
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.4628297362110312,
|
| 944 |
+
"grad_norm": 0.23832756129503824,
|
| 945 |
+
"learning_rate": 2.848888888888889e-07,
|
| 946 |
+
"logits/chosen": -1.296875,
|
| 947 |
+
"logits/rejected": -1.6875,
|
| 948 |
+
"logps/chosen": -241.0,
|
| 949 |
+
"logps/rejected": -376.0,
|
| 950 |
+
"loss": 0.0177,
|
| 951 |
+
"rewards/accuracies": 1.0,
|
| 952 |
+
"rewards/chosen": 1.015625,
|
| 953 |
+
"rewards/margins": 10.9375,
|
| 954 |
+
"rewards/rejected": -9.9375,
|
| 955 |
+
"step": 610
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.486810551558753,
|
| 959 |
+
"grad_norm": 14.900430603440345,
|
| 960 |
+
"learning_rate": 2.8044444444444445e-07,
|
| 961 |
+
"logits/chosen": -1.3515625,
|
| 962 |
+
"logits/rejected": -1.765625,
|
| 963 |
+
"logps/chosen": -274.0,
|
| 964 |
+
"logps/rejected": -372.0,
|
| 965 |
+
"loss": 0.0184,
|
| 966 |
+
"rewards/accuracies": 0.987500011920929,
|
| 967 |
+
"rewards/chosen": 1.46875,
|
| 968 |
+
"rewards/margins": 11.6875,
|
| 969 |
+
"rewards/rejected": -10.25,
|
| 970 |
+
"step": 620
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.5107913669064748,
|
| 974 |
+
"grad_norm": 29.0158558612294,
|
| 975 |
+
"learning_rate": 2.7600000000000004e-07,
|
| 976 |
+
"logits/chosen": -1.296875,
|
| 977 |
+
"logits/rejected": -1.875,
|
| 978 |
+
"logps/chosen": -264.0,
|
| 979 |
+
"logps/rejected": -392.0,
|
| 980 |
+
"loss": 0.0258,
|
| 981 |
+
"rewards/accuracies": 1.0,
|
| 982 |
+
"rewards/chosen": 0.8515625,
|
| 983 |
+
"rewards/margins": 11.0625,
|
| 984 |
+
"rewards/rejected": -10.1875,
|
| 985 |
+
"step": 630
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.5347721822541966,
|
| 989 |
+
"grad_norm": 0.10758431473862176,
|
| 990 |
+
"learning_rate": 2.715555555555555e-07,
|
| 991 |
+
"logits/chosen": -1.2734375,
|
| 992 |
+
"logits/rejected": -1.8125,
|
| 993 |
+
"logps/chosen": -258.0,
|
| 994 |
+
"logps/rejected": -380.0,
|
| 995 |
+
"loss": 0.0205,
|
| 996 |
+
"rewards/accuracies": 0.987500011920929,
|
| 997 |
+
"rewards/chosen": 1.390625,
|
| 998 |
+
"rewards/margins": 12.6875,
|
| 999 |
+
"rewards/rejected": -11.3125,
|
| 1000 |
+
"step": 640
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.5587529976019185,
|
| 1004 |
+
"grad_norm": 0.6643883173223759,
|
| 1005 |
+
"learning_rate": 2.671111111111111e-07,
|
| 1006 |
+
"logits/chosen": -1.3359375,
|
| 1007 |
+
"logits/rejected": -1.734375,
|
| 1008 |
+
"logps/chosen": -247.0,
|
| 1009 |
+
"logps/rejected": -388.0,
|
| 1010 |
+
"loss": 0.001,
|
| 1011 |
+
"rewards/accuracies": 1.0,
|
| 1012 |
+
"rewards/chosen": 1.34375,
|
| 1013 |
+
"rewards/margins": 11.3125,
|
| 1014 |
+
"rewards/rejected": -10.0,
|
| 1015 |
+
"step": 650
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.5827338129496402,
|
| 1019 |
+
"grad_norm": 0.19081443363615663,
|
| 1020 |
+
"learning_rate": 2.6266666666666664e-07,
|
| 1021 |
+
"logits/chosen": -1.359375,
|
| 1022 |
+
"logits/rejected": -1.7109375,
|
| 1023 |
+
"logps/chosen": -268.0,
|
| 1024 |
+
"logps/rejected": -376.0,
|
| 1025 |
+
"loss": 0.0085,
|
| 1026 |
+
"rewards/accuracies": 1.0,
|
| 1027 |
+
"rewards/chosen": 1.1796875,
|
| 1028 |
+
"rewards/margins": 10.8125,
|
| 1029 |
+
"rewards/rejected": -9.625,
|
| 1030 |
+
"step": 660
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.6067146282973621,
|
| 1034 |
+
"grad_norm": 0.6802954971662545,
|
| 1035 |
+
"learning_rate": 2.582222222222222e-07,
|
| 1036 |
+
"logits/chosen": -1.2734375,
|
| 1037 |
+
"logits/rejected": -1.71875,
|
| 1038 |
+
"logps/chosen": -241.0,
|
| 1039 |
+
"logps/rejected": -382.0,
|
| 1040 |
+
"loss": 0.004,
|
| 1041 |
+
"rewards/accuracies": 1.0,
|
| 1042 |
+
"rewards/chosen": 1.609375,
|
| 1043 |
+
"rewards/margins": 10.6875,
|
| 1044 |
+
"rewards/rejected": -9.125,
|
| 1045 |
+
"step": 670
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.630695443645084,
|
| 1049 |
+
"grad_norm": 0.21946623844305102,
|
| 1050 |
+
"learning_rate": 2.5377777777777776e-07,
|
| 1051 |
+
"logits/chosen": -1.3125,
|
| 1052 |
+
"logits/rejected": -1.734375,
|
| 1053 |
+
"logps/chosen": -246.0,
|
| 1054 |
+
"logps/rejected": -378.0,
|
| 1055 |
+
"loss": 0.0077,
|
| 1056 |
+
"rewards/accuracies": 1.0,
|
| 1057 |
+
"rewards/chosen": 1.34375,
|
| 1058 |
+
"rewards/margins": 12.1875,
|
| 1059 |
+
"rewards/rejected": -10.875,
|
| 1060 |
+
"step": 680
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.6546762589928057,
|
| 1064 |
+
"grad_norm": 14.387030473360179,
|
| 1065 |
+
"learning_rate": 2.493333333333333e-07,
|
| 1066 |
+
"logits/chosen": -1.2890625,
|
| 1067 |
+
"logits/rejected": -1.7265625,
|
| 1068 |
+
"logps/chosen": -252.0,
|
| 1069 |
+
"logps/rejected": -382.0,
|
| 1070 |
+
"loss": 0.0055,
|
| 1071 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1072 |
+
"rewards/chosen": 0.8046875,
|
| 1073 |
+
"rewards/margins": 11.0,
|
| 1074 |
+
"rewards/rejected": -10.1875,
|
| 1075 |
+
"step": 690
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.6786570743405276,
|
| 1079 |
+
"grad_norm": 4.254572167354162,
|
| 1080 |
+
"learning_rate": 2.448888888888889e-07,
|
| 1081 |
+
"logits/chosen": -1.265625,
|
| 1082 |
+
"logits/rejected": -1.796875,
|
| 1083 |
+
"logps/chosen": -248.0,
|
| 1084 |
+
"logps/rejected": -384.0,
|
| 1085 |
+
"loss": 0.0027,
|
| 1086 |
+
"rewards/accuracies": 1.0,
|
| 1087 |
+
"rewards/chosen": 1.578125,
|
| 1088 |
+
"rewards/margins": 11.5,
|
| 1089 |
+
"rewards/rejected": -9.875,
|
| 1090 |
+
"step": 700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.7026378896882495,
|
| 1094 |
+
"grad_norm": 2.551492265163294,
|
| 1095 |
+
"learning_rate": 2.404444444444444e-07,
|
| 1096 |
+
"logits/chosen": -1.3125,
|
| 1097 |
+
"logits/rejected": -1.734375,
|
| 1098 |
+
"logps/chosen": -241.0,
|
| 1099 |
+
"logps/rejected": -384.0,
|
| 1100 |
+
"loss": 0.0037,
|
| 1101 |
+
"rewards/accuracies": 1.0,
|
| 1102 |
+
"rewards/chosen": 1.546875,
|
| 1103 |
+
"rewards/margins": 12.5,
|
| 1104 |
+
"rewards/rejected": -10.9375,
|
| 1105 |
+
"step": 710
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.7266187050359711,
|
| 1109 |
+
"grad_norm": 3.5788184390591766,
|
| 1110 |
+
"learning_rate": 2.3599999999999997e-07,
|
| 1111 |
+
"logits/chosen": -1.296875,
|
| 1112 |
+
"logits/rejected": -1.6875,
|
| 1113 |
+
"logps/chosen": -270.0,
|
| 1114 |
+
"logps/rejected": -370.0,
|
| 1115 |
+
"loss": 0.002,
|
| 1116 |
+
"rewards/accuracies": 1.0,
|
| 1117 |
+
"rewards/chosen": 0.94921875,
|
| 1118 |
+
"rewards/margins": 11.9375,
|
| 1119 |
+
"rewards/rejected": -11.0,
|
| 1120 |
+
"step": 720
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.750599520383693,
|
| 1124 |
+
"grad_norm": 0.22812646760052926,
|
| 1125 |
+
"learning_rate": 2.3155555555555553e-07,
|
| 1126 |
+
"logits/chosen": -1.3125,
|
| 1127 |
+
"logits/rejected": -1.7265625,
|
| 1128 |
+
"logps/chosen": -252.0,
|
| 1129 |
+
"logps/rejected": -376.0,
|
| 1130 |
+
"loss": 0.003,
|
| 1131 |
+
"rewards/accuracies": 1.0,
|
| 1132 |
+
"rewards/chosen": 1.421875,
|
| 1133 |
+
"rewards/margins": 11.1875,
|
| 1134 |
+
"rewards/rejected": -9.8125,
|
| 1135 |
+
"step": 730
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 1.774580335731415,
|
| 1139 |
+
"grad_norm": 3.5369041443316966,
|
| 1140 |
+
"learning_rate": 2.2711111111111112e-07,
|
| 1141 |
+
"logits/chosen": -1.28125,
|
| 1142 |
+
"logits/rejected": -1.765625,
|
| 1143 |
+
"logps/chosen": -247.0,
|
| 1144 |
+
"logps/rejected": -374.0,
|
| 1145 |
+
"loss": 0.0019,
|
| 1146 |
+
"rewards/accuracies": 1.0,
|
| 1147 |
+
"rewards/chosen": 1.453125,
|
| 1148 |
+
"rewards/margins": 10.9375,
|
| 1149 |
+
"rewards/rejected": -9.4375,
|
| 1150 |
+
"step": 740
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.7985611510791366,
|
| 1154 |
+
"grad_norm": 0.15385299522364304,
|
| 1155 |
+
"learning_rate": 2.2266666666666668e-07,
|
| 1156 |
+
"logits/chosen": -1.296875,
|
| 1157 |
+
"logits/rejected": -1.6640625,
|
| 1158 |
+
"logps/chosen": -247.0,
|
| 1159 |
+
"logps/rejected": -372.0,
|
| 1160 |
+
"loss": 0.0083,
|
| 1161 |
+
"rewards/accuracies": 1.0,
|
| 1162 |
+
"rewards/chosen": 1.109375,
|
| 1163 |
+
"rewards/margins": 11.25,
|
| 1164 |
+
"rewards/rejected": -10.125,
|
| 1165 |
+
"step": 750
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.8225419664268585,
|
| 1169 |
+
"grad_norm": 1.8182203002712007,
|
| 1170 |
+
"learning_rate": 2.1822222222222224e-07,
|
| 1171 |
+
"logits/chosen": -1.34375,
|
| 1172 |
+
"logits/rejected": -1.7421875,
|
| 1173 |
+
"logps/chosen": -260.0,
|
| 1174 |
+
"logps/rejected": -378.0,
|
| 1175 |
+
"loss": 0.001,
|
| 1176 |
+
"rewards/accuracies": 1.0,
|
| 1177 |
+
"rewards/chosen": 1.03125,
|
| 1178 |
+
"rewards/margins": 11.5625,
|
| 1179 |
+
"rewards/rejected": -10.5,
|
| 1180 |
+
"step": 760
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.8465227817745804,
|
| 1184 |
+
"grad_norm": 0.12014261187968397,
|
| 1185 |
+
"learning_rate": 2.1377777777777777e-07,
|
| 1186 |
+
"logits/chosen": -1.234375,
|
| 1187 |
+
"logits/rejected": -1.71875,
|
| 1188 |
+
"logps/chosen": -237.0,
|
| 1189 |
+
"logps/rejected": -366.0,
|
| 1190 |
+
"loss": 0.0005,
|
| 1191 |
+
"rewards/accuracies": 1.0,
|
| 1192 |
+
"rewards/chosen": 1.4609375,
|
| 1193 |
+
"rewards/margins": 11.75,
|
| 1194 |
+
"rewards/rejected": -10.25,
|
| 1195 |
+
"step": 770
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.870503597122302,
|
| 1199 |
+
"grad_norm": 0.017562287950709583,
|
| 1200 |
+
"learning_rate": 2.0933333333333333e-07,
|
| 1201 |
+
"logits/chosen": -1.328125,
|
| 1202 |
+
"logits/rejected": -1.7109375,
|
| 1203 |
+
"logps/chosen": -256.0,
|
| 1204 |
+
"logps/rejected": -378.0,
|
| 1205 |
+
"loss": 0.0054,
|
| 1206 |
+
"rewards/accuracies": 1.0,
|
| 1207 |
+
"rewards/chosen": 1.4765625,
|
| 1208 |
+
"rewards/margins": 11.6875,
|
| 1209 |
+
"rewards/rejected": -10.1875,
|
| 1210 |
+
"step": 780
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.894484412470024,
|
| 1214 |
+
"grad_norm": 0.15749847503299128,
|
| 1215 |
+
"learning_rate": 2.048888888888889e-07,
|
| 1216 |
+
"logits/chosen": -1.34375,
|
| 1217 |
+
"logits/rejected": -1.65625,
|
| 1218 |
+
"logps/chosen": -242.0,
|
| 1219 |
+
"logps/rejected": -378.0,
|
| 1220 |
+
"loss": 0.023,
|
| 1221 |
+
"rewards/accuracies": 1.0,
|
| 1222 |
+
"rewards/chosen": 1.59375,
|
| 1223 |
+
"rewards/margins": 11.0,
|
| 1224 |
+
"rewards/rejected": -9.375,
|
| 1225 |
+
"step": 790
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 1.9184652278177459,
|
| 1229 |
+
"grad_norm": 0.08616776985304095,
|
| 1230 |
+
"learning_rate": 2.0044444444444445e-07,
|
| 1231 |
+
"logits/chosen": -1.3046875,
|
| 1232 |
+
"logits/rejected": -1.6875,
|
| 1233 |
+
"logps/chosen": -235.0,
|
| 1234 |
+
"logps/rejected": -388.0,
|
| 1235 |
+
"loss": 0.0009,
|
| 1236 |
+
"rewards/accuracies": 1.0,
|
| 1237 |
+
"rewards/chosen": 0.68359375,
|
| 1238 |
+
"rewards/margins": 11.5625,
|
| 1239 |
+
"rewards/rejected": -10.875,
|
| 1240 |
+
"step": 800
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 1.9424460431654675,
|
| 1244 |
+
"grad_norm": 0.11034132303896381,
|
| 1245 |
+
"learning_rate": 1.96e-07,
|
| 1246 |
+
"logits/chosen": -1.3515625,
|
| 1247 |
+
"logits/rejected": -1.7109375,
|
| 1248 |
+
"logps/chosen": -258.0,
|
| 1249 |
+
"logps/rejected": -374.0,
|
| 1250 |
+
"loss": 0.0052,
|
| 1251 |
+
"rewards/accuracies": 1.0,
|
| 1252 |
+
"rewards/chosen": 1.375,
|
| 1253 |
+
"rewards/margins": 11.9375,
|
| 1254 |
+
"rewards/rejected": -10.625,
|
| 1255 |
+
"step": 810
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.9664268585131894,
|
| 1259 |
+
"grad_norm": 0.1125888676469132,
|
| 1260 |
+
"learning_rate": 1.9155555555555554e-07,
|
| 1261 |
+
"logits/chosen": -1.203125,
|
| 1262 |
+
"logits/rejected": -1.7109375,
|
| 1263 |
+
"logps/chosen": -264.0,
|
| 1264 |
+
"logps/rejected": -380.0,
|
| 1265 |
+
"loss": 0.006,
|
| 1266 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1267 |
+
"rewards/chosen": 1.4921875,
|
| 1268 |
+
"rewards/margins": 12.1875,
|
| 1269 |
+
"rewards/rejected": -10.75,
|
| 1270 |
+
"step": 820
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.9904076738609113,
|
| 1274 |
+
"grad_norm": 0.15670668500960064,
|
| 1275 |
+
"learning_rate": 1.871111111111111e-07,
|
| 1276 |
+
"logits/chosen": -1.296875,
|
| 1277 |
+
"logits/rejected": -1.6953125,
|
| 1278 |
+
"logps/chosen": -252.0,
|
| 1279 |
+
"logps/rejected": -392.0,
|
| 1280 |
+
"loss": 0.0101,
|
| 1281 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1282 |
+
"rewards/chosen": 1.1640625,
|
| 1283 |
+
"rewards/margins": 11.0625,
|
| 1284 |
+
"rewards/rejected": -9.875,
|
| 1285 |
+
"step": 830
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 2.0,
|
| 1289 |
+
"eval_logits/chosen": -1.2109375,
|
| 1290 |
+
"eval_logits/rejected": -1.6796875,
|
| 1291 |
+
"eval_logps/chosen": -227.0,
|
| 1292 |
+
"eval_logps/rejected": -376.0,
|
| 1293 |
+
"eval_loss": 0.11118045449256897,
|
| 1294 |
+
"eval_rewards/accuracies": 0.9583333134651184,
|
| 1295 |
+
"eval_rewards/chosen": 1.1171875,
|
| 1296 |
+
"eval_rewards/margins": 10.125,
|
| 1297 |
+
"eval_rewards/rejected": -9.0,
|
| 1298 |
+
"eval_runtime": 11.778,
|
| 1299 |
+
"eval_samples_per_second": 16.981,
|
| 1300 |
+
"eval_steps_per_second": 0.764,
|
| 1301 |
+
"step": 834
|
| 1302 |
}
|
| 1303 |
],
|
| 1304 |
"logging_steps": 10,
|