Commit
•
d5a5b05
1
Parent(s):
0b9c80a
Training in progress, epoch 3, checkpoint
Browse files- last-checkpoint/global_step939/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +1 -1
- last-checkpoint/global_step939/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +1 -1
- last-checkpoint/global_step939/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +1 -1
- last-checkpoint/global_step939/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +1 -1
- 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/rng_state_3.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +484 -3
last-checkpoint/global_step939/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16060527788
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05913d20787049efa6f824d72634683bd1f7c033d1cecedd4d145e7004455eb9
|
| 3 |
size 16060527788
|
last-checkpoint/global_step939/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16060527788
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aefda786d9f77daf98567fa45de15ba99eea0ce8264d2ac64ddb64e84142f95c
|
| 3 |
size 16060527788
|
last-checkpoint/global_step939/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16060527788
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:321f40b28d9d745c38f4a28c5d2a0f4e22cda431ff78463d45c843efafabd3ea
|
| 3 |
size 16060527788
|
last-checkpoint/global_step939/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16060527788
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:173cbd92e7431ae3bcd5833d4031309b911b5ac8a690cb0807de6316d6ffe747
|
| 3 |
size 16060527788
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step939
|
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 4976698672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e328c4b9e408094b769bcfdf47a15208d5acb0eb5819e898d72b4d7196e8ba4c
|
| 3 |
size 4976698672
|
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 4999802720
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09d5c223ef59d11fe1dfa4cb6735afd256a9a354d3f5951af9b2f47767185aee
|
| 3 |
size 4999802720
|
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 4915916176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee090c536140ebe9822778c771b86d853685174c66af36ffb76a6b8a796df814
|
| 3 |
size 4915916176
|
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 1168138808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c6c3237f3c1804ed5da951f36e10eb384ff1b824ca5e50a8b56a30ca3290e67
|
| 3 |
size 1168138808
|
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 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab7cf0d34d60dfda516cf9661904550e2e294e723edd07c25c738f05e8ba92d1
|
| 3 |
size 15024
|
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 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7b183324e8227a51a9556d86b2ad893a8c4c52205ed4a737356c6611dac7353
|
| 3 |
size 15024
|
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 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac69e994090f4818cb1fa6f6cefa363178552c3c731c6507ff195bcb07fd5bef
|
| 3 |
size 15024
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15024
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f68e61b63402f8afb1f69c960f7944965655dac11e3ccf29919c282f23931f86
|
| 3 |
size 15024
|
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:fe7d17fbcfa5bee1bf6cb5aedab3ff5a70436912c200d7301d173be443809d63
|
| 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,
|
|
@@ -984,6 +984,487 @@
|
|
| 984 |
"eval_samples_per_second": 15.854,
|
| 985 |
"eval_steps_per_second": 0.555,
|
| 986 |
"step": 626
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 987 |
}
|
| 988 |
],
|
| 989 |
"logging_steps": 10,
|
|
@@ -998,7 +1479,7 @@
|
|
| 998 |
"should_evaluate": false,
|
| 999 |
"should_log": false,
|
| 1000 |
"should_save": true,
|
| 1001 |
-
"should_training_stop":
|
| 1002 |
},
|
| 1003 |
"attributes": {}
|
| 1004 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 939,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 984 |
"eval_samples_per_second": 15.854,
|
| 985 |
"eval_steps_per_second": 0.555,
|
| 986 |
"step": 626
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.012779552715655,
|
| 990 |
+
"grad_norm": 15.229138030548278,
|
| 991 |
+
"learning_rate": 1.8284023668639053e-07,
|
| 992 |
+
"logits/chosen": -1.609375,
|
| 993 |
+
"logits/rejected": -1.6171875,
|
| 994 |
+
"logps/chosen": -240.0,
|
| 995 |
+
"logps/rejected": -268.0,
|
| 996 |
+
"loss": 0.2259,
|
| 997 |
+
"rewards/accuracies": 0.949999988079071,
|
| 998 |
+
"rewards/chosen": -0.201171875,
|
| 999 |
+
"rewards/margins": 1.8046875,
|
| 1000 |
+
"rewards/rejected": -2.0,
|
| 1001 |
+
"step": 630
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 2.0447284345047922,
|
| 1005 |
+
"grad_norm": 17.70681792433403,
|
| 1006 |
+
"learning_rate": 1.7692307692307693e-07,
|
| 1007 |
+
"logits/chosen": -1.6171875,
|
| 1008 |
+
"logits/rejected": -1.609375,
|
| 1009 |
+
"logps/chosen": -237.0,
|
| 1010 |
+
"logps/rejected": -264.0,
|
| 1011 |
+
"loss": 0.2073,
|
| 1012 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1013 |
+
"rewards/chosen": 0.208984375,
|
| 1014 |
+
"rewards/margins": 2.203125,
|
| 1015 |
+
"rewards/rejected": -2.0,
|
| 1016 |
+
"step": 640
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 2.07667731629393,
|
| 1020 |
+
"grad_norm": 13.769490197345323,
|
| 1021 |
+
"learning_rate": 1.710059171597633e-07,
|
| 1022 |
+
"logits/chosen": -1.6484375,
|
| 1023 |
+
"logits/rejected": -1.65625,
|
| 1024 |
+
"logps/chosen": -242.0,
|
| 1025 |
+
"logps/rejected": -270.0,
|
| 1026 |
+
"loss": 0.2069,
|
| 1027 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1028 |
+
"rewards/chosen": -0.0859375,
|
| 1029 |
+
"rewards/margins": 2.015625,
|
| 1030 |
+
"rewards/rejected": -2.109375,
|
| 1031 |
+
"step": 650
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 2.108626198083067,
|
| 1035 |
+
"grad_norm": 10.553376765871462,
|
| 1036 |
+
"learning_rate": 1.650887573964497e-07,
|
| 1037 |
+
"logits/chosen": -1.640625,
|
| 1038 |
+
"logits/rejected": -1.65625,
|
| 1039 |
+
"logps/chosen": -238.0,
|
| 1040 |
+
"logps/rejected": -266.0,
|
| 1041 |
+
"loss": 0.188,
|
| 1042 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1043 |
+
"rewards/chosen": -0.06640625,
|
| 1044 |
+
"rewards/margins": 2.21875,
|
| 1045 |
+
"rewards/rejected": -2.28125,
|
| 1046 |
+
"step": 660
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.1405750798722045,
|
| 1050 |
+
"grad_norm": 12.717503262116809,
|
| 1051 |
+
"learning_rate": 1.591715976331361e-07,
|
| 1052 |
+
"logits/chosen": -1.6640625,
|
| 1053 |
+
"logits/rejected": -1.625,
|
| 1054 |
+
"logps/chosen": -240.0,
|
| 1055 |
+
"logps/rejected": -258.0,
|
| 1056 |
+
"loss": 0.2087,
|
| 1057 |
+
"rewards/accuracies": 0.9375,
|
| 1058 |
+
"rewards/chosen": -0.05419921875,
|
| 1059 |
+
"rewards/margins": 1.828125,
|
| 1060 |
+
"rewards/rejected": -1.8828125,
|
| 1061 |
+
"step": 670
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 2.1725239616613417,
|
| 1065 |
+
"grad_norm": 17.368724259557464,
|
| 1066 |
+
"learning_rate": 1.5325443786982248e-07,
|
| 1067 |
+
"logits/chosen": -1.6484375,
|
| 1068 |
+
"logits/rejected": -1.6875,
|
| 1069 |
+
"logps/chosen": -244.0,
|
| 1070 |
+
"logps/rejected": -264.0,
|
| 1071 |
+
"loss": 0.196,
|
| 1072 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1073 |
+
"rewards/chosen": -0.09423828125,
|
| 1074 |
+
"rewards/margins": 2.28125,
|
| 1075 |
+
"rewards/rejected": -2.375,
|
| 1076 |
+
"step": 680
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.2044728434504792,
|
| 1080 |
+
"grad_norm": 15.997180398783524,
|
| 1081 |
+
"learning_rate": 1.4733727810650885e-07,
|
| 1082 |
+
"logits/chosen": -1.625,
|
| 1083 |
+
"logits/rejected": -1.640625,
|
| 1084 |
+
"logps/chosen": -239.0,
|
| 1085 |
+
"logps/rejected": -262.0,
|
| 1086 |
+
"loss": 0.1955,
|
| 1087 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1088 |
+
"rewards/chosen": -0.07763671875,
|
| 1089 |
+
"rewards/margins": 1.890625,
|
| 1090 |
+
"rewards/rejected": -1.96875,
|
| 1091 |
+
"step": 690
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.236421725239617,
|
| 1095 |
+
"grad_norm": 16.465231501289626,
|
| 1096 |
+
"learning_rate": 1.4142011834319526e-07,
|
| 1097 |
+
"logits/chosen": -1.6171875,
|
| 1098 |
+
"logits/rejected": -1.6796875,
|
| 1099 |
+
"logps/chosen": -247.0,
|
| 1100 |
+
"logps/rejected": -264.0,
|
| 1101 |
+
"loss": 0.1818,
|
| 1102 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1103 |
+
"rewards/chosen": 0.0087890625,
|
| 1104 |
+
"rewards/margins": 2.03125,
|
| 1105 |
+
"rewards/rejected": -2.03125,
|
| 1106 |
+
"step": 700
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 2.268370607028754,
|
| 1110 |
+
"grad_norm": 15.303382006736875,
|
| 1111 |
+
"learning_rate": 1.3550295857988164e-07,
|
| 1112 |
+
"logits/chosen": -1.640625,
|
| 1113 |
+
"logits/rejected": -1.671875,
|
| 1114 |
+
"logps/chosen": -240.0,
|
| 1115 |
+
"logps/rejected": -266.0,
|
| 1116 |
+
"loss": 0.196,
|
| 1117 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1118 |
+
"rewards/chosen": -0.02001953125,
|
| 1119 |
+
"rewards/margins": 1.9375,
|
| 1120 |
+
"rewards/rejected": -1.9609375,
|
| 1121 |
+
"step": 710
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 2.3003194888178915,
|
| 1125 |
+
"grad_norm": 27.787991315567947,
|
| 1126 |
+
"learning_rate": 1.2958579881656802e-07,
|
| 1127 |
+
"logits/chosen": -1.6328125,
|
| 1128 |
+
"logits/rejected": -1.609375,
|
| 1129 |
+
"logps/chosen": -247.0,
|
| 1130 |
+
"logps/rejected": -268.0,
|
| 1131 |
+
"loss": 0.2045,
|
| 1132 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1133 |
+
"rewards/chosen": -0.34765625,
|
| 1134 |
+
"rewards/margins": 1.9375,
|
| 1135 |
+
"rewards/rejected": -2.28125,
|
| 1136 |
+
"step": 720
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 2.3322683706070286,
|
| 1140 |
+
"grad_norm": 20.488749808987933,
|
| 1141 |
+
"learning_rate": 1.2366863905325443e-07,
|
| 1142 |
+
"logits/chosen": -1.609375,
|
| 1143 |
+
"logits/rejected": -1.6484375,
|
| 1144 |
+
"logps/chosen": -236.0,
|
| 1145 |
+
"logps/rejected": -262.0,
|
| 1146 |
+
"loss": 0.1955,
|
| 1147 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1148 |
+
"rewards/chosen": 0.076171875,
|
| 1149 |
+
"rewards/margins": 2.15625,
|
| 1150 |
+
"rewards/rejected": -2.078125,
|
| 1151 |
+
"step": 730
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.364217252396166,
|
| 1155 |
+
"grad_norm": 16.34392558119839,
|
| 1156 |
+
"learning_rate": 1.1775147928994082e-07,
|
| 1157 |
+
"logits/chosen": -1.6328125,
|
| 1158 |
+
"logits/rejected": -1.6328125,
|
| 1159 |
+
"logps/chosen": -235.0,
|
| 1160 |
+
"logps/rejected": -260.0,
|
| 1161 |
+
"loss": 0.2017,
|
| 1162 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1163 |
+
"rewards/chosen": 0.171875,
|
| 1164 |
+
"rewards/margins": 2.171875,
|
| 1165 |
+
"rewards/rejected": -2.0,
|
| 1166 |
+
"step": 740
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 2.3961661341853033,
|
| 1170 |
+
"grad_norm": 22.986892420642484,
|
| 1171 |
+
"learning_rate": 1.1183431952662721e-07,
|
| 1172 |
+
"logits/chosen": -1.59375,
|
| 1173 |
+
"logits/rejected": -1.640625,
|
| 1174 |
+
"logps/chosen": -238.0,
|
| 1175 |
+
"logps/rejected": -262.0,
|
| 1176 |
+
"loss": 0.1846,
|
| 1177 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1178 |
+
"rewards/chosen": 0.033935546875,
|
| 1179 |
+
"rewards/margins": 2.03125,
|
| 1180 |
+
"rewards/rejected": -1.9921875,
|
| 1181 |
+
"step": 750
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 2.428115015974441,
|
| 1185 |
+
"grad_norm": 22.954617926133928,
|
| 1186 |
+
"learning_rate": 1.059171597633136e-07,
|
| 1187 |
+
"logits/chosen": -1.6171875,
|
| 1188 |
+
"logits/rejected": -1.640625,
|
| 1189 |
+
"logps/chosen": -238.0,
|
| 1190 |
+
"logps/rejected": -264.0,
|
| 1191 |
+
"loss": 0.1675,
|
| 1192 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1193 |
+
"rewards/chosen": 0.1943359375,
|
| 1194 |
+
"rewards/margins": 2.3125,
|
| 1195 |
+
"rewards/rejected": -2.125,
|
| 1196 |
+
"step": 760
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.460063897763578,
|
| 1200 |
+
"grad_norm": 10.72220228289897,
|
| 1201 |
+
"learning_rate": 1e-07,
|
| 1202 |
+
"logits/chosen": -1.59375,
|
| 1203 |
+
"logits/rejected": -1.6015625,
|
| 1204 |
+
"logps/chosen": -239.0,
|
| 1205 |
+
"logps/rejected": -268.0,
|
| 1206 |
+
"loss": 0.1909,
|
| 1207 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1208 |
+
"rewards/chosen": -0.083984375,
|
| 1209 |
+
"rewards/margins": 2.171875,
|
| 1210 |
+
"rewards/rejected": -2.25,
|
| 1211 |
+
"step": 770
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 2.4920127795527156,
|
| 1215 |
+
"grad_norm": 22.698230184573,
|
| 1216 |
+
"learning_rate": 9.408284023668639e-08,
|
| 1217 |
+
"logits/chosen": -1.6328125,
|
| 1218 |
+
"logits/rejected": -1.6328125,
|
| 1219 |
+
"logps/chosen": -247.0,
|
| 1220 |
+
"logps/rejected": -262.0,
|
| 1221 |
+
"loss": 0.1904,
|
| 1222 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1223 |
+
"rewards/chosen": 0.044921875,
|
| 1224 |
+
"rewards/margins": 2.03125,
|
| 1225 |
+
"rewards/rejected": -1.9921875,
|
| 1226 |
+
"step": 780
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 2.523961661341853,
|
| 1230 |
+
"grad_norm": 17.131418963453264,
|
| 1231 |
+
"learning_rate": 8.816568047337278e-08,
|
| 1232 |
+
"logits/chosen": -1.4921875,
|
| 1233 |
+
"logits/rejected": -1.65625,
|
| 1234 |
+
"logps/chosen": -249.0,
|
| 1235 |
+
"logps/rejected": -264.0,
|
| 1236 |
+
"loss": 0.2065,
|
| 1237 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1238 |
+
"rewards/chosen": -0.2119140625,
|
| 1239 |
+
"rewards/margins": 2.09375,
|
| 1240 |
+
"rewards/rejected": -2.296875,
|
| 1241 |
+
"step": 790
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.5559105431309903,
|
| 1245 |
+
"grad_norm": 14.204160115025998,
|
| 1246 |
+
"learning_rate": 8.224852071005916e-08,
|
| 1247 |
+
"logits/chosen": -1.640625,
|
| 1248 |
+
"logits/rejected": -1.6328125,
|
| 1249 |
+
"logps/chosen": -247.0,
|
| 1250 |
+
"logps/rejected": -264.0,
|
| 1251 |
+
"loss": 0.1824,
|
| 1252 |
+
"rewards/accuracies": 0.9375,
|
| 1253 |
+
"rewards/chosen": -0.130859375,
|
| 1254 |
+
"rewards/margins": 1.9296875,
|
| 1255 |
+
"rewards/rejected": -2.0625,
|
| 1256 |
+
"step": 800
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.587859424920128,
|
| 1260 |
+
"grad_norm": 16.438831527877248,
|
| 1261 |
+
"learning_rate": 7.633136094674555e-08,
|
| 1262 |
+
"logits/chosen": -1.6328125,
|
| 1263 |
+
"logits/rejected": -1.6640625,
|
| 1264 |
+
"logps/chosen": -241.0,
|
| 1265 |
+
"logps/rejected": -266.0,
|
| 1266 |
+
"loss": 0.1841,
|
| 1267 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1268 |
+
"rewards/chosen": -0.08154296875,
|
| 1269 |
+
"rewards/margins": 2.125,
|
| 1270 |
+
"rewards/rejected": -2.203125,
|
| 1271 |
+
"step": 810
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 2.619808306709265,
|
| 1275 |
+
"grad_norm": 23.382793800196303,
|
| 1276 |
+
"learning_rate": 7.041420118343195e-08,
|
| 1277 |
+
"logits/chosen": -1.609375,
|
| 1278 |
+
"logits/rejected": -1.609375,
|
| 1279 |
+
"logps/chosen": -240.0,
|
| 1280 |
+
"logps/rejected": -266.0,
|
| 1281 |
+
"loss": 0.1805,
|
| 1282 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1283 |
+
"rewards/chosen": -0.03271484375,
|
| 1284 |
+
"rewards/margins": 2.265625,
|
| 1285 |
+
"rewards/rejected": -2.296875,
|
| 1286 |
+
"step": 820
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 2.6517571884984026,
|
| 1290 |
+
"grad_norm": 17.97702724195434,
|
| 1291 |
+
"learning_rate": 6.449704142011835e-08,
|
| 1292 |
+
"logits/chosen": -1.6640625,
|
| 1293 |
+
"logits/rejected": -1.640625,
|
| 1294 |
+
"logps/chosen": -239.0,
|
| 1295 |
+
"logps/rejected": -258.0,
|
| 1296 |
+
"loss": 0.1621,
|
| 1297 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1298 |
+
"rewards/chosen": 0.004974365234375,
|
| 1299 |
+
"rewards/margins": 2.21875,
|
| 1300 |
+
"rewards/rejected": -2.21875,
|
| 1301 |
+
"step": 830
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.68370607028754,
|
| 1305 |
+
"grad_norm": 15.72139882933752,
|
| 1306 |
+
"learning_rate": 5.857988165680473e-08,
|
| 1307 |
+
"logits/chosen": -1.640625,
|
| 1308 |
+
"logits/rejected": -1.6484375,
|
| 1309 |
+
"logps/chosen": -244.0,
|
| 1310 |
+
"logps/rejected": -268.0,
|
| 1311 |
+
"loss": 0.1928,
|
| 1312 |
+
"rewards/accuracies": 0.9375,
|
| 1313 |
+
"rewards/chosen": -0.142578125,
|
| 1314 |
+
"rewards/margins": 2.296875,
|
| 1315 |
+
"rewards/rejected": -2.4375,
|
| 1316 |
+
"step": 840
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 2.7156549520766773,
|
| 1320 |
+
"grad_norm": 20.685915337750906,
|
| 1321 |
+
"learning_rate": 5.266272189349112e-08,
|
| 1322 |
+
"logits/chosen": -1.6171875,
|
| 1323 |
+
"logits/rejected": -1.6484375,
|
| 1324 |
+
"logps/chosen": -237.0,
|
| 1325 |
+
"logps/rejected": -268.0,
|
| 1326 |
+
"loss": 0.1831,
|
| 1327 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1328 |
+
"rewards/chosen": 0.1875,
|
| 1329 |
+
"rewards/margins": 2.390625,
|
| 1330 |
+
"rewards/rejected": -2.203125,
|
| 1331 |
+
"step": 850
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 2.747603833865815,
|
| 1335 |
+
"grad_norm": 20.038756432535152,
|
| 1336 |
+
"learning_rate": 4.674556213017751e-08,
|
| 1337 |
+
"logits/chosen": -1.609375,
|
| 1338 |
+
"logits/rejected": -1.6015625,
|
| 1339 |
+
"logps/chosen": -246.0,
|
| 1340 |
+
"logps/rejected": -268.0,
|
| 1341 |
+
"loss": 0.1985,
|
| 1342 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1343 |
+
"rewards/chosen": -0.298828125,
|
| 1344 |
+
"rewards/margins": 2.03125,
|
| 1345 |
+
"rewards/rejected": -2.328125,
|
| 1346 |
+
"step": 860
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.779552715654952,
|
| 1350 |
+
"grad_norm": 21.447728631568275,
|
| 1351 |
+
"learning_rate": 4.082840236686391e-08,
|
| 1352 |
+
"logits/chosen": -1.625,
|
| 1353 |
+
"logits/rejected": -1.6015625,
|
| 1354 |
+
"logps/chosen": -246.0,
|
| 1355 |
+
"logps/rejected": -266.0,
|
| 1356 |
+
"loss": 0.2021,
|
| 1357 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1358 |
+
"rewards/chosen": -0.1845703125,
|
| 1359 |
+
"rewards/margins": 1.9609375,
|
| 1360 |
+
"rewards/rejected": -2.140625,
|
| 1361 |
+
"step": 870
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.8115015974440896,
|
| 1365 |
+
"grad_norm": 13.38540049073994,
|
| 1366 |
+
"learning_rate": 3.4911242603550294e-08,
|
| 1367 |
+
"logits/chosen": -1.640625,
|
| 1368 |
+
"logits/rejected": -1.65625,
|
| 1369 |
+
"logps/chosen": -240.0,
|
| 1370 |
+
"logps/rejected": -270.0,
|
| 1371 |
+
"loss": 0.18,
|
| 1372 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1373 |
+
"rewards/chosen": -0.00701904296875,
|
| 1374 |
+
"rewards/margins": 2.25,
|
| 1375 |
+
"rewards/rejected": -2.265625,
|
| 1376 |
+
"step": 880
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 2.8434504792332267,
|
| 1380 |
+
"grad_norm": 17.91486249665347,
|
| 1381 |
+
"learning_rate": 2.8994082840236687e-08,
|
| 1382 |
+
"logits/chosen": -1.625,
|
| 1383 |
+
"logits/rejected": -1.6484375,
|
| 1384 |
+
"logps/chosen": -241.0,
|
| 1385 |
+
"logps/rejected": -264.0,
|
| 1386 |
+
"loss": 0.1964,
|
| 1387 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1388 |
+
"rewards/chosen": 0.057373046875,
|
| 1389 |
+
"rewards/margins": 2.109375,
|
| 1390 |
+
"rewards/rejected": -2.0625,
|
| 1391 |
+
"step": 890
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 2.8753993610223643,
|
| 1395 |
+
"grad_norm": 22.402213957503538,
|
| 1396 |
+
"learning_rate": 2.3076923076923076e-08,
|
| 1397 |
+
"logits/chosen": -1.625,
|
| 1398 |
+
"logits/rejected": -1.625,
|
| 1399 |
+
"logps/chosen": -240.0,
|
| 1400 |
+
"logps/rejected": -274.0,
|
| 1401 |
+
"loss": 0.1627,
|
| 1402 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1403 |
+
"rewards/chosen": 0.1201171875,
|
| 1404 |
+
"rewards/margins": 2.578125,
|
| 1405 |
+
"rewards/rejected": -2.46875,
|
| 1406 |
+
"step": 900
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.9073482428115014,
|
| 1410 |
+
"grad_norm": 16.63729631800027,
|
| 1411 |
+
"learning_rate": 1.7159763313609465e-08,
|
| 1412 |
+
"logits/chosen": -1.59375,
|
| 1413 |
+
"logits/rejected": -1.625,
|
| 1414 |
+
"logps/chosen": -238.0,
|
| 1415 |
+
"logps/rejected": -270.0,
|
| 1416 |
+
"loss": 0.2028,
|
| 1417 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1418 |
+
"rewards/chosen": 0.09765625,
|
| 1419 |
+
"rewards/margins": 2.265625,
|
| 1420 |
+
"rewards/rejected": -2.171875,
|
| 1421 |
+
"step": 910
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 2.939297124600639,
|
| 1425 |
+
"grad_norm": 13.806768716470549,
|
| 1426 |
+
"learning_rate": 1.1242603550295858e-08,
|
| 1427 |
+
"logits/chosen": -1.625,
|
| 1428 |
+
"logits/rejected": -1.6171875,
|
| 1429 |
+
"logps/chosen": -244.0,
|
| 1430 |
+
"logps/rejected": -268.0,
|
| 1431 |
+
"loss": 0.1701,
|
| 1432 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1433 |
+
"rewards/chosen": -0.208984375,
|
| 1434 |
+
"rewards/margins": 2.015625,
|
| 1435 |
+
"rewards/rejected": -2.21875,
|
| 1436 |
+
"step": 920
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 2.9712460063897765,
|
| 1440 |
+
"grad_norm": 12.821411876121983,
|
| 1441 |
+
"learning_rate": 5.325443786982248e-09,
|
| 1442 |
+
"logits/chosen": -1.6015625,
|
| 1443 |
+
"logits/rejected": -1.6171875,
|
| 1444 |
+
"logps/chosen": -243.0,
|
| 1445 |
+
"logps/rejected": -264.0,
|
| 1446 |
+
"loss": 0.1777,
|
| 1447 |
+
"rewards/accuracies": 0.9375,
|
| 1448 |
+
"rewards/chosen": -0.1640625,
|
| 1449 |
+
"rewards/margins": 2.171875,
|
| 1450 |
+
"rewards/rejected": -2.34375,
|
| 1451 |
+
"step": 930
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 3.0,
|
| 1455 |
+
"eval_logits/chosen": -1.640625,
|
| 1456 |
+
"eval_logits/rejected": -1.6484375,
|
| 1457 |
+
"eval_logps/chosen": -251.0,
|
| 1458 |
+
"eval_logps/rejected": -255.0,
|
| 1459 |
+
"eval_loss": 0.6092968583106995,
|
| 1460 |
+
"eval_rewards/accuracies": 0.6607142686843872,
|
| 1461 |
+
"eval_rewards/chosen": -0.83984375,
|
| 1462 |
+
"eval_rewards/margins": 0.56640625,
|
| 1463 |
+
"eval_rewards/rejected": -1.40625,
|
| 1464 |
+
"eval_runtime": 15.3355,
|
| 1465 |
+
"eval_samples_per_second": 13.042,
|
| 1466 |
+
"eval_steps_per_second": 0.456,
|
| 1467 |
+
"step": 939
|
| 1468 |
}
|
| 1469 |
],
|
| 1470 |
"logging_steps": 10,
|
|
|
|
| 1479 |
"should_evaluate": false,
|
| 1480 |
"should_log": false,
|
| 1481 |
"should_save": true,
|
| 1482 |
+
"should_training_stop": true
|
| 1483 |
},
|
| 1484 |
"attributes": {}
|
| 1485 |
}
|