Training in progress, epoch 3, checkpoint
Browse files- last-checkpoint/global_step1872/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1872/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1872/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step1872/zero_pp_rank_1_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/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +964 -3
last-checkpoint/global_step1872/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 sha256:62c1b0737d9ef833cc5447393be9645063199f9c62e660078c52e845aa32896b
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step1872/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 sha256:817359d4e1fee825f428c471135c1a7e867c2eed7ed945e8a13b0997f00d4daa
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step1872/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 sha256:2f20044a320dfaceff8d5f009e24b0c20b18aa9918609113a30e834ccbce7a2a
|
| 3 |
+
size 168021
|
last-checkpoint/global_step1872/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:e4c4a2977f7897858d681153444d2ebd90e00463a70a1038df61619876f7f93b
|
| 3 |
+
size 168021
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step1872
|
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 sha256:80fcbf5f63a031db5e7b7cc0059d8610275f60c9994a6affd53349d73d62303d
|
| 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:b40d8045439d7259e9b7354ad01954826a7161688838b7467a7e317daee489fd
|
| 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:f96435a411a27508b63821508e0f9ac352278204b2c94d715d3f47b37f0c7e59
|
| 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:f1cad27ebfd179a88097c9008af3e226adae1d17acab641c6855cd3bc85c3d7d
|
| 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 sha256:95430508d31bbe1a66a940e2572bf04addefae3e1c4e861e8657f66d302aa23e
|
| 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:b70a6983e6085768b4d2b447a8bd35374bce9cf4ea8e8fbefc1260ca2e054a70
|
| 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 sha256:2b15b652e34e702b048a2fe65a8149c25795b8f29765f41806646e987f007a10
|
| 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,
|
|
@@ -1914,6 +1914,967 @@
|
|
| 1914 |
"eval_samples_per_second": 6.892,
|
| 1915 |
"eval_steps_per_second": 0.448,
|
| 1916 |
"step": 1248
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1917 |
}
|
| 1918 |
],
|
| 1919 |
"logging_steps": 10,
|
|
@@ -1928,7 +2889,7 @@
|
|
| 1928 |
"should_evaluate": false,
|
| 1929 |
"should_log": false,
|
| 1930 |
"should_save": true,
|
| 1931 |
-
"should_training_stop":
|
| 1932 |
},
|
| 1933 |
"attributes": {}
|
| 1934 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1872,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1914 |
"eval_samples_per_second": 6.892,
|
| 1915 |
"eval_steps_per_second": 0.448,
|
| 1916 |
"step": 1248
|
| 1917 |
+
},
|
| 1918 |
+
{
|
| 1919 |
+
"epoch": 2.003205128205128,
|
| 1920 |
+
"grad_norm": 2.3679308142140885,
|
| 1921 |
+
"learning_rate": 1.846793349168646e-07,
|
| 1922 |
+
"logits/chosen": -0.341796875,
|
| 1923 |
+
"logits/rejected": -0.46484375,
|
| 1924 |
+
"logps/chosen": -366.0,
|
| 1925 |
+
"logps/rejected": -352.0,
|
| 1926 |
+
"loss": 0.0315,
|
| 1927 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1928 |
+
"rewards/chosen": 2.609375,
|
| 1929 |
+
"rewards/margins": 6.9375,
|
| 1930 |
+
"rewards/rejected": -4.3125,
|
| 1931 |
+
"step": 1250
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 2.019230769230769,
|
| 1935 |
+
"grad_norm": 2.590719502204235,
|
| 1936 |
+
"learning_rate": 1.8171021377672207e-07,
|
| 1937 |
+
"logits/chosen": -0.25,
|
| 1938 |
+
"logits/rejected": -0.404296875,
|
| 1939 |
+
"logps/chosen": -378.0,
|
| 1940 |
+
"logps/rejected": -364.0,
|
| 1941 |
+
"loss": 0.0081,
|
| 1942 |
+
"rewards/accuracies": 1.0,
|
| 1943 |
+
"rewards/chosen": 2.015625,
|
| 1944 |
+
"rewards/margins": 6.90625,
|
| 1945 |
+
"rewards/rejected": -4.875,
|
| 1946 |
+
"step": 1260
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 2.03525641025641,
|
| 1950 |
+
"grad_norm": 22.175507349451703,
|
| 1951 |
+
"learning_rate": 1.7874109263657958e-07,
|
| 1952 |
+
"logits/chosen": -0.2275390625,
|
| 1953 |
+
"logits/rejected": -0.37890625,
|
| 1954 |
+
"logps/chosen": -364.0,
|
| 1955 |
+
"logps/rejected": -384.0,
|
| 1956 |
+
"loss": 0.0156,
|
| 1957 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1958 |
+
"rewards/chosen": 2.0625,
|
| 1959 |
+
"rewards/margins": 7.375,
|
| 1960 |
+
"rewards/rejected": -5.3125,
|
| 1961 |
+
"step": 1270
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 2.051282051282051,
|
| 1965 |
+
"grad_norm": 4.568251724068892,
|
| 1966 |
+
"learning_rate": 1.7577197149643706e-07,
|
| 1967 |
+
"logits/chosen": -0.25,
|
| 1968 |
+
"logits/rejected": -0.482421875,
|
| 1969 |
+
"logps/chosen": -380.0,
|
| 1970 |
+
"logps/rejected": -370.0,
|
| 1971 |
+
"loss": 0.0145,
|
| 1972 |
+
"rewards/accuracies": 1.0,
|
| 1973 |
+
"rewards/chosen": 1.953125,
|
| 1974 |
+
"rewards/margins": 7.46875,
|
| 1975 |
+
"rewards/rejected": -5.53125,
|
| 1976 |
+
"step": 1280
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 2.0673076923076925,
|
| 1980 |
+
"grad_norm": 4.904685892165479,
|
| 1981 |
+
"learning_rate": 1.728028503562945e-07,
|
| 1982 |
+
"logits/chosen": -0.22265625,
|
| 1983 |
+
"logits/rejected": -0.466796875,
|
| 1984 |
+
"logps/chosen": -376.0,
|
| 1985 |
+
"logps/rejected": -366.0,
|
| 1986 |
+
"loss": 0.0088,
|
| 1987 |
+
"rewards/accuracies": 1.0,
|
| 1988 |
+
"rewards/chosen": 2.515625,
|
| 1989 |
+
"rewards/margins": 7.3125,
|
| 1990 |
+
"rewards/rejected": -4.78125,
|
| 1991 |
+
"step": 1290
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 2.0833333333333335,
|
| 1995 |
+
"grad_norm": 0.6187256370987634,
|
| 1996 |
+
"learning_rate": 1.6983372921615202e-07,
|
| 1997 |
+
"logits/chosen": -0.2138671875,
|
| 1998 |
+
"logits/rejected": -0.458984375,
|
| 1999 |
+
"logps/chosen": -362.0,
|
| 2000 |
+
"logps/rejected": -370.0,
|
| 2001 |
+
"loss": 0.0123,
|
| 2002 |
+
"rewards/accuracies": 1.0,
|
| 2003 |
+
"rewards/chosen": 2.265625,
|
| 2004 |
+
"rewards/margins": 7.625,
|
| 2005 |
+
"rewards/rejected": -5.375,
|
| 2006 |
+
"step": 1300
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 2.0993589743589745,
|
| 2010 |
+
"grad_norm": 0.7331904803685036,
|
| 2011 |
+
"learning_rate": 1.668646080760095e-07,
|
| 2012 |
+
"logits/chosen": -0.15234375,
|
| 2013 |
+
"logits/rejected": -0.40625,
|
| 2014 |
+
"logps/chosen": -354.0,
|
| 2015 |
+
"logps/rejected": -356.0,
|
| 2016 |
+
"loss": 0.0061,
|
| 2017 |
+
"rewards/accuracies": 1.0,
|
| 2018 |
+
"rewards/chosen": 2.59375,
|
| 2019 |
+
"rewards/margins": 7.65625,
|
| 2020 |
+
"rewards/rejected": -5.0625,
|
| 2021 |
+
"step": 1310
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 2.1153846153846154,
|
| 2025 |
+
"grad_norm": 2.0296944258203036,
|
| 2026 |
+
"learning_rate": 1.6389548693586697e-07,
|
| 2027 |
+
"logits/chosen": -0.2412109375,
|
| 2028 |
+
"logits/rejected": -0.1748046875,
|
| 2029 |
+
"logps/chosen": -348.0,
|
| 2030 |
+
"logps/rejected": -348.0,
|
| 2031 |
+
"loss": 0.0231,
|
| 2032 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2033 |
+
"rewards/chosen": 2.484375,
|
| 2034 |
+
"rewards/margins": 7.125,
|
| 2035 |
+
"rewards/rejected": -4.625,
|
| 2036 |
+
"step": 1320
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 2.1314102564102564,
|
| 2040 |
+
"grad_norm": 23.71632295557207,
|
| 2041 |
+
"learning_rate": 1.6092636579572448e-07,
|
| 2042 |
+
"logits/chosen": -0.2275390625,
|
| 2043 |
+
"logits/rejected": -0.3125,
|
| 2044 |
+
"logps/chosen": -358.0,
|
| 2045 |
+
"logps/rejected": -342.0,
|
| 2046 |
+
"loss": 0.019,
|
| 2047 |
+
"rewards/accuracies": 1.0,
|
| 2048 |
+
"rewards/chosen": 2.78125,
|
| 2049 |
+
"rewards/margins": 7.90625,
|
| 2050 |
+
"rewards/rejected": -5.125,
|
| 2051 |
+
"step": 1330
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 2.1474358974358974,
|
| 2055 |
+
"grad_norm": 4.284093803737362,
|
| 2056 |
+
"learning_rate": 1.5795724465558193e-07,
|
| 2057 |
+
"logits/chosen": -0.2255859375,
|
| 2058 |
+
"logits/rejected": -0.5546875,
|
| 2059 |
+
"logps/chosen": -364.0,
|
| 2060 |
+
"logps/rejected": -384.0,
|
| 2061 |
+
"loss": 0.0159,
|
| 2062 |
+
"rewards/accuracies": 1.0,
|
| 2063 |
+
"rewards/chosen": 2.484375,
|
| 2064 |
+
"rewards/margins": 7.4375,
|
| 2065 |
+
"rewards/rejected": -4.9375,
|
| 2066 |
+
"step": 1340
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 2.1634615384615383,
|
| 2070 |
+
"grad_norm": 0.7862983762455222,
|
| 2071 |
+
"learning_rate": 1.549881235154394e-07,
|
| 2072 |
+
"logits/chosen": -0.1435546875,
|
| 2073 |
+
"logits/rejected": -0.337890625,
|
| 2074 |
+
"logps/chosen": -376.0,
|
| 2075 |
+
"logps/rejected": -354.0,
|
| 2076 |
+
"loss": 0.0114,
|
| 2077 |
+
"rewards/accuracies": 1.0,
|
| 2078 |
+
"rewards/chosen": 2.828125,
|
| 2079 |
+
"rewards/margins": 7.40625,
|
| 2080 |
+
"rewards/rejected": -4.59375,
|
| 2081 |
+
"step": 1350
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 2.1794871794871793,
|
| 2085 |
+
"grad_norm": 1.8750426567792726,
|
| 2086 |
+
"learning_rate": 1.520190023752969e-07,
|
| 2087 |
+
"logits/chosen": -0.259765625,
|
| 2088 |
+
"logits/rejected": -0.59765625,
|
| 2089 |
+
"logps/chosen": -364.0,
|
| 2090 |
+
"logps/rejected": -330.0,
|
| 2091 |
+
"loss": 0.0138,
|
| 2092 |
+
"rewards/accuracies": 1.0,
|
| 2093 |
+
"rewards/chosen": 3.015625,
|
| 2094 |
+
"rewards/margins": 8.1875,
|
| 2095 |
+
"rewards/rejected": -5.15625,
|
| 2096 |
+
"step": 1360
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 2.1955128205128207,
|
| 2100 |
+
"grad_norm": 7.170950559094129,
|
| 2101 |
+
"learning_rate": 1.490498812351544e-07,
|
| 2102 |
+
"logits/chosen": -0.251953125,
|
| 2103 |
+
"logits/rejected": -0.36328125,
|
| 2104 |
+
"logps/chosen": -362.0,
|
| 2105 |
+
"logps/rejected": -356.0,
|
| 2106 |
+
"loss": 0.0323,
|
| 2107 |
+
"rewards/accuracies": 1.0,
|
| 2108 |
+
"rewards/chosen": 3.0,
|
| 2109 |
+
"rewards/margins": 8.1875,
|
| 2110 |
+
"rewards/rejected": -5.1875,
|
| 2111 |
+
"step": 1370
|
| 2112 |
+
},
|
| 2113 |
+
{
|
| 2114 |
+
"epoch": 2.2115384615384617,
|
| 2115 |
+
"grad_norm": 21.23713813309802,
|
| 2116 |
+
"learning_rate": 1.4608076009501184e-07,
|
| 2117 |
+
"logits/chosen": -0.2431640625,
|
| 2118 |
+
"logits/rejected": -0.2353515625,
|
| 2119 |
+
"logps/chosen": -366.0,
|
| 2120 |
+
"logps/rejected": -356.0,
|
| 2121 |
+
"loss": 0.024,
|
| 2122 |
+
"rewards/accuracies": 1.0,
|
| 2123 |
+
"rewards/chosen": 2.6875,
|
| 2124 |
+
"rewards/margins": 7.8125,
|
| 2125 |
+
"rewards/rejected": -5.125,
|
| 2126 |
+
"step": 1380
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 2.2275641025641026,
|
| 2130 |
+
"grad_norm": 7.169471258104354,
|
| 2131 |
+
"learning_rate": 1.4311163895486935e-07,
|
| 2132 |
+
"logits/chosen": -0.18359375,
|
| 2133 |
+
"logits/rejected": -0.29296875,
|
| 2134 |
+
"logps/chosen": -348.0,
|
| 2135 |
+
"logps/rejected": -364.0,
|
| 2136 |
+
"loss": 0.0257,
|
| 2137 |
+
"rewards/accuracies": 1.0,
|
| 2138 |
+
"rewards/chosen": 2.90625,
|
| 2139 |
+
"rewards/margins": 8.1875,
|
| 2140 |
+
"rewards/rejected": -5.25,
|
| 2141 |
+
"step": 1390
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 2.2435897435897436,
|
| 2145 |
+
"grad_norm": 4.891546899421327,
|
| 2146 |
+
"learning_rate": 1.4014251781472683e-07,
|
| 2147 |
+
"logits/chosen": -0.22265625,
|
| 2148 |
+
"logits/rejected": -0.55078125,
|
| 2149 |
+
"logps/chosen": -344.0,
|
| 2150 |
+
"logps/rejected": -374.0,
|
| 2151 |
+
"loss": 0.0165,
|
| 2152 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2153 |
+
"rewards/chosen": 2.71875,
|
| 2154 |
+
"rewards/margins": 7.4375,
|
| 2155 |
+
"rewards/rejected": -4.71875,
|
| 2156 |
+
"step": 1400
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 2.2596153846153846,
|
| 2160 |
+
"grad_norm": 14.0088561337579,
|
| 2161 |
+
"learning_rate": 1.3717339667458433e-07,
|
| 2162 |
+
"logits/chosen": -0.171875,
|
| 2163 |
+
"logits/rejected": -0.375,
|
| 2164 |
+
"logps/chosen": -336.0,
|
| 2165 |
+
"logps/rejected": -380.0,
|
| 2166 |
+
"loss": 0.023,
|
| 2167 |
+
"rewards/accuracies": 1.0,
|
| 2168 |
+
"rewards/chosen": 2.328125,
|
| 2169 |
+
"rewards/margins": 8.125,
|
| 2170 |
+
"rewards/rejected": -5.8125,
|
| 2171 |
+
"step": 1410
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 2.2756410256410255,
|
| 2175 |
+
"grad_norm": 5.478498841262275,
|
| 2176 |
+
"learning_rate": 1.342042755344418e-07,
|
| 2177 |
+
"logits/chosen": -0.244140625,
|
| 2178 |
+
"logits/rejected": -0.60546875,
|
| 2179 |
+
"logps/chosen": -340.0,
|
| 2180 |
+
"logps/rejected": -368.0,
|
| 2181 |
+
"loss": 0.0101,
|
| 2182 |
+
"rewards/accuracies": 1.0,
|
| 2183 |
+
"rewards/chosen": 2.078125,
|
| 2184 |
+
"rewards/margins": 7.125,
|
| 2185 |
+
"rewards/rejected": -5.0625,
|
| 2186 |
+
"step": 1420
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 2.2916666666666665,
|
| 2190 |
+
"grad_norm": 4.08870815918569,
|
| 2191 |
+
"learning_rate": 1.3123515439429926e-07,
|
| 2192 |
+
"logits/chosen": -0.216796875,
|
| 2193 |
+
"logits/rejected": -0.5,
|
| 2194 |
+
"logps/chosen": -348.0,
|
| 2195 |
+
"logps/rejected": -380.0,
|
| 2196 |
+
"loss": 0.0094,
|
| 2197 |
+
"rewards/accuracies": 1.0,
|
| 2198 |
+
"rewards/chosen": 2.40625,
|
| 2199 |
+
"rewards/margins": 7.875,
|
| 2200 |
+
"rewards/rejected": -5.46875,
|
| 2201 |
+
"step": 1430
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 2.3076923076923075,
|
| 2205 |
+
"grad_norm": 4.0828939935609885,
|
| 2206 |
+
"learning_rate": 1.2826603325415677e-07,
|
| 2207 |
+
"logits/chosen": -0.310546875,
|
| 2208 |
+
"logits/rejected": -0.3515625,
|
| 2209 |
+
"logps/chosen": -362.0,
|
| 2210 |
+
"logps/rejected": -372.0,
|
| 2211 |
+
"loss": 0.0115,
|
| 2212 |
+
"rewards/accuracies": 1.0,
|
| 2213 |
+
"rewards/chosen": 1.9296875,
|
| 2214 |
+
"rewards/margins": 7.90625,
|
| 2215 |
+
"rewards/rejected": -5.96875,
|
| 2216 |
+
"step": 1440
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 2.323717948717949,
|
| 2220 |
+
"grad_norm": 2.2611774310220714,
|
| 2221 |
+
"learning_rate": 1.2529691211401425e-07,
|
| 2222 |
+
"logits/chosen": -0.2890625,
|
| 2223 |
+
"logits/rejected": -0.3671875,
|
| 2224 |
+
"logps/chosen": -350.0,
|
| 2225 |
+
"logps/rejected": -380.0,
|
| 2226 |
+
"loss": 0.0071,
|
| 2227 |
+
"rewards/accuracies": 1.0,
|
| 2228 |
+
"rewards/chosen": 2.359375,
|
| 2229 |
+
"rewards/margins": 8.1875,
|
| 2230 |
+
"rewards/rejected": -5.84375,
|
| 2231 |
+
"step": 1450
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 2.33974358974359,
|
| 2235 |
+
"grad_norm": 35.51723177781465,
|
| 2236 |
+
"learning_rate": 1.2232779097387173e-07,
|
| 2237 |
+
"logits/chosen": -0.244140625,
|
| 2238 |
+
"logits/rejected": -0.36328125,
|
| 2239 |
+
"logps/chosen": -378.0,
|
| 2240 |
+
"logps/rejected": -368.0,
|
| 2241 |
+
"loss": 0.0773,
|
| 2242 |
+
"rewards/accuracies": 1.0,
|
| 2243 |
+
"rewards/chosen": 2.078125,
|
| 2244 |
+
"rewards/margins": 8.125,
|
| 2245 |
+
"rewards/rejected": -6.0,
|
| 2246 |
+
"step": 1460
|
| 2247 |
+
},
|
| 2248 |
+
{
|
| 2249 |
+
"epoch": 2.355769230769231,
|
| 2250 |
+
"grad_norm": 5.534009913713299,
|
| 2251 |
+
"learning_rate": 1.193586698337292e-07,
|
| 2252 |
+
"logits/chosen": -0.1767578125,
|
| 2253 |
+
"logits/rejected": -0.5390625,
|
| 2254 |
+
"logps/chosen": -368.0,
|
| 2255 |
+
"logps/rejected": -358.0,
|
| 2256 |
+
"loss": 0.012,
|
| 2257 |
+
"rewards/accuracies": 1.0,
|
| 2258 |
+
"rewards/chosen": 2.484375,
|
| 2259 |
+
"rewards/margins": 8.0625,
|
| 2260 |
+
"rewards/rejected": -5.59375,
|
| 2261 |
+
"step": 1470
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 2.371794871794872,
|
| 2265 |
+
"grad_norm": 18.31189968205467,
|
| 2266 |
+
"learning_rate": 1.163895486935867e-07,
|
| 2267 |
+
"logits/chosen": -0.25390625,
|
| 2268 |
+
"logits/rejected": -0.353515625,
|
| 2269 |
+
"logps/chosen": -368.0,
|
| 2270 |
+
"logps/rejected": -356.0,
|
| 2271 |
+
"loss": 0.0359,
|
| 2272 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2273 |
+
"rewards/chosen": 2.3125,
|
| 2274 |
+
"rewards/margins": 7.78125,
|
| 2275 |
+
"rewards/rejected": -5.46875,
|
| 2276 |
+
"step": 1480
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 2.3878205128205128,
|
| 2280 |
+
"grad_norm": 14.190561755907082,
|
| 2281 |
+
"learning_rate": 1.1342042755344417e-07,
|
| 2282 |
+
"logits/chosen": -0.1484375,
|
| 2283 |
+
"logits/rejected": -0.35546875,
|
| 2284 |
+
"logps/chosen": -376.0,
|
| 2285 |
+
"logps/rejected": -370.0,
|
| 2286 |
+
"loss": 0.0065,
|
| 2287 |
+
"rewards/accuracies": 1.0,
|
| 2288 |
+
"rewards/chosen": 2.734375,
|
| 2289 |
+
"rewards/margins": 8.4375,
|
| 2290 |
+
"rewards/rejected": -5.6875,
|
| 2291 |
+
"step": 1490
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 2.4038461538461537,
|
| 2295 |
+
"grad_norm": 57.38633989383038,
|
| 2296 |
+
"learning_rate": 1.1045130641330165e-07,
|
| 2297 |
+
"logits/chosen": -0.2177734375,
|
| 2298 |
+
"logits/rejected": -0.24609375,
|
| 2299 |
+
"logps/chosen": -356.0,
|
| 2300 |
+
"logps/rejected": -376.0,
|
| 2301 |
+
"loss": 0.0762,
|
| 2302 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2303 |
+
"rewards/chosen": 1.25,
|
| 2304 |
+
"rewards/margins": 7.25,
|
| 2305 |
+
"rewards/rejected": -6.0,
|
| 2306 |
+
"step": 1500
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 2.4198717948717947,
|
| 2310 |
+
"grad_norm": 13.945337786807688,
|
| 2311 |
+
"learning_rate": 1.0748218527315913e-07,
|
| 2312 |
+
"logits/chosen": -0.2353515625,
|
| 2313 |
+
"logits/rejected": -0.486328125,
|
| 2314 |
+
"logps/chosen": -374.0,
|
| 2315 |
+
"logps/rejected": -386.0,
|
| 2316 |
+
"loss": 0.0146,
|
| 2317 |
+
"rewards/accuracies": 1.0,
|
| 2318 |
+
"rewards/chosen": 2.28125,
|
| 2319 |
+
"rewards/margins": 8.625,
|
| 2320 |
+
"rewards/rejected": -6.34375,
|
| 2321 |
+
"step": 1510
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 2.435897435897436,
|
| 2325 |
+
"grad_norm": 2.025562206086781,
|
| 2326 |
+
"learning_rate": 1.0451306413301662e-07,
|
| 2327 |
+
"logits/chosen": -0.2890625,
|
| 2328 |
+
"logits/rejected": -0.404296875,
|
| 2329 |
+
"logps/chosen": -362.0,
|
| 2330 |
+
"logps/rejected": -360.0,
|
| 2331 |
+
"loss": 0.0058,
|
| 2332 |
+
"rewards/accuracies": 1.0,
|
| 2333 |
+
"rewards/chosen": 2.96875,
|
| 2334 |
+
"rewards/margins": 8.875,
|
| 2335 |
+
"rewards/rejected": -5.90625,
|
| 2336 |
+
"step": 1520
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 2.451923076923077,
|
| 2340 |
+
"grad_norm": 2.5818458781278997,
|
| 2341 |
+
"learning_rate": 1.0154394299287411e-07,
|
| 2342 |
+
"logits/chosen": -0.20703125,
|
| 2343 |
+
"logits/rejected": -0.380859375,
|
| 2344 |
+
"logps/chosen": -356.0,
|
| 2345 |
+
"logps/rejected": -378.0,
|
| 2346 |
+
"loss": 0.0084,
|
| 2347 |
+
"rewards/accuracies": 1.0,
|
| 2348 |
+
"rewards/chosen": 2.75,
|
| 2349 |
+
"rewards/margins": 8.375,
|
| 2350 |
+
"rewards/rejected": -5.59375,
|
| 2351 |
+
"step": 1530
|
| 2352 |
+
},
|
| 2353 |
+
{
|
| 2354 |
+
"epoch": 2.467948717948718,
|
| 2355 |
+
"grad_norm": 0.5147935469188503,
|
| 2356 |
+
"learning_rate": 9.857482185273158e-08,
|
| 2357 |
+
"logits/chosen": -0.240234375,
|
| 2358 |
+
"logits/rejected": -0.47265625,
|
| 2359 |
+
"logps/chosen": -348.0,
|
| 2360 |
+
"logps/rejected": -354.0,
|
| 2361 |
+
"loss": 0.0156,
|
| 2362 |
+
"rewards/accuracies": 1.0,
|
| 2363 |
+
"rewards/chosen": 2.609375,
|
| 2364 |
+
"rewards/margins": 8.125,
|
| 2365 |
+
"rewards/rejected": -5.46875,
|
| 2366 |
+
"step": 1540
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 2.483974358974359,
|
| 2370 |
+
"grad_norm": 2.5527476910216764,
|
| 2371 |
+
"learning_rate": 9.560570071258907e-08,
|
| 2372 |
+
"logits/chosen": -0.1748046875,
|
| 2373 |
+
"logits/rejected": -0.392578125,
|
| 2374 |
+
"logps/chosen": -378.0,
|
| 2375 |
+
"logps/rejected": -336.0,
|
| 2376 |
+
"loss": 0.0132,
|
| 2377 |
+
"rewards/accuracies": 1.0,
|
| 2378 |
+
"rewards/chosen": 2.34375,
|
| 2379 |
+
"rewards/margins": 8.125,
|
| 2380 |
+
"rewards/rejected": -5.8125,
|
| 2381 |
+
"step": 1550
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 2.5,
|
| 2385 |
+
"grad_norm": 2.6084669807533807,
|
| 2386 |
+
"learning_rate": 9.263657957244655e-08,
|
| 2387 |
+
"logits/chosen": -0.28125,
|
| 2388 |
+
"logits/rejected": -0.40234375,
|
| 2389 |
+
"logps/chosen": -340.0,
|
| 2390 |
+
"logps/rejected": -386.0,
|
| 2391 |
+
"loss": 0.0112,
|
| 2392 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2393 |
+
"rewards/chosen": 2.78125,
|
| 2394 |
+
"rewards/margins": 8.4375,
|
| 2395 |
+
"rewards/rejected": -5.6875,
|
| 2396 |
+
"step": 1560
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 2.516025641025641,
|
| 2400 |
+
"grad_norm": 0.5142214115866551,
|
| 2401 |
+
"learning_rate": 8.966745843230403e-08,
|
| 2402 |
+
"logits/chosen": -0.26953125,
|
| 2403 |
+
"logits/rejected": -0.34765625,
|
| 2404 |
+
"logps/chosen": -332.0,
|
| 2405 |
+
"logps/rejected": -338.0,
|
| 2406 |
+
"loss": 0.0087,
|
| 2407 |
+
"rewards/accuracies": 1.0,
|
| 2408 |
+
"rewards/chosen": 2.625,
|
| 2409 |
+
"rewards/margins": 8.25,
|
| 2410 |
+
"rewards/rejected": -5.625,
|
| 2411 |
+
"step": 1570
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 2.532051282051282,
|
| 2415 |
+
"grad_norm": 0.4686133876015494,
|
| 2416 |
+
"learning_rate": 8.669833729216151e-08,
|
| 2417 |
+
"logits/chosen": -0.2080078125,
|
| 2418 |
+
"logits/rejected": -0.36328125,
|
| 2419 |
+
"logps/chosen": -374.0,
|
| 2420 |
+
"logps/rejected": -382.0,
|
| 2421 |
+
"loss": 0.0066,
|
| 2422 |
+
"rewards/accuracies": 1.0,
|
| 2423 |
+
"rewards/chosen": 2.0625,
|
| 2424 |
+
"rewards/margins": 7.84375,
|
| 2425 |
+
"rewards/rejected": -5.78125,
|
| 2426 |
+
"step": 1580
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 2.5480769230769234,
|
| 2430 |
+
"grad_norm": 4.377458752115824,
|
| 2431 |
+
"learning_rate": 8.3729216152019e-08,
|
| 2432 |
+
"logits/chosen": -0.240234375,
|
| 2433 |
+
"logits/rejected": -0.50390625,
|
| 2434 |
+
"logps/chosen": -358.0,
|
| 2435 |
+
"logps/rejected": -370.0,
|
| 2436 |
+
"loss": 0.0077,
|
| 2437 |
+
"rewards/accuracies": 1.0,
|
| 2438 |
+
"rewards/chosen": 2.375,
|
| 2439 |
+
"rewards/margins": 8.4375,
|
| 2440 |
+
"rewards/rejected": -6.03125,
|
| 2441 |
+
"step": 1590
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 2.564102564102564,
|
| 2445 |
+
"grad_norm": 8.575422939319253,
|
| 2446 |
+
"learning_rate": 8.076009501187649e-08,
|
| 2447 |
+
"logits/chosen": -0.2490234375,
|
| 2448 |
+
"logits/rejected": -0.337890625,
|
| 2449 |
+
"logps/chosen": -348.0,
|
| 2450 |
+
"logps/rejected": -370.0,
|
| 2451 |
+
"loss": 0.0241,
|
| 2452 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2453 |
+
"rewards/chosen": 2.484375,
|
| 2454 |
+
"rewards/margins": 7.5,
|
| 2455 |
+
"rewards/rejected": -5.0,
|
| 2456 |
+
"step": 1600
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 2.5801282051282053,
|
| 2460 |
+
"grad_norm": 2.551052943978534,
|
| 2461 |
+
"learning_rate": 7.779097387173396e-08,
|
| 2462 |
+
"logits/chosen": -0.2734375,
|
| 2463 |
+
"logits/rejected": -0.375,
|
| 2464 |
+
"logps/chosen": -342.0,
|
| 2465 |
+
"logps/rejected": -360.0,
|
| 2466 |
+
"loss": 0.0082,
|
| 2467 |
+
"rewards/accuracies": 1.0,
|
| 2468 |
+
"rewards/chosen": 2.8125,
|
| 2469 |
+
"rewards/margins": 7.8125,
|
| 2470 |
+
"rewards/rejected": -4.96875,
|
| 2471 |
+
"step": 1610
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 2.5961538461538463,
|
| 2475 |
+
"grad_norm": 6.000172766868698,
|
| 2476 |
+
"learning_rate": 7.482185273159145e-08,
|
| 2477 |
+
"logits/chosen": -0.263671875,
|
| 2478 |
+
"logits/rejected": -0.451171875,
|
| 2479 |
+
"logps/chosen": -368.0,
|
| 2480 |
+
"logps/rejected": -378.0,
|
| 2481 |
+
"loss": 0.0303,
|
| 2482 |
+
"rewards/accuracies": 1.0,
|
| 2483 |
+
"rewards/chosen": 2.46875,
|
| 2484 |
+
"rewards/margins": 7.96875,
|
| 2485 |
+
"rewards/rejected": -5.5,
|
| 2486 |
+
"step": 1620
|
| 2487 |
+
},
|
| 2488 |
+
{
|
| 2489 |
+
"epoch": 2.6121794871794872,
|
| 2490 |
+
"grad_norm": 19.722033841376266,
|
| 2491 |
+
"learning_rate": 7.185273159144893e-08,
|
| 2492 |
+
"logits/chosen": -0.294921875,
|
| 2493 |
+
"logits/rejected": -0.392578125,
|
| 2494 |
+
"logps/chosen": -366.0,
|
| 2495 |
+
"logps/rejected": -398.0,
|
| 2496 |
+
"loss": 0.0123,
|
| 2497 |
+
"rewards/accuracies": 1.0,
|
| 2498 |
+
"rewards/chosen": 2.71875,
|
| 2499 |
+
"rewards/margins": 8.25,
|
| 2500 |
+
"rewards/rejected": -5.53125,
|
| 2501 |
+
"step": 1630
|
| 2502 |
+
},
|
| 2503 |
+
{
|
| 2504 |
+
"epoch": 2.628205128205128,
|
| 2505 |
+
"grad_norm": 11.30540085696599,
|
| 2506 |
+
"learning_rate": 6.88836104513064e-08,
|
| 2507 |
+
"logits/chosen": -0.18359375,
|
| 2508 |
+
"logits/rejected": -0.314453125,
|
| 2509 |
+
"logps/chosen": -380.0,
|
| 2510 |
+
"logps/rejected": -354.0,
|
| 2511 |
+
"loss": 0.0078,
|
| 2512 |
+
"rewards/accuracies": 1.0,
|
| 2513 |
+
"rewards/chosen": 2.734375,
|
| 2514 |
+
"rewards/margins": 8.6875,
|
| 2515 |
+
"rewards/rejected": -5.96875,
|
| 2516 |
+
"step": 1640
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 2.644230769230769,
|
| 2520 |
+
"grad_norm": 1.1773482574998353,
|
| 2521 |
+
"learning_rate": 6.591448931116388e-08,
|
| 2522 |
+
"logits/chosen": -0.162109375,
|
| 2523 |
+
"logits/rejected": -0.4609375,
|
| 2524 |
+
"logps/chosen": -378.0,
|
| 2525 |
+
"logps/rejected": -374.0,
|
| 2526 |
+
"loss": 0.0175,
|
| 2527 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2528 |
+
"rewards/chosen": 2.125,
|
| 2529 |
+
"rewards/margins": 8.0625,
|
| 2530 |
+
"rewards/rejected": -5.9375,
|
| 2531 |
+
"step": 1650
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 2.66025641025641,
|
| 2535 |
+
"grad_norm": 52.67450880620409,
|
| 2536 |
+
"learning_rate": 6.294536817102138e-08,
|
| 2537 |
+
"logits/chosen": -0.294921875,
|
| 2538 |
+
"logits/rejected": -0.58203125,
|
| 2539 |
+
"logps/chosen": -386.0,
|
| 2540 |
+
"logps/rejected": -376.0,
|
| 2541 |
+
"loss": 0.0316,
|
| 2542 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2543 |
+
"rewards/chosen": 1.828125,
|
| 2544 |
+
"rewards/margins": 7.53125,
|
| 2545 |
+
"rewards/rejected": -5.6875,
|
| 2546 |
+
"step": 1660
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 2.676282051282051,
|
| 2550 |
+
"grad_norm": 2.444649208747581,
|
| 2551 |
+
"learning_rate": 5.997624703087885e-08,
|
| 2552 |
+
"logits/chosen": -0.2734375,
|
| 2553 |
+
"logits/rejected": -0.30859375,
|
| 2554 |
+
"logps/chosen": -348.0,
|
| 2555 |
+
"logps/rejected": -330.0,
|
| 2556 |
+
"loss": 0.0139,
|
| 2557 |
+
"rewards/accuracies": 1.0,
|
| 2558 |
+
"rewards/chosen": 2.34375,
|
| 2559 |
+
"rewards/margins": 7.96875,
|
| 2560 |
+
"rewards/rejected": -5.625,
|
| 2561 |
+
"step": 1670
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 2.6923076923076925,
|
| 2565 |
+
"grad_norm": 1.8309710467577502,
|
| 2566 |
+
"learning_rate": 5.700712589073634e-08,
|
| 2567 |
+
"logits/chosen": -0.33203125,
|
| 2568 |
+
"logits/rejected": -0.5546875,
|
| 2569 |
+
"logps/chosen": -368.0,
|
| 2570 |
+
"logps/rejected": -372.0,
|
| 2571 |
+
"loss": 0.0044,
|
| 2572 |
+
"rewards/accuracies": 1.0,
|
| 2573 |
+
"rewards/chosen": 2.578125,
|
| 2574 |
+
"rewards/margins": 8.3125,
|
| 2575 |
+
"rewards/rejected": -5.75,
|
| 2576 |
+
"step": 1680
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 2.7083333333333335,
|
| 2580 |
+
"grad_norm": 0.974183927512936,
|
| 2581 |
+
"learning_rate": 5.4038004750593824e-08,
|
| 2582 |
+
"logits/chosen": -0.212890625,
|
| 2583 |
+
"logits/rejected": -0.384765625,
|
| 2584 |
+
"logps/chosen": -370.0,
|
| 2585 |
+
"logps/rejected": -366.0,
|
| 2586 |
+
"loss": 0.0061,
|
| 2587 |
+
"rewards/accuracies": 1.0,
|
| 2588 |
+
"rewards/chosen": 2.4375,
|
| 2589 |
+
"rewards/margins": 8.625,
|
| 2590 |
+
"rewards/rejected": -6.21875,
|
| 2591 |
+
"step": 1690
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 2.7243589743589745,
|
| 2595 |
+
"grad_norm": 8.112271810670284,
|
| 2596 |
+
"learning_rate": 5.10688836104513e-08,
|
| 2597 |
+
"logits/chosen": -0.287109375,
|
| 2598 |
+
"logits/rejected": -0.34375,
|
| 2599 |
+
"logps/chosen": -364.0,
|
| 2600 |
+
"logps/rejected": -384.0,
|
| 2601 |
+
"loss": 0.0074,
|
| 2602 |
+
"rewards/accuracies": 1.0,
|
| 2603 |
+
"rewards/chosen": 2.53125,
|
| 2604 |
+
"rewards/margins": 8.625,
|
| 2605 |
+
"rewards/rejected": -6.09375,
|
| 2606 |
+
"step": 1700
|
| 2607 |
+
},
|
| 2608 |
+
{
|
| 2609 |
+
"epoch": 2.7403846153846154,
|
| 2610 |
+
"grad_norm": 48.323303487774965,
|
| 2611 |
+
"learning_rate": 4.809976247030879e-08,
|
| 2612 |
+
"logits/chosen": -0.2734375,
|
| 2613 |
+
"logits/rejected": -0.39453125,
|
| 2614 |
+
"logps/chosen": -378.0,
|
| 2615 |
+
"logps/rejected": -368.0,
|
| 2616 |
+
"loss": 0.0187,
|
| 2617 |
+
"rewards/accuracies": 1.0,
|
| 2618 |
+
"rewards/chosen": 2.28125,
|
| 2619 |
+
"rewards/margins": 7.9375,
|
| 2620 |
+
"rewards/rejected": -5.65625,
|
| 2621 |
+
"step": 1710
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 2.7564102564102564,
|
| 2625 |
+
"grad_norm": 0.6216279700601488,
|
| 2626 |
+
"learning_rate": 4.5130641330166267e-08,
|
| 2627 |
+
"logits/chosen": -0.2470703125,
|
| 2628 |
+
"logits/rejected": -0.3828125,
|
| 2629 |
+
"logps/chosen": -340.0,
|
| 2630 |
+
"logps/rejected": -372.0,
|
| 2631 |
+
"loss": 0.0305,
|
| 2632 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2633 |
+
"rewards/chosen": 2.25,
|
| 2634 |
+
"rewards/margins": 7.6875,
|
| 2635 |
+
"rewards/rejected": -5.4375,
|
| 2636 |
+
"step": 1720
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 2.7724358974358974,
|
| 2640 |
+
"grad_norm": 8.386988307728698,
|
| 2641 |
+
"learning_rate": 4.216152019002375e-08,
|
| 2642 |
+
"logits/chosen": -0.2060546875,
|
| 2643 |
+
"logits/rejected": -0.349609375,
|
| 2644 |
+
"logps/chosen": -360.0,
|
| 2645 |
+
"logps/rejected": -364.0,
|
| 2646 |
+
"loss": 0.0062,
|
| 2647 |
+
"rewards/accuracies": 1.0,
|
| 2648 |
+
"rewards/chosen": 2.734375,
|
| 2649 |
+
"rewards/margins": 8.4375,
|
| 2650 |
+
"rewards/rejected": -5.6875,
|
| 2651 |
+
"step": 1730
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 2.7884615384615383,
|
| 2655 |
+
"grad_norm": 2.6310329285858587,
|
| 2656 |
+
"learning_rate": 3.919239904988123e-08,
|
| 2657 |
+
"logits/chosen": -0.1826171875,
|
| 2658 |
+
"logits/rejected": -0.333984375,
|
| 2659 |
+
"logps/chosen": -358.0,
|
| 2660 |
+
"logps/rejected": -390.0,
|
| 2661 |
+
"loss": 0.0088,
|
| 2662 |
+
"rewards/accuracies": 1.0,
|
| 2663 |
+
"rewards/chosen": 2.28125,
|
| 2664 |
+
"rewards/margins": 7.875,
|
| 2665 |
+
"rewards/rejected": -5.59375,
|
| 2666 |
+
"step": 1740
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 2.8044871794871797,
|
| 2670 |
+
"grad_norm": 3.066349322026837,
|
| 2671 |
+
"learning_rate": 3.6223277909738715e-08,
|
| 2672 |
+
"logits/chosen": -0.2236328125,
|
| 2673 |
+
"logits/rejected": -0.357421875,
|
| 2674 |
+
"logps/chosen": -350.0,
|
| 2675 |
+
"logps/rejected": -368.0,
|
| 2676 |
+
"loss": 0.0092,
|
| 2677 |
+
"rewards/accuracies": 1.0,
|
| 2678 |
+
"rewards/chosen": 2.40625,
|
| 2679 |
+
"rewards/margins": 7.96875,
|
| 2680 |
+
"rewards/rejected": -5.5625,
|
| 2681 |
+
"step": 1750
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 2.8205128205128203,
|
| 2685 |
+
"grad_norm": 8.079242737705172,
|
| 2686 |
+
"learning_rate": 3.32541567695962e-08,
|
| 2687 |
+
"logits/chosen": -0.0966796875,
|
| 2688 |
+
"logits/rejected": -0.2431640625,
|
| 2689 |
+
"logps/chosen": -342.0,
|
| 2690 |
+
"logps/rejected": -360.0,
|
| 2691 |
+
"loss": 0.0237,
|
| 2692 |
+
"rewards/accuracies": 1.0,
|
| 2693 |
+
"rewards/chosen": 2.421875,
|
| 2694 |
+
"rewards/margins": 8.4375,
|
| 2695 |
+
"rewards/rejected": -6.03125,
|
| 2696 |
+
"step": 1760
|
| 2697 |
+
},
|
| 2698 |
+
{
|
| 2699 |
+
"epoch": 2.8365384615384617,
|
| 2700 |
+
"grad_norm": 36.85905904660706,
|
| 2701 |
+
"learning_rate": 3.028503562945368e-08,
|
| 2702 |
+
"logits/chosen": -0.1962890625,
|
| 2703 |
+
"logits/rejected": -0.1435546875,
|
| 2704 |
+
"logps/chosen": -376.0,
|
| 2705 |
+
"logps/rejected": -366.0,
|
| 2706 |
+
"loss": 0.0138,
|
| 2707 |
+
"rewards/accuracies": 1.0,
|
| 2708 |
+
"rewards/chosen": 3.015625,
|
| 2709 |
+
"rewards/margins": 8.75,
|
| 2710 |
+
"rewards/rejected": -5.71875,
|
| 2711 |
+
"step": 1770
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 2.8525641025641026,
|
| 2715 |
+
"grad_norm": 1.3700769518291278,
|
| 2716 |
+
"learning_rate": 2.7315914489311164e-08,
|
| 2717 |
+
"logits/chosen": -0.177734375,
|
| 2718 |
+
"logits/rejected": -0.29296875,
|
| 2719 |
+
"logps/chosen": -354.0,
|
| 2720 |
+
"logps/rejected": -366.0,
|
| 2721 |
+
"loss": 0.0047,
|
| 2722 |
+
"rewards/accuracies": 1.0,
|
| 2723 |
+
"rewards/chosen": 2.1875,
|
| 2724 |
+
"rewards/margins": 8.375,
|
| 2725 |
+
"rewards/rejected": -6.15625,
|
| 2726 |
+
"step": 1780
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 2.8685897435897436,
|
| 2730 |
+
"grad_norm": 2.5772413830882437,
|
| 2731 |
+
"learning_rate": 2.4346793349168646e-08,
|
| 2732 |
+
"logits/chosen": -0.3515625,
|
| 2733 |
+
"logits/rejected": -0.40625,
|
| 2734 |
+
"logps/chosen": -356.0,
|
| 2735 |
+
"logps/rejected": -380.0,
|
| 2736 |
+
"loss": 0.0071,
|
| 2737 |
+
"rewards/accuracies": 1.0,
|
| 2738 |
+
"rewards/chosen": 2.265625,
|
| 2739 |
+
"rewards/margins": 8.25,
|
| 2740 |
+
"rewards/rejected": -5.96875,
|
| 2741 |
+
"step": 1790
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 2.8846153846153846,
|
| 2745 |
+
"grad_norm": 0.2522421688634616,
|
| 2746 |
+
"learning_rate": 2.1377672209026125e-08,
|
| 2747 |
+
"logits/chosen": -0.1884765625,
|
| 2748 |
+
"logits/rejected": -0.296875,
|
| 2749 |
+
"logps/chosen": -354.0,
|
| 2750 |
+
"logps/rejected": -362.0,
|
| 2751 |
+
"loss": 0.0155,
|
| 2752 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2753 |
+
"rewards/chosen": 2.609375,
|
| 2754 |
+
"rewards/margins": 8.625,
|
| 2755 |
+
"rewards/rejected": -5.96875,
|
| 2756 |
+
"step": 1800
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 2.9006410256410255,
|
| 2760 |
+
"grad_norm": 2.0378946111088347,
|
| 2761 |
+
"learning_rate": 1.840855106888361e-08,
|
| 2762 |
+
"logits/chosen": -0.298828125,
|
| 2763 |
+
"logits/rejected": -0.4765625,
|
| 2764 |
+
"logps/chosen": -346.0,
|
| 2765 |
+
"logps/rejected": -370.0,
|
| 2766 |
+
"loss": 0.0427,
|
| 2767 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2768 |
+
"rewards/chosen": 1.828125,
|
| 2769 |
+
"rewards/margins": 7.90625,
|
| 2770 |
+
"rewards/rejected": -6.0625,
|
| 2771 |
+
"step": 1810
|
| 2772 |
+
},
|
| 2773 |
+
{
|
| 2774 |
+
"epoch": 2.9166666666666665,
|
| 2775 |
+
"grad_norm": 2.2386092025677784,
|
| 2776 |
+
"learning_rate": 1.5439429928741092e-08,
|
| 2777 |
+
"logits/chosen": -0.21875,
|
| 2778 |
+
"logits/rejected": -0.2138671875,
|
| 2779 |
+
"logps/chosen": -380.0,
|
| 2780 |
+
"logps/rejected": -372.0,
|
| 2781 |
+
"loss": 0.0179,
|
| 2782 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2783 |
+
"rewards/chosen": 1.84375,
|
| 2784 |
+
"rewards/margins": 8.5,
|
| 2785 |
+
"rewards/rejected": -6.65625,
|
| 2786 |
+
"step": 1820
|
| 2787 |
+
},
|
| 2788 |
+
{
|
| 2789 |
+
"epoch": 2.9326923076923075,
|
| 2790 |
+
"grad_norm": 5.606669828245322,
|
| 2791 |
+
"learning_rate": 1.2470308788598574e-08,
|
| 2792 |
+
"logits/chosen": -0.26953125,
|
| 2793 |
+
"logits/rejected": -0.224609375,
|
| 2794 |
+
"logps/chosen": -364.0,
|
| 2795 |
+
"logps/rejected": -346.0,
|
| 2796 |
+
"loss": 0.0302,
|
| 2797 |
+
"rewards/accuracies": 1.0,
|
| 2798 |
+
"rewards/chosen": 2.5625,
|
| 2799 |
+
"rewards/margins": 8.5625,
|
| 2800 |
+
"rewards/rejected": -5.96875,
|
| 2801 |
+
"step": 1830
|
| 2802 |
+
},
|
| 2803 |
+
{
|
| 2804 |
+
"epoch": 2.948717948717949,
|
| 2805 |
+
"grad_norm": 7.459897331240726,
|
| 2806 |
+
"learning_rate": 9.501187648456057e-09,
|
| 2807 |
+
"logits/chosen": -0.166015625,
|
| 2808 |
+
"logits/rejected": -0.27734375,
|
| 2809 |
+
"logps/chosen": -338.0,
|
| 2810 |
+
"logps/rejected": -372.0,
|
| 2811 |
+
"loss": 0.0069,
|
| 2812 |
+
"rewards/accuracies": 1.0,
|
| 2813 |
+
"rewards/chosen": 1.9453125,
|
| 2814 |
+
"rewards/margins": 8.375,
|
| 2815 |
+
"rewards/rejected": -6.40625,
|
| 2816 |
+
"step": 1840
|
| 2817 |
+
},
|
| 2818 |
+
{
|
| 2819 |
+
"epoch": 2.96474358974359,
|
| 2820 |
+
"grad_norm": 46.309709716339626,
|
| 2821 |
+
"learning_rate": 6.532066508313539e-09,
|
| 2822 |
+
"logits/chosen": -0.265625,
|
| 2823 |
+
"logits/rejected": -0.3046875,
|
| 2824 |
+
"logps/chosen": -368.0,
|
| 2825 |
+
"logps/rejected": -392.0,
|
| 2826 |
+
"loss": 0.0144,
|
| 2827 |
+
"rewards/accuracies": 1.0,
|
| 2828 |
+
"rewards/chosen": 2.125,
|
| 2829 |
+
"rewards/margins": 8.5,
|
| 2830 |
+
"rewards/rejected": -6.375,
|
| 2831 |
+
"step": 1850
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 2.980769230769231,
|
| 2835 |
+
"grad_norm": 2.757504775523463,
|
| 2836 |
+
"learning_rate": 3.562945368171021e-09,
|
| 2837 |
+
"logits/chosen": -0.2314453125,
|
| 2838 |
+
"logits/rejected": -0.3515625,
|
| 2839 |
+
"logps/chosen": -366.0,
|
| 2840 |
+
"logps/rejected": -382.0,
|
| 2841 |
+
"loss": 0.0081,
|
| 2842 |
+
"rewards/accuracies": 1.0,
|
| 2843 |
+
"rewards/chosen": 2.3125,
|
| 2844 |
+
"rewards/margins": 8.125,
|
| 2845 |
+
"rewards/rejected": -5.84375,
|
| 2846 |
+
"step": 1860
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 2.996794871794872,
|
| 2850 |
+
"grad_norm": 0.7690916064350508,
|
| 2851 |
+
"learning_rate": 5.938242280285036e-10,
|
| 2852 |
+
"logits/chosen": -0.146484375,
|
| 2853 |
+
"logits/rejected": -0.361328125,
|
| 2854 |
+
"logps/chosen": -352.0,
|
| 2855 |
+
"logps/rejected": -378.0,
|
| 2856 |
+
"loss": 0.0214,
|
| 2857 |
+
"rewards/accuracies": 1.0,
|
| 2858 |
+
"rewards/chosen": 2.9375,
|
| 2859 |
+
"rewards/margins": 8.75,
|
| 2860 |
+
"rewards/rejected": -5.84375,
|
| 2861 |
+
"step": 1870
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 3.0,
|
| 2865 |
+
"eval_logits/chosen": -0.2392578125,
|
| 2866 |
+
"eval_logits/rejected": -0.2734375,
|
| 2867 |
+
"eval_logps/chosen": -364.0,
|
| 2868 |
+
"eval_logps/rejected": -370.0,
|
| 2869 |
+
"eval_loss": 0.21664032340049744,
|
| 2870 |
+
"eval_rewards/accuracies": 0.9134615659713745,
|
| 2871 |
+
"eval_rewards/chosen": 1.7734375,
|
| 2872 |
+
"eval_rewards/margins": 6.65625,
|
| 2873 |
+
"eval_rewards/rejected": -4.875,
|
| 2874 |
+
"eval_runtime": 26.3085,
|
| 2875 |
+
"eval_samples_per_second": 7.602,
|
| 2876 |
+
"eval_steps_per_second": 0.494,
|
| 2877 |
+
"step": 1872
|
| 2878 |
}
|
| 2879 |
],
|
| 2880 |
"logging_steps": 10,
|
|
|
|
| 2889 |
"should_evaluate": false,
|
| 2890 |
"should_log": false,
|
| 2891 |
"should_save": true,
|
| 2892 |
+
"should_training_stop": true
|
| 2893 |
},
|
| 2894 |
"attributes": {}
|
| 2895 |
}
|