Update latest checkpoint for run 15940d1d-7d8c-4253-8810-2ea934bedee4 at step 500
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1303 -3
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 29529752
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:faa071c041589c9409cf0d5f6467276c758e59b808318330624b57f8742527f0
|
| 3 |
size 29529752
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 59228491
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b522da25da05fa57b972b261d5cb773382c5262d2408a209bb9e0e645c5dd172
|
| 3 |
size 59228491
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff241110fffc6d7f1dfe5219180751eb7c13fda2f32cae1897e796d0d82dcadb
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0217f1177dc7a87c11647d1abc84763fe1e837429688a9d0244781cf2d1571c1
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -11708,11 +11708,1311 @@
|
|
| 11708 |
"rewards/reward_func/mean": 0.48919999599456787,
|
| 11709 |
"rewards/reward_func/std": 0.522976815700531,
|
| 11710 |
"step": 450
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11711 |
}
|
| 11712 |
],
|
| 11713 |
"logging_steps": 1,
|
| 11714 |
"max_steps": 950,
|
| 11715 |
-
"num_input_tokens_seen":
|
| 11716 |
"num_train_epochs": 2,
|
| 11717 |
"save_steps": 50,
|
| 11718 |
"stateful_callbacks": {
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9765625,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 11708 |
"rewards/reward_func/mean": 0.48919999599456787,
|
| 11709 |
"rewards/reward_func/std": 0.522976815700531,
|
| 11710 |
"step": 450
|
| 11711 |
+
},
|
| 11712 |
+
{
|
| 11713 |
+
"clip_ratio/high_max": 0.0,
|
| 11714 |
+
"clip_ratio/high_mean": 0.0,
|
| 11715 |
+
"clip_ratio/low_mean": 0.0,
|
| 11716 |
+
"clip_ratio/low_min": 0.0,
|
| 11717 |
+
"clip_ratio/region_mean": 0.0,
|
| 11718 |
+
"completions/clipped_ratio": 1.0,
|
| 11719 |
+
"completions/max_length": 384.0,
|
| 11720 |
+
"completions/max_terminated_length": 0.0,
|
| 11721 |
+
"completions/mean_length": 384.0,
|
| 11722 |
+
"completions/mean_terminated_length": 0.0,
|
| 11723 |
+
"completions/min_length": 384.0,
|
| 11724 |
+
"completions/min_terminated_length": 0.0,
|
| 11725 |
+
"entropy": 0.6474330313503742,
|
| 11726 |
+
"epoch": 0.880859375,
|
| 11727 |
+
"frac_reward_zero_std": 0.5,
|
| 11728 |
+
"grad_norm": 0.13752923905849457,
|
| 11729 |
+
"learning_rate": 2.631578947368421e-06,
|
| 11730 |
+
"loss": 0.0,
|
| 11731 |
+
"num_tokens": 2121664.0,
|
| 11732 |
+
"reward": 0.875,
|
| 11733 |
+
"reward_std": 0.25,
|
| 11734 |
+
"rewards/reward_func/mean": 0.875,
|
| 11735 |
+
"rewards/reward_func/std": 0.3535533845424652,
|
| 11736 |
+
"step": 451
|
| 11737 |
+
},
|
| 11738 |
+
{
|
| 11739 |
+
"clip_ratio/high_max": 0.0,
|
| 11740 |
+
"clip_ratio/high_mean": 0.0,
|
| 11741 |
+
"clip_ratio/low_mean": 0.0,
|
| 11742 |
+
"clip_ratio/low_min": 0.0,
|
| 11743 |
+
"clip_ratio/region_mean": 0.0,
|
| 11744 |
+
"completions/clipped_ratio": 1.0,
|
| 11745 |
+
"completions/max_length": 384.0,
|
| 11746 |
+
"completions/max_terminated_length": 0.0,
|
| 11747 |
+
"completions/mean_length": 384.0,
|
| 11748 |
+
"completions/mean_terminated_length": 0.0,
|
| 11749 |
+
"completions/min_length": 384.0,
|
| 11750 |
+
"completions/min_terminated_length": 0.0,
|
| 11751 |
+
"entropy": 0.47194433584809303,
|
| 11752 |
+
"epoch": 0.8828125,
|
| 11753 |
+
"frac_reward_zero_std": 0.0,
|
| 11754 |
+
"grad_norm": 0.22527579963207245,
|
| 11755 |
+
"learning_rate": 2.6263157894736847e-06,
|
| 11756 |
+
"loss": 0.0,
|
| 11757 |
+
"num_tokens": 2126328.0,
|
| 11758 |
+
"reward": 0.5039750337600708,
|
| 11759 |
+
"reward_std": 0.560730516910553,
|
| 11760 |
+
"rewards/reward_func/mean": 0.5039750337600708,
|
| 11761 |
+
"rewards/reward_func/std": 0.5192855596542358,
|
| 11762 |
+
"step": 452
|
| 11763 |
+
},
|
| 11764 |
+
{
|
| 11765 |
+
"clip_ratio/high_max": 0.0,
|
| 11766 |
+
"clip_ratio/high_mean": 0.0,
|
| 11767 |
+
"clip_ratio/low_mean": 0.0,
|
| 11768 |
+
"clip_ratio/low_min": 0.0,
|
| 11769 |
+
"clip_ratio/region_mean": 0.0,
|
| 11770 |
+
"completions/clipped_ratio": 1.0,
|
| 11771 |
+
"completions/max_length": 384.0,
|
| 11772 |
+
"completions/max_terminated_length": 0.0,
|
| 11773 |
+
"completions/mean_length": 384.0,
|
| 11774 |
+
"completions/mean_terminated_length": 0.0,
|
| 11775 |
+
"completions/min_length": 384.0,
|
| 11776 |
+
"completions/min_terminated_length": 0.0,
|
| 11777 |
+
"entropy": 0.696033101528883,
|
| 11778 |
+
"epoch": 0.884765625,
|
| 11779 |
+
"frac_reward_zero_std": 0.0,
|
| 11780 |
+
"grad_norm": 0.21090690791606903,
|
| 11781 |
+
"learning_rate": 2.6210526315789474e-06,
|
| 11782 |
+
"loss": -0.0,
|
| 11783 |
+
"num_tokens": 2130864.0,
|
| 11784 |
+
"reward": 0.7397750020027161,
|
| 11785 |
+
"reward_std": 0.28458285331726074,
|
| 11786 |
+
"rewards/reward_func/mean": 0.7397749423980713,
|
| 11787 |
+
"rewards/reward_func/std": 0.4491344392299652,
|
| 11788 |
+
"step": 453
|
| 11789 |
+
},
|
| 11790 |
+
{
|
| 11791 |
+
"clip_ratio/high_max": 0.0,
|
| 11792 |
+
"clip_ratio/high_mean": 0.0,
|
| 11793 |
+
"clip_ratio/low_mean": 0.0,
|
| 11794 |
+
"clip_ratio/low_min": 0.0,
|
| 11795 |
+
"clip_ratio/region_mean": 0.0,
|
| 11796 |
+
"completions/clipped_ratio": 1.0,
|
| 11797 |
+
"completions/max_length": 384.0,
|
| 11798 |
+
"completions/max_terminated_length": 0.0,
|
| 11799 |
+
"completions/mean_length": 384.0,
|
| 11800 |
+
"completions/mean_terminated_length": 0.0,
|
| 11801 |
+
"completions/min_length": 384.0,
|
| 11802 |
+
"completions/min_terminated_length": 0.0,
|
| 11803 |
+
"entropy": 0.5411256179213524,
|
| 11804 |
+
"epoch": 0.88671875,
|
| 11805 |
+
"frac_reward_zero_std": 0.0,
|
| 11806 |
+
"grad_norm": 0.18700988590717316,
|
| 11807 |
+
"learning_rate": 2.615789473684211e-06,
|
| 11808 |
+
"loss": 0.0,
|
| 11809 |
+
"num_tokens": 2135832.0,
|
| 11810 |
+
"reward": 0.3330000042915344,
|
| 11811 |
+
"reward_std": 0.47834351658821106,
|
| 11812 |
+
"rewards/reward_func/mean": 0.3330000042915344,
|
| 11813 |
+
"rewards/reward_func/std": 0.45958366990089417,
|
| 11814 |
+
"step": 454
|
| 11815 |
+
},
|
| 11816 |
+
{
|
| 11817 |
+
"clip_ratio/high_max": 0.0,
|
| 11818 |
+
"clip_ratio/high_mean": 0.0,
|
| 11819 |
+
"clip_ratio/low_mean": 0.0,
|
| 11820 |
+
"clip_ratio/low_min": 0.0,
|
| 11821 |
+
"clip_ratio/region_mean": 0.0,
|
| 11822 |
+
"completions/clipped_ratio": 1.0,
|
| 11823 |
+
"completions/max_length": 384.0,
|
| 11824 |
+
"completions/max_terminated_length": 0.0,
|
| 11825 |
+
"completions/mean_length": 384.0,
|
| 11826 |
+
"completions/mean_terminated_length": 0.0,
|
| 11827 |
+
"completions/min_length": 384.0,
|
| 11828 |
+
"completions/min_terminated_length": 0.0,
|
| 11829 |
+
"entropy": 0.602843264117837,
|
| 11830 |
+
"epoch": 0.888671875,
|
| 11831 |
+
"frac_reward_zero_std": 0.0,
|
| 11832 |
+
"grad_norm": 0.19545380771160126,
|
| 11833 |
+
"learning_rate": 2.6105263157894738e-06,
|
| 11834 |
+
"loss": -0.0,
|
| 11835 |
+
"num_tokens": 2140624.0,
|
| 11836 |
+
"reward": 0.24056249856948853,
|
| 11837 |
+
"reward_std": 0.47697657346725464,
|
| 11838 |
+
"rewards/reward_func/mean": 0.24056249856948853,
|
| 11839 |
+
"rewards/reward_func/std": 0.44159868359565735,
|
| 11840 |
+
"step": 455
|
| 11841 |
+
},
|
| 11842 |
+
{
|
| 11843 |
+
"clip_ratio/high_max": 0.0,
|
| 11844 |
+
"clip_ratio/high_mean": 0.0,
|
| 11845 |
+
"clip_ratio/low_mean": 0.0,
|
| 11846 |
+
"clip_ratio/low_min": 0.0,
|
| 11847 |
+
"clip_ratio/region_mean": 0.0,
|
| 11848 |
+
"completions/clipped_ratio": 1.0,
|
| 11849 |
+
"completions/max_length": 384.0,
|
| 11850 |
+
"completions/max_terminated_length": 0.0,
|
| 11851 |
+
"completions/mean_length": 384.0,
|
| 11852 |
+
"completions/mean_terminated_length": 0.0,
|
| 11853 |
+
"completions/min_length": 384.0,
|
| 11854 |
+
"completions/min_terminated_length": 0.0,
|
| 11855 |
+
"entropy": 0.7841911315917969,
|
| 11856 |
+
"epoch": 0.890625,
|
| 11857 |
+
"frac_reward_zero_std": 0.5,
|
| 11858 |
+
"grad_norm": 0.16834405064582825,
|
| 11859 |
+
"learning_rate": 2.605263157894737e-06,
|
| 11860 |
+
"loss": 0.0,
|
| 11861 |
+
"num_tokens": 2145096.0,
|
| 11862 |
+
"reward": 0.7562500238418579,
|
| 11863 |
+
"reward_std": 0.2814582586288452,
|
| 11864 |
+
"rewards/reward_func/mean": 0.7562500238418579,
|
| 11865 |
+
"rewards/reward_func/std": 0.4513373076915741,
|
| 11866 |
+
"step": 456
|
| 11867 |
+
},
|
| 11868 |
+
{
|
| 11869 |
+
"clip_ratio/high_max": 0.0,
|
| 11870 |
+
"clip_ratio/high_mean": 0.0,
|
| 11871 |
+
"clip_ratio/low_mean": 0.0,
|
| 11872 |
+
"clip_ratio/low_min": 0.0,
|
| 11873 |
+
"clip_ratio/region_mean": 0.0,
|
| 11874 |
+
"completions/clipped_ratio": 1.0,
|
| 11875 |
+
"completions/max_length": 384.0,
|
| 11876 |
+
"completions/max_terminated_length": 0.0,
|
| 11877 |
+
"completions/mean_length": 384.0,
|
| 11878 |
+
"completions/mean_terminated_length": 0.0,
|
| 11879 |
+
"completions/min_length": 384.0,
|
| 11880 |
+
"completions/min_terminated_length": 0.0,
|
| 11881 |
+
"entropy": 0.4610623624175787,
|
| 11882 |
+
"epoch": 0.892578125,
|
| 11883 |
+
"frac_reward_zero_std": 0.0,
|
| 11884 |
+
"grad_norm": 0.22504417598247528,
|
| 11885 |
+
"learning_rate": 2.6e-06,
|
| 11886 |
+
"loss": 0.0,
|
| 11887 |
+
"num_tokens": 2149848.0,
|
| 11888 |
+
"reward": 0.7124999761581421,
|
| 11889 |
+
"reward_std": 0.47541630268096924,
|
| 11890 |
+
"rewards/reward_func/mean": 0.7125000357627869,
|
| 11891 |
+
"rewards/reward_func/std": 0.44025152921676636,
|
| 11892 |
+
"step": 457
|
| 11893 |
+
},
|
| 11894 |
+
{
|
| 11895 |
+
"clip_ratio/high_max": 0.0,
|
| 11896 |
+
"clip_ratio/high_mean": 0.0,
|
| 11897 |
+
"clip_ratio/low_mean": 0.0,
|
| 11898 |
+
"clip_ratio/low_min": 0.0,
|
| 11899 |
+
"clip_ratio/region_mean": 0.0,
|
| 11900 |
+
"completions/clipped_ratio": 1.0,
|
| 11901 |
+
"completions/max_length": 384.0,
|
| 11902 |
+
"completions/max_terminated_length": 0.0,
|
| 11903 |
+
"completions/mean_length": 384.0,
|
| 11904 |
+
"completions/mean_terminated_length": 0.0,
|
| 11905 |
+
"completions/min_length": 384.0,
|
| 11906 |
+
"completions/min_terminated_length": 0.0,
|
| 11907 |
+
"entropy": 0.5260163843631744,
|
| 11908 |
+
"epoch": 0.89453125,
|
| 11909 |
+
"frac_reward_zero_std": 0.0,
|
| 11910 |
+
"grad_norm": 0.2690121531486511,
|
| 11911 |
+
"learning_rate": 2.5947368421052633e-06,
|
| 11912 |
+
"loss": 0.0,
|
| 11913 |
+
"num_tokens": 2154696.0,
|
| 11914 |
+
"reward": 0.350600004196167,
|
| 11915 |
+
"reward_std": 0.4991450905799866,
|
| 11916 |
+
"rewards/reward_func/mean": 0.350600004196167,
|
| 11917 |
+
"rewards/reward_func/std": 0.48443493247032166,
|
| 11918 |
+
"step": 458
|
| 11919 |
+
},
|
| 11920 |
+
{
|
| 11921 |
+
"clip_ratio/high_max": 0.0,
|
| 11922 |
+
"clip_ratio/high_mean": 0.0,
|
| 11923 |
+
"clip_ratio/low_mean": 0.0,
|
| 11924 |
+
"clip_ratio/low_min": 0.0,
|
| 11925 |
+
"clip_ratio/region_mean": 0.0,
|
| 11926 |
+
"completions/clipped_ratio": 1.0,
|
| 11927 |
+
"completions/max_length": 384.0,
|
| 11928 |
+
"completions/max_terminated_length": 0.0,
|
| 11929 |
+
"completions/mean_length": 384.0,
|
| 11930 |
+
"completions/mean_terminated_length": 0.0,
|
| 11931 |
+
"completions/min_length": 384.0,
|
| 11932 |
+
"completions/min_terminated_length": 0.0,
|
| 11933 |
+
"entropy": 0.8946198597550392,
|
| 11934 |
+
"epoch": 0.896484375,
|
| 11935 |
+
"frac_reward_zero_std": 0.0,
|
| 11936 |
+
"grad_norm": 0.25450319051742554,
|
| 11937 |
+
"learning_rate": 2.589473684210527e-06,
|
| 11938 |
+
"loss": 0.0,
|
| 11939 |
+
"num_tokens": 2159176.0,
|
| 11940 |
+
"reward": 0.9945999979972839,
|
| 11941 |
+
"reward_std": 0.010800004005432129,
|
| 11942 |
+
"rewards/reward_func/mean": 0.9945999979972839,
|
| 11943 |
+
"rewards/reward_func/std": 0.009998860768973827,
|
| 11944 |
+
"step": 459
|
| 11945 |
+
},
|
| 11946 |
+
{
|
| 11947 |
+
"clip_ratio/high_max": 0.0,
|
| 11948 |
+
"clip_ratio/high_mean": 0.0,
|
| 11949 |
+
"clip_ratio/low_mean": 0.0,
|
| 11950 |
+
"clip_ratio/low_min": 0.0,
|
| 11951 |
+
"clip_ratio/region_mean": 0.0,
|
| 11952 |
+
"completions/clipped_ratio": 1.0,
|
| 11953 |
+
"completions/max_length": 384.0,
|
| 11954 |
+
"completions/max_terminated_length": 0.0,
|
| 11955 |
+
"completions/mean_length": 384.0,
|
| 11956 |
+
"completions/mean_terminated_length": 0.0,
|
| 11957 |
+
"completions/min_length": 384.0,
|
| 11958 |
+
"completions/min_terminated_length": 0.0,
|
| 11959 |
+
"entropy": 0.3343417225405574,
|
| 11960 |
+
"epoch": 0.8984375,
|
| 11961 |
+
"frac_reward_zero_std": 0.5,
|
| 11962 |
+
"grad_norm": 0.23701933026313782,
|
| 11963 |
+
"learning_rate": 2.5842105263157896e-06,
|
| 11964 |
+
"loss": -0.0,
|
| 11965 |
+
"num_tokens": 2163664.0,
|
| 11966 |
+
"reward": 0.8134000301361084,
|
| 11967 |
+
"reward_std": 0.23240000009536743,
|
| 11968 |
+
"rewards/reward_func/mean": 0.8134000301361084,
|
| 11969 |
+
"rewards/reward_func/std": 0.3286632299423218,
|
| 11970 |
+
"step": 460
|
| 11971 |
+
},
|
| 11972 |
+
{
|
| 11973 |
+
"clip_ratio/high_max": 0.0,
|
| 11974 |
+
"clip_ratio/high_mean": 0.0,
|
| 11975 |
+
"clip_ratio/low_mean": 0.0,
|
| 11976 |
+
"clip_ratio/low_min": 0.0,
|
| 11977 |
+
"clip_ratio/region_mean": 0.0,
|
| 11978 |
+
"completions/clipped_ratio": 1.0,
|
| 11979 |
+
"completions/max_length": 384.0,
|
| 11980 |
+
"completions/max_terminated_length": 0.0,
|
| 11981 |
+
"completions/mean_length": 384.0,
|
| 11982 |
+
"completions/mean_terminated_length": 0.0,
|
| 11983 |
+
"completions/min_length": 384.0,
|
| 11984 |
+
"completions/min_terminated_length": 0.0,
|
| 11985 |
+
"entropy": 0.5373089350759983,
|
| 11986 |
+
"epoch": 0.900390625,
|
| 11987 |
+
"frac_reward_zero_std": 0.5,
|
| 11988 |
+
"grad_norm": 0.14844948053359985,
|
| 11989 |
+
"learning_rate": 2.578947368421053e-06,
|
| 11990 |
+
"loss": 0.0,
|
| 11991 |
+
"num_tokens": 2168448.0,
|
| 11992 |
+
"reward": 0.23899999260902405,
|
| 11993 |
+
"reward_std": 0.27597343921661377,
|
| 11994 |
+
"rewards/reward_func/mean": 0.23899999260902405,
|
| 11995 |
+
"rewards/reward_func/std": 0.44254201650619507,
|
| 11996 |
+
"step": 461
|
| 11997 |
+
},
|
| 11998 |
+
{
|
| 11999 |
+
"clip_ratio/high_max": 0.0,
|
| 12000 |
+
"clip_ratio/high_mean": 0.0,
|
| 12001 |
+
"clip_ratio/low_mean": 0.0,
|
| 12002 |
+
"clip_ratio/low_min": 0.0,
|
| 12003 |
+
"clip_ratio/region_mean": 0.0,
|
| 12004 |
+
"completions/clipped_ratio": 1.0,
|
| 12005 |
+
"completions/max_length": 384.0,
|
| 12006 |
+
"completions/max_terminated_length": 0.0,
|
| 12007 |
+
"completions/mean_length": 384.0,
|
| 12008 |
+
"completions/mean_terminated_length": 0.0,
|
| 12009 |
+
"completions/min_length": 384.0,
|
| 12010 |
+
"completions/min_terminated_length": 0.0,
|
| 12011 |
+
"entropy": 0.6289132609963417,
|
| 12012 |
+
"epoch": 0.90234375,
|
| 12013 |
+
"frac_reward_zero_std": 0.0,
|
| 12014 |
+
"grad_norm": 0.2263883650302887,
|
| 12015 |
+
"learning_rate": 2.573684210526316e-06,
|
| 12016 |
+
"loss": 0.0,
|
| 12017 |
+
"num_tokens": 2173248.0,
|
| 12018 |
+
"reward": 0.6249375343322754,
|
| 12019 |
+
"reward_std": 0.5195988416671753,
|
| 12020 |
+
"rewards/reward_func/mean": 0.6249375343322754,
|
| 12021 |
+
"rewards/reward_func/std": 0.4944932758808136,
|
| 12022 |
+
"step": 462
|
| 12023 |
+
},
|
| 12024 |
+
{
|
| 12025 |
+
"clip_ratio/high_max": 0.0,
|
| 12026 |
+
"clip_ratio/high_mean": 0.0,
|
| 12027 |
+
"clip_ratio/low_mean": 0.0,
|
| 12028 |
+
"clip_ratio/low_min": 0.0,
|
| 12029 |
+
"clip_ratio/region_mean": 0.0,
|
| 12030 |
+
"completions/clipped_ratio": 1.0,
|
| 12031 |
+
"completions/max_length": 384.0,
|
| 12032 |
+
"completions/max_terminated_length": 0.0,
|
| 12033 |
+
"completions/mean_length": 384.0,
|
| 12034 |
+
"completions/mean_terminated_length": 0.0,
|
| 12035 |
+
"completions/min_length": 384.0,
|
| 12036 |
+
"completions/min_terminated_length": 0.0,
|
| 12037 |
+
"entropy": 0.9023691639304161,
|
| 12038 |
+
"epoch": 0.904296875,
|
| 12039 |
+
"frac_reward_zero_std": 0.0,
|
| 12040 |
+
"grad_norm": 0.3016529679298401,
|
| 12041 |
+
"learning_rate": 2.568421052631579e-06,
|
| 12042 |
+
"loss": 0.0,
|
| 12043 |
+
"num_tokens": 2177720.0,
|
| 12044 |
+
"reward": 0.6285499930381775,
|
| 12045 |
+
"reward_std": 0.25130394101142883,
|
| 12046 |
+
"rewards/reward_func/mean": 0.6285499930381775,
|
| 12047 |
+
"rewards/reward_func/std": 0.5067988634109497,
|
| 12048 |
+
"step": 463
|
| 12049 |
+
},
|
| 12050 |
+
{
|
| 12051 |
+
"clip_ratio/high_max": 0.0,
|
| 12052 |
+
"clip_ratio/high_mean": 0.0,
|
| 12053 |
+
"clip_ratio/low_mean": 0.0,
|
| 12054 |
+
"clip_ratio/low_min": 0.0,
|
| 12055 |
+
"clip_ratio/region_mean": 0.0,
|
| 12056 |
+
"completions/clipped_ratio": 1.0,
|
| 12057 |
+
"completions/max_length": 384.0,
|
| 12058 |
+
"completions/max_terminated_length": 0.0,
|
| 12059 |
+
"completions/mean_length": 384.0,
|
| 12060 |
+
"completions/mean_terminated_length": 0.0,
|
| 12061 |
+
"completions/min_length": 384.0,
|
| 12062 |
+
"completions/min_terminated_length": 0.0,
|
| 12063 |
+
"entropy": 0.5828330963850021,
|
| 12064 |
+
"epoch": 0.90625,
|
| 12065 |
+
"frac_reward_zero_std": 0.0,
|
| 12066 |
+
"grad_norm": 0.20633725821971893,
|
| 12067 |
+
"learning_rate": 2.5631578947368422e-06,
|
| 12068 |
+
"loss": 0.0,
|
| 12069 |
+
"num_tokens": 2182560.0,
|
| 12070 |
+
"reward": 0.6124000549316406,
|
| 12071 |
+
"reward_std": 0.48071789741516113,
|
| 12072 |
+
"rewards/reward_func/mean": 0.6124000549316406,
|
| 12073 |
+
"rewards/reward_func/std": 0.46755528450012207,
|
| 12074 |
+
"step": 464
|
| 12075 |
+
},
|
| 12076 |
+
{
|
| 12077 |
+
"clip_ratio/high_max": 0.0,
|
| 12078 |
+
"clip_ratio/high_mean": 0.0,
|
| 12079 |
+
"clip_ratio/low_mean": 0.0,
|
| 12080 |
+
"clip_ratio/low_min": 0.0,
|
| 12081 |
+
"clip_ratio/region_mean": 0.0,
|
| 12082 |
+
"completions/clipped_ratio": 1.0,
|
| 12083 |
+
"completions/max_length": 384.0,
|
| 12084 |
+
"completions/max_terminated_length": 0.0,
|
| 12085 |
+
"completions/mean_length": 384.0,
|
| 12086 |
+
"completions/mean_terminated_length": 0.0,
|
| 12087 |
+
"completions/min_length": 384.0,
|
| 12088 |
+
"completions/min_terminated_length": 0.0,
|
| 12089 |
+
"entropy": 0.421830615028739,
|
| 12090 |
+
"epoch": 0.908203125,
|
| 12091 |
+
"frac_reward_zero_std": 0.0,
|
| 12092 |
+
"grad_norm": 0.1946926712989807,
|
| 12093 |
+
"learning_rate": 2.5578947368421054e-06,
|
| 12094 |
+
"loss": -0.0,
|
| 12095 |
+
"num_tokens": 2187512.0,
|
| 12096 |
+
"reward": 0.24700000882148743,
|
| 12097 |
+
"reward_std": 0.2572190761566162,
|
| 12098 |
+
"rewards/reward_func/mean": 0.24700000882148743,
|
| 12099 |
+
"rewards/reward_func/std": 0.39629149436950684,
|
| 12100 |
+
"step": 465
|
| 12101 |
+
},
|
| 12102 |
+
{
|
| 12103 |
+
"clip_ratio/high_max": 0.0,
|
| 12104 |
+
"clip_ratio/high_mean": 0.0,
|
| 12105 |
+
"clip_ratio/low_mean": 0.0,
|
| 12106 |
+
"clip_ratio/low_min": 0.0,
|
| 12107 |
+
"clip_ratio/region_mean": 0.0,
|
| 12108 |
+
"completions/clipped_ratio": 1.0,
|
| 12109 |
+
"completions/max_length": 384.0,
|
| 12110 |
+
"completions/max_terminated_length": 0.0,
|
| 12111 |
+
"completions/mean_length": 384.0,
|
| 12112 |
+
"completions/mean_terminated_length": 0.0,
|
| 12113 |
+
"completions/min_length": 384.0,
|
| 12114 |
+
"completions/min_terminated_length": 0.0,
|
| 12115 |
+
"entropy": 0.38339319732040167,
|
| 12116 |
+
"epoch": 0.91015625,
|
| 12117 |
+
"frac_reward_zero_std": 0.5,
|
| 12118 |
+
"grad_norm": 0.11662346869707108,
|
| 12119 |
+
"learning_rate": 2.552631578947369e-06,
|
| 12120 |
+
"loss": 0.0,
|
| 12121 |
+
"num_tokens": 2192024.0,
|
| 12122 |
+
"reward": 0.8113000392913818,
|
| 12123 |
+
"reward_std": 0.23103395104408264,
|
| 12124 |
+
"rewards/reward_func/mean": 0.8113000392913818,
|
| 12125 |
+
"rewards/reward_func/std": 0.32786741852760315,
|
| 12126 |
+
"step": 466
|
| 12127 |
+
},
|
| 12128 |
+
{
|
| 12129 |
+
"clip_ratio/high_max": 0.0,
|
| 12130 |
+
"clip_ratio/high_mean": 0.0,
|
| 12131 |
+
"clip_ratio/low_mean": 0.0,
|
| 12132 |
+
"clip_ratio/low_min": 0.0,
|
| 12133 |
+
"clip_ratio/region_mean": 0.0,
|
| 12134 |
+
"completions/clipped_ratio": 1.0,
|
| 12135 |
+
"completions/max_length": 384.0,
|
| 12136 |
+
"completions/max_terminated_length": 0.0,
|
| 12137 |
+
"completions/mean_length": 384.0,
|
| 12138 |
+
"completions/mean_terminated_length": 0.0,
|
| 12139 |
+
"completions/min_length": 384.0,
|
| 12140 |
+
"completions/min_terminated_length": 0.0,
|
| 12141 |
+
"entropy": 0.733765222132206,
|
| 12142 |
+
"epoch": 0.912109375,
|
| 12143 |
+
"frac_reward_zero_std": 0.5,
|
| 12144 |
+
"grad_norm": 0.13307461142539978,
|
| 12145 |
+
"learning_rate": 2.5473684210526317e-06,
|
| 12146 |
+
"loss": -0.0,
|
| 12147 |
+
"num_tokens": 2197040.0,
|
| 12148 |
+
"reward": 0.11410000175237656,
|
| 12149 |
+
"reward_std": 0.2282000035047531,
|
| 12150 |
+
"rewards/reward_func/mean": 0.11410000175237656,
|
| 12151 |
+
"rewards/reward_func/std": 0.3227235674858093,
|
| 12152 |
+
"step": 467
|
| 12153 |
+
},
|
| 12154 |
+
{
|
| 12155 |
+
"clip_ratio/high_max": 0.0,
|
| 12156 |
+
"clip_ratio/high_mean": 0.0,
|
| 12157 |
+
"clip_ratio/low_mean": 0.0,
|
| 12158 |
+
"clip_ratio/low_min": 0.0,
|
| 12159 |
+
"clip_ratio/region_mean": 0.0,
|
| 12160 |
+
"completions/clipped_ratio": 1.0,
|
| 12161 |
+
"completions/max_length": 384.0,
|
| 12162 |
+
"completions/max_terminated_length": 0.0,
|
| 12163 |
+
"completions/mean_length": 384.0,
|
| 12164 |
+
"completions/mean_terminated_length": 0.0,
|
| 12165 |
+
"completions/min_length": 384.0,
|
| 12166 |
+
"completions/min_terminated_length": 0.0,
|
| 12167 |
+
"entropy": 0.5777375996112823,
|
| 12168 |
+
"epoch": 0.9140625,
|
| 12169 |
+
"frac_reward_zero_std": 0.0,
|
| 12170 |
+
"grad_norm": 0.2451069951057434,
|
| 12171 |
+
"learning_rate": 2.5421052631578953e-06,
|
| 12172 |
+
"loss": 0.0,
|
| 12173 |
+
"num_tokens": 2201888.0,
|
| 12174 |
+
"reward": 0.58160001039505,
|
| 12175 |
+
"reward_std": 0.5065810680389404,
|
| 12176 |
+
"rewards/reward_func/mean": 0.58160001039505,
|
| 12177 |
+
"rewards/reward_func/std": 0.4829333424568176,
|
| 12178 |
+
"step": 468
|
| 12179 |
+
},
|
| 12180 |
+
{
|
| 12181 |
+
"clip_ratio/high_max": 0.0,
|
| 12182 |
+
"clip_ratio/high_mean": 0.0,
|
| 12183 |
+
"clip_ratio/low_mean": 0.0,
|
| 12184 |
+
"clip_ratio/low_min": 0.0,
|
| 12185 |
+
"clip_ratio/region_mean": 0.0,
|
| 12186 |
+
"completions/clipped_ratio": 1.0,
|
| 12187 |
+
"completions/max_length": 384.0,
|
| 12188 |
+
"completions/max_terminated_length": 0.0,
|
| 12189 |
+
"completions/mean_length": 384.0,
|
| 12190 |
+
"completions/mean_terminated_length": 0.0,
|
| 12191 |
+
"completions/min_length": 384.0,
|
| 12192 |
+
"completions/min_terminated_length": 0.0,
|
| 12193 |
+
"entropy": 0.5028033722192049,
|
| 12194 |
+
"epoch": 0.916015625,
|
| 12195 |
+
"frac_reward_zero_std": 0.0,
|
| 12196 |
+
"grad_norm": 0.23731045424938202,
|
| 12197 |
+
"learning_rate": 2.536842105263158e-06,
|
| 12198 |
+
"loss": 0.0,
|
| 12199 |
+
"num_tokens": 2206744.0,
|
| 12200 |
+
"reward": 0.7462999820709229,
|
| 12201 |
+
"reward_std": 0.4641999900341034,
|
| 12202 |
+
"rewards/reward_func/mean": 0.7462999820709229,
|
| 12203 |
+
"rewards/reward_func/std": 0.4297657012939453,
|
| 12204 |
+
"step": 469
|
| 12205 |
+
},
|
| 12206 |
+
{
|
| 12207 |
+
"clip_ratio/high_max": 0.0,
|
| 12208 |
+
"clip_ratio/high_mean": 0.0,
|
| 12209 |
+
"clip_ratio/low_mean": 0.0,
|
| 12210 |
+
"clip_ratio/low_min": 0.0,
|
| 12211 |
+
"clip_ratio/region_mean": 0.0,
|
| 12212 |
+
"completions/clipped_ratio": 1.0,
|
| 12213 |
+
"completions/max_length": 384.0,
|
| 12214 |
+
"completions/max_terminated_length": 0.0,
|
| 12215 |
+
"completions/mean_length": 384.0,
|
| 12216 |
+
"completions/mean_terminated_length": 0.0,
|
| 12217 |
+
"completions/min_length": 384.0,
|
| 12218 |
+
"completions/min_terminated_length": 0.0,
|
| 12219 |
+
"entropy": 0.49811795353889465,
|
| 12220 |
+
"epoch": 0.91796875,
|
| 12221 |
+
"frac_reward_zero_std": 0.0,
|
| 12222 |
+
"grad_norm": 0.20878010988235474,
|
| 12223 |
+
"learning_rate": 2.5315789473684212e-06,
|
| 12224 |
+
"loss": 0.0,
|
| 12225 |
+
"num_tokens": 2211416.0,
|
| 12226 |
+
"reward": 0.7425000071525574,
|
| 12227 |
+
"reward_std": 0.30367511510849,
|
| 12228 |
+
"rewards/reward_func/mean": 0.7425000071525574,
|
| 12229 |
+
"rewards/reward_func/std": 0.4587482810020447,
|
| 12230 |
+
"step": 470
|
| 12231 |
+
},
|
| 12232 |
+
{
|
| 12233 |
+
"clip_ratio/high_max": 0.0,
|
| 12234 |
+
"clip_ratio/high_mean": 0.0,
|
| 12235 |
+
"clip_ratio/low_mean": 0.0,
|
| 12236 |
+
"clip_ratio/low_min": 0.0,
|
| 12237 |
+
"clip_ratio/region_mean": 0.0,
|
| 12238 |
+
"completions/clipped_ratio": 1.0,
|
| 12239 |
+
"completions/max_length": 384.0,
|
| 12240 |
+
"completions/max_terminated_length": 0.0,
|
| 12241 |
+
"completions/mean_length": 384.0,
|
| 12242 |
+
"completions/mean_terminated_length": 0.0,
|
| 12243 |
+
"completions/min_length": 384.0,
|
| 12244 |
+
"completions/min_terminated_length": 0.0,
|
| 12245 |
+
"entropy": 0.36679044365882874,
|
| 12246 |
+
"epoch": 0.919921875,
|
| 12247 |
+
"frac_reward_zero_std": 0.5,
|
| 12248 |
+
"grad_norm": 0.12484736740589142,
|
| 12249 |
+
"learning_rate": 2.5263157894736844e-06,
|
| 12250 |
+
"loss": -0.0,
|
| 12251 |
+
"num_tokens": 2215912.0,
|
| 12252 |
+
"reward": 0.8134000301361084,
|
| 12253 |
+
"reward_std": 0.23240000009536743,
|
| 12254 |
+
"rewards/reward_func/mean": 0.8134000301361084,
|
| 12255 |
+
"rewards/reward_func/std": 0.3286632299423218,
|
| 12256 |
+
"step": 471
|
| 12257 |
+
},
|
| 12258 |
+
{
|
| 12259 |
+
"clip_ratio/high_max": 0.0,
|
| 12260 |
+
"clip_ratio/high_mean": 0.0,
|
| 12261 |
+
"clip_ratio/low_mean": 0.0,
|
| 12262 |
+
"clip_ratio/low_min": 0.0,
|
| 12263 |
+
"clip_ratio/region_mean": 0.0,
|
| 12264 |
+
"completions/clipped_ratio": 1.0,
|
| 12265 |
+
"completions/max_length": 384.0,
|
| 12266 |
+
"completions/max_terminated_length": 0.0,
|
| 12267 |
+
"completions/mean_length": 384.0,
|
| 12268 |
+
"completions/mean_terminated_length": 0.0,
|
| 12269 |
+
"completions/min_length": 384.0,
|
| 12270 |
+
"completions/min_terminated_length": 0.0,
|
| 12271 |
+
"entropy": 0.5753449611365795,
|
| 12272 |
+
"epoch": 0.921875,
|
| 12273 |
+
"frac_reward_zero_std": 0.0,
|
| 12274 |
+
"grad_norm": 0.23447397351264954,
|
| 12275 |
+
"learning_rate": 2.5210526315789475e-06,
|
| 12276 |
+
"loss": 0.0,
|
| 12277 |
+
"num_tokens": 2220568.0,
|
| 12278 |
+
"reward": 0.38792502880096436,
|
| 12279 |
+
"reward_std": 0.5182595252990723,
|
| 12280 |
+
"rewards/reward_func/mean": 0.38792499899864197,
|
| 12281 |
+
"rewards/reward_func/std": 0.5017743706703186,
|
| 12282 |
+
"step": 472
|
| 12283 |
+
},
|
| 12284 |
+
{
|
| 12285 |
+
"clip_ratio/high_max": 0.0,
|
| 12286 |
+
"clip_ratio/high_mean": 0.0,
|
| 12287 |
+
"clip_ratio/low_mean": 0.0,
|
| 12288 |
+
"clip_ratio/low_min": 0.0,
|
| 12289 |
+
"clip_ratio/region_mean": 0.0,
|
| 12290 |
+
"completions/clipped_ratio": 1.0,
|
| 12291 |
+
"completions/max_length": 384.0,
|
| 12292 |
+
"completions/max_terminated_length": 0.0,
|
| 12293 |
+
"completions/mean_length": 384.0,
|
| 12294 |
+
"completions/mean_terminated_length": 0.0,
|
| 12295 |
+
"completions/min_length": 384.0,
|
| 12296 |
+
"completions/min_terminated_length": 0.0,
|
| 12297 |
+
"entropy": 0.4622494773939252,
|
| 12298 |
+
"epoch": 0.923828125,
|
| 12299 |
+
"frac_reward_zero_std": 0.0,
|
| 12300 |
+
"grad_norm": 0.25002261996269226,
|
| 12301 |
+
"learning_rate": 2.5157894736842107e-06,
|
| 12302 |
+
"loss": 0.0,
|
| 12303 |
+
"num_tokens": 2225352.0,
|
| 12304 |
+
"reward": 0.249937504529953,
|
| 12305 |
+
"reward_std": 0.47128826379776,
|
| 12306 |
+
"rewards/reward_func/mean": 0.249937504529953,
|
| 12307 |
+
"rewards/reward_func/std": 0.4365290105342865,
|
| 12308 |
+
"step": 473
|
| 12309 |
+
},
|
| 12310 |
+
{
|
| 12311 |
+
"clip_ratio/high_max": 0.0,
|
| 12312 |
+
"clip_ratio/high_mean": 0.0,
|
| 12313 |
+
"clip_ratio/low_mean": 0.0,
|
| 12314 |
+
"clip_ratio/low_min": 0.0,
|
| 12315 |
+
"clip_ratio/region_mean": 0.0,
|
| 12316 |
+
"completions/clipped_ratio": 1.0,
|
| 12317 |
+
"completions/max_length": 384.0,
|
| 12318 |
+
"completions/max_terminated_length": 0.0,
|
| 12319 |
+
"completions/mean_length": 384.0,
|
| 12320 |
+
"completions/mean_terminated_length": 0.0,
|
| 12321 |
+
"completions/min_length": 384.0,
|
| 12322 |
+
"completions/min_terminated_length": 0.0,
|
| 12323 |
+
"entropy": 0.5469172280281782,
|
| 12324 |
+
"epoch": 0.92578125,
|
| 12325 |
+
"frac_reward_zero_std": 0.0,
|
| 12326 |
+
"grad_norm": 0.20600056648254395,
|
| 12327 |
+
"learning_rate": 2.510526315789474e-06,
|
| 12328 |
+
"loss": 0.0,
|
| 12329 |
+
"num_tokens": 2230336.0,
|
| 12330 |
+
"reward": 0.22462499141693115,
|
| 12331 |
+
"reward_std": 0.44099122285842896,
|
| 12332 |
+
"rewards/reward_func/mean": 0.22462499141693115,
|
| 12333 |
+
"rewards/reward_func/std": 0.4093196392059326,
|
| 12334 |
+
"step": 474
|
| 12335 |
+
},
|
| 12336 |
+
{
|
| 12337 |
+
"clip_ratio/high_max": 0.0,
|
| 12338 |
+
"clip_ratio/high_mean": 0.0,
|
| 12339 |
+
"clip_ratio/low_mean": 0.0,
|
| 12340 |
+
"clip_ratio/low_min": 0.0,
|
| 12341 |
+
"clip_ratio/region_mean": 0.0,
|
| 12342 |
+
"completions/clipped_ratio": 1.0,
|
| 12343 |
+
"completions/max_length": 384.0,
|
| 12344 |
+
"completions/max_terminated_length": 0.0,
|
| 12345 |
+
"completions/mean_length": 384.0,
|
| 12346 |
+
"completions/mean_terminated_length": 0.0,
|
| 12347 |
+
"completions/min_length": 384.0,
|
| 12348 |
+
"completions/min_terminated_length": 0.0,
|
| 12349 |
+
"entropy": 0.6319026313722134,
|
| 12350 |
+
"epoch": 0.927734375,
|
| 12351 |
+
"frac_reward_zero_std": 0.0,
|
| 12352 |
+
"grad_norm": 0.3944365382194519,
|
| 12353 |
+
"learning_rate": 2.5052631578947375e-06,
|
| 12354 |
+
"loss": 0.0,
|
| 12355 |
+
"num_tokens": 2235096.0,
|
| 12356 |
+
"reward": 0.5051125288009644,
|
| 12357 |
+
"reward_std": 0.4934867322444916,
|
| 12358 |
+
"rewards/reward_func/mean": 0.5051125288009644,
|
| 12359 |
+
"rewards/reward_func/std": 0.5237316489219666,
|
| 12360 |
+
"step": 475
|
| 12361 |
+
},
|
| 12362 |
+
{
|
| 12363 |
+
"clip_ratio/high_max": 0.0,
|
| 12364 |
+
"clip_ratio/high_mean": 0.0,
|
| 12365 |
+
"clip_ratio/low_mean": 0.0,
|
| 12366 |
+
"clip_ratio/low_min": 0.0,
|
| 12367 |
+
"clip_ratio/region_mean": 0.0,
|
| 12368 |
+
"completions/clipped_ratio": 1.0,
|
| 12369 |
+
"completions/max_length": 384.0,
|
| 12370 |
+
"completions/max_terminated_length": 0.0,
|
| 12371 |
+
"completions/mean_length": 384.0,
|
| 12372 |
+
"completions/mean_terminated_length": 0.0,
|
| 12373 |
+
"completions/min_length": 384.0,
|
| 12374 |
+
"completions/min_terminated_length": 0.0,
|
| 12375 |
+
"entropy": 0.34385241754353046,
|
| 12376 |
+
"epoch": 0.9296875,
|
| 12377 |
+
"frac_reward_zero_std": 0.0,
|
| 12378 |
+
"grad_norm": 0.18634098768234253,
|
| 12379 |
+
"learning_rate": 2.5e-06,
|
| 12380 |
+
"loss": 0.0,
|
| 12381 |
+
"num_tokens": 2240048.0,
|
| 12382 |
+
"reward": 0.4440000057220459,
|
| 12383 |
+
"reward_std": 0.5151644945144653,
|
| 12384 |
+
"rewards/reward_func/mean": 0.4440000057220459,
|
| 12385 |
+
"rewards/reward_func/std": 0.47747913002967834,
|
| 12386 |
+
"step": 476
|
| 12387 |
+
},
|
| 12388 |
+
{
|
| 12389 |
+
"clip_ratio/high_max": 0.0,
|
| 12390 |
+
"clip_ratio/high_mean": 0.0,
|
| 12391 |
+
"clip_ratio/low_mean": 0.0,
|
| 12392 |
+
"clip_ratio/low_min": 0.0,
|
| 12393 |
+
"clip_ratio/region_mean": 0.0,
|
| 12394 |
+
"completions/clipped_ratio": 1.0,
|
| 12395 |
+
"completions/max_length": 384.0,
|
| 12396 |
+
"completions/max_terminated_length": 0.0,
|
| 12397 |
+
"completions/mean_length": 384.0,
|
| 12398 |
+
"completions/mean_terminated_length": 0.0,
|
| 12399 |
+
"completions/min_length": 384.0,
|
| 12400 |
+
"completions/min_terminated_length": 0.0,
|
| 12401 |
+
"entropy": 0.5607694983482361,
|
| 12402 |
+
"epoch": 0.931640625,
|
| 12403 |
+
"frac_reward_zero_std": 0.0,
|
| 12404 |
+
"grad_norm": 0.19195789098739624,
|
| 12405 |
+
"learning_rate": 2.4947368421052634e-06,
|
| 12406 |
+
"loss": -0.0,
|
| 12407 |
+
"num_tokens": 2244704.0,
|
| 12408 |
+
"reward": 0.6195999979972839,
|
| 12409 |
+
"reward_std": 0.25623539090156555,
|
| 12410 |
+
"rewards/reward_func/mean": 0.6195999979972839,
|
| 12411 |
+
"rewards/reward_func/std": 0.513155460357666,
|
| 12412 |
+
"step": 477
|
| 12413 |
+
},
|
| 12414 |
+
{
|
| 12415 |
+
"clip_ratio/high_max": 0.0,
|
| 12416 |
+
"clip_ratio/high_mean": 0.0,
|
| 12417 |
+
"clip_ratio/low_mean": 0.0,
|
| 12418 |
+
"clip_ratio/low_min": 0.0,
|
| 12419 |
+
"clip_ratio/region_mean": 0.0,
|
| 12420 |
+
"completions/clipped_ratio": 1.0,
|
| 12421 |
+
"completions/max_length": 384.0,
|
| 12422 |
+
"completions/max_terminated_length": 0.0,
|
| 12423 |
+
"completions/mean_length": 384.0,
|
| 12424 |
+
"completions/mean_terminated_length": 0.0,
|
| 12425 |
+
"completions/min_length": 384.0,
|
| 12426 |
+
"completions/min_terminated_length": 0.0,
|
| 12427 |
+
"entropy": 0.6188374534249306,
|
| 12428 |
+
"epoch": 0.93359375,
|
| 12429 |
+
"frac_reward_zero_std": 0.0,
|
| 12430 |
+
"grad_norm": 0.19206063449382782,
|
| 12431 |
+
"learning_rate": 2.4894736842105265e-06,
|
| 12432 |
+
"loss": 0.0,
|
| 12433 |
+
"num_tokens": 2249352.0,
|
| 12434 |
+
"reward": 0.7438000440597534,
|
| 12435 |
+
"reward_std": 0.4960067868232727,
|
| 12436 |
+
"rewards/reward_func/mean": 0.7437999844551086,
|
| 12437 |
+
"rewards/reward_func/std": 0.45921406149864197,
|
| 12438 |
+
"step": 478
|
| 12439 |
+
},
|
| 12440 |
+
{
|
| 12441 |
+
"clip_ratio/high_max": 0.0,
|
| 12442 |
+
"clip_ratio/high_mean": 0.0,
|
| 12443 |
+
"clip_ratio/low_mean": 0.0,
|
| 12444 |
+
"clip_ratio/low_min": 0.0,
|
| 12445 |
+
"clip_ratio/region_mean": 0.0,
|
| 12446 |
+
"completions/clipped_ratio": 1.0,
|
| 12447 |
+
"completions/max_length": 384.0,
|
| 12448 |
+
"completions/max_terminated_length": 0.0,
|
| 12449 |
+
"completions/mean_length": 384.0,
|
| 12450 |
+
"completions/mean_terminated_length": 0.0,
|
| 12451 |
+
"completions/min_length": 384.0,
|
| 12452 |
+
"completions/min_terminated_length": 0.0,
|
| 12453 |
+
"entropy": 0.7727699540555477,
|
| 12454 |
+
"epoch": 0.935546875,
|
| 12455 |
+
"frac_reward_zero_std": 0.0,
|
| 12456 |
+
"grad_norm": 0.24975650012493134,
|
| 12457 |
+
"learning_rate": 2.4842105263157897e-06,
|
| 12458 |
+
"loss": -0.0,
|
| 12459 |
+
"num_tokens": 2254096.0,
|
| 12460 |
+
"reward": 0.7468999624252319,
|
| 12461 |
+
"reward_std": 0.46515488624572754,
|
| 12462 |
+
"rewards/reward_func/mean": 0.7468999624252319,
|
| 12463 |
+
"rewards/reward_func/std": 0.43068981170654297,
|
| 12464 |
+
"step": 479
|
| 12465 |
+
},
|
| 12466 |
+
{
|
| 12467 |
+
"clip_ratio/high_max": 0.0,
|
| 12468 |
+
"clip_ratio/high_mean": 0.0,
|
| 12469 |
+
"clip_ratio/low_mean": 0.0,
|
| 12470 |
+
"clip_ratio/low_min": 0.0,
|
| 12471 |
+
"clip_ratio/region_mean": 0.0,
|
| 12472 |
+
"completions/clipped_ratio": 1.0,
|
| 12473 |
+
"completions/max_length": 384.0,
|
| 12474 |
+
"completions/max_terminated_length": 0.0,
|
| 12475 |
+
"completions/mean_length": 384.0,
|
| 12476 |
+
"completions/mean_terminated_length": 0.0,
|
| 12477 |
+
"completions/min_length": 384.0,
|
| 12478 |
+
"completions/min_terminated_length": 0.0,
|
| 12479 |
+
"entropy": 0.5285328812897205,
|
| 12480 |
+
"epoch": 0.9375,
|
| 12481 |
+
"frac_reward_zero_std": 0.0,
|
| 12482 |
+
"grad_norm": 0.18033409118652344,
|
| 12483 |
+
"learning_rate": 2.478947368421053e-06,
|
| 12484 |
+
"loss": 0.0,
|
| 12485 |
+
"num_tokens": 2259064.0,
|
| 12486 |
+
"reward": 0.22200000286102295,
|
| 12487 |
+
"reward_std": 0.4440000057220459,
|
| 12488 |
+
"rewards/reward_func/mean": 0.22200000286102295,
|
| 12489 |
+
"rewards/reward_func/std": 0.41106414794921875,
|
| 12490 |
+
"step": 480
|
| 12491 |
+
},
|
| 12492 |
+
{
|
| 12493 |
+
"clip_ratio/high_max": 0.0,
|
| 12494 |
+
"clip_ratio/high_mean": 0.0,
|
| 12495 |
+
"clip_ratio/low_mean": 0.0,
|
| 12496 |
+
"clip_ratio/low_min": 0.0,
|
| 12497 |
+
"clip_ratio/region_mean": 0.0,
|
| 12498 |
+
"completions/clipped_ratio": 1.0,
|
| 12499 |
+
"completions/max_length": 384.0,
|
| 12500 |
+
"completions/max_terminated_length": 0.0,
|
| 12501 |
+
"completions/mean_length": 384.0,
|
| 12502 |
+
"completions/mean_terminated_length": 0.0,
|
| 12503 |
+
"completions/min_length": 384.0,
|
| 12504 |
+
"completions/min_terminated_length": 0.0,
|
| 12505 |
+
"entropy": 0.6708570122718811,
|
| 12506 |
+
"epoch": 0.939453125,
|
| 12507 |
+
"frac_reward_zero_std": 0.0,
|
| 12508 |
+
"grad_norm": 0.23435814678668976,
|
| 12509 |
+
"learning_rate": 2.473684210526316e-06,
|
| 12510 |
+
"loss": 0.0,
|
| 12511 |
+
"num_tokens": 2263720.0,
|
| 12512 |
+
"reward": 0.3711249828338623,
|
| 12513 |
+
"reward_std": 0.5285789370536804,
|
| 12514 |
+
"rewards/reward_func/mean": 0.3711249828338623,
|
| 12515 |
+
"rewards/reward_func/std": 0.5054450631141663,
|
| 12516 |
+
"step": 481
|
| 12517 |
+
},
|
| 12518 |
+
{
|
| 12519 |
+
"clip_ratio/high_max": 0.0,
|
| 12520 |
+
"clip_ratio/high_mean": 0.0,
|
| 12521 |
+
"clip_ratio/low_mean": 0.0,
|
| 12522 |
+
"clip_ratio/low_min": 0.0,
|
| 12523 |
+
"clip_ratio/region_mean": 0.0,
|
| 12524 |
+
"completions/clipped_ratio": 1.0,
|
| 12525 |
+
"completions/max_length": 384.0,
|
| 12526 |
+
"completions/max_terminated_length": 0.0,
|
| 12527 |
+
"completions/mean_length": 384.0,
|
| 12528 |
+
"completions/mean_terminated_length": 0.0,
|
| 12529 |
+
"completions/min_length": 384.0,
|
| 12530 |
+
"completions/min_terminated_length": 0.0,
|
| 12531 |
+
"entropy": 0.6528993677347898,
|
| 12532 |
+
"epoch": 0.94140625,
|
| 12533 |
+
"frac_reward_zero_std": 0.0,
|
| 12534 |
+
"grad_norm": 0.23905879259109497,
|
| 12535 |
+
"learning_rate": 2.468421052631579e-06,
|
| 12536 |
+
"loss": -0.0,
|
| 12537 |
+
"num_tokens": 2268472.0,
|
| 12538 |
+
"reward": 0.6135500073432922,
|
| 12539 |
+
"reward_std": 0.5113159418106079,
|
| 12540 |
+
"rewards/reward_func/mean": 0.6135500073432922,
|
| 12541 |
+
"rewards/reward_func/std": 0.4950350224971771,
|
| 12542 |
+
"step": 482
|
| 12543 |
+
},
|
| 12544 |
+
{
|
| 12545 |
+
"clip_ratio/high_max": 0.0,
|
| 12546 |
+
"clip_ratio/high_mean": 0.0,
|
| 12547 |
+
"clip_ratio/low_mean": 0.0,
|
| 12548 |
+
"clip_ratio/low_min": 0.0,
|
| 12549 |
+
"clip_ratio/region_mean": 0.0,
|
| 12550 |
+
"completions/clipped_ratio": 1.0,
|
| 12551 |
+
"completions/max_length": 384.0,
|
| 12552 |
+
"completions/max_terminated_length": 0.0,
|
| 12553 |
+
"completions/mean_length": 384.0,
|
| 12554 |
+
"completions/mean_terminated_length": 0.0,
|
| 12555 |
+
"completions/min_length": 384.0,
|
| 12556 |
+
"completions/min_terminated_length": 0.0,
|
| 12557 |
+
"entropy": 0.47676665522158146,
|
| 12558 |
+
"epoch": 0.943359375,
|
| 12559 |
+
"frac_reward_zero_std": 0.5,
|
| 12560 |
+
"grad_norm": 0.145980104804039,
|
| 12561 |
+
"learning_rate": 2.4631578947368424e-06,
|
| 12562 |
+
"loss": 0.0,
|
| 12563 |
+
"num_tokens": 2273312.0,
|
| 12564 |
+
"reward": 0.8654749989509583,
|
| 12565 |
+
"reward_std": 0.22585000097751617,
|
| 12566 |
+
"rewards/reward_func/mean": 0.8654749989509583,
|
| 12567 |
+
"rewards/reward_func/std": 0.3194001615047455,
|
| 12568 |
+
"step": 483
|
| 12569 |
+
},
|
| 12570 |
+
{
|
| 12571 |
+
"clip_ratio/high_max": 0.0,
|
| 12572 |
+
"clip_ratio/high_mean": 0.0,
|
| 12573 |
+
"clip_ratio/low_mean": 0.0,
|
| 12574 |
+
"clip_ratio/low_min": 0.0,
|
| 12575 |
+
"clip_ratio/region_mean": 0.0,
|
| 12576 |
+
"completions/clipped_ratio": 1.0,
|
| 12577 |
+
"completions/max_length": 384.0,
|
| 12578 |
+
"completions/max_terminated_length": 0.0,
|
| 12579 |
+
"completions/mean_length": 384.0,
|
| 12580 |
+
"completions/mean_terminated_length": 0.0,
|
| 12581 |
+
"completions/min_length": 384.0,
|
| 12582 |
+
"completions/min_terminated_length": 0.0,
|
| 12583 |
+
"entropy": 0.5406187996268272,
|
| 12584 |
+
"epoch": 0.9453125,
|
| 12585 |
+
"frac_reward_zero_std": 0.0,
|
| 12586 |
+
"grad_norm": 0.23957812786102295,
|
| 12587 |
+
"learning_rate": 2.4578947368421055e-06,
|
| 12588 |
+
"loss": 0.0,
|
| 12589 |
+
"num_tokens": 2278288.0,
|
| 12590 |
+
"reward": 0.2150000035762787,
|
| 12591 |
+
"reward_std": 0.4300000071525574,
|
| 12592 |
+
"rewards/reward_func/mean": 0.2150000035762787,
|
| 12593 |
+
"rewards/reward_func/std": 0.39838388562202454,
|
| 12594 |
+
"step": 484
|
| 12595 |
+
},
|
| 12596 |
+
{
|
| 12597 |
+
"clip_ratio/high_max": 0.0,
|
| 12598 |
+
"clip_ratio/high_mean": 0.0,
|
| 12599 |
+
"clip_ratio/low_mean": 0.0,
|
| 12600 |
+
"clip_ratio/low_min": 0.0,
|
| 12601 |
+
"clip_ratio/region_mean": 0.0,
|
| 12602 |
+
"completions/clipped_ratio": 1.0,
|
| 12603 |
+
"completions/max_length": 384.0,
|
| 12604 |
+
"completions/max_terminated_length": 0.0,
|
| 12605 |
+
"completions/mean_length": 384.0,
|
| 12606 |
+
"completions/mean_terminated_length": 0.0,
|
| 12607 |
+
"completions/min_length": 384.0,
|
| 12608 |
+
"completions/min_terminated_length": 0.0,
|
| 12609 |
+
"entropy": 0.5646249614655972,
|
| 12610 |
+
"epoch": 0.947265625,
|
| 12611 |
+
"frac_reward_zero_std": 0.0,
|
| 12612 |
+
"grad_norm": 0.2331782728433609,
|
| 12613 |
+
"learning_rate": 2.4526315789473687e-06,
|
| 12614 |
+
"loss": 0.0,
|
| 12615 |
+
"num_tokens": 2283248.0,
|
| 12616 |
+
"reward": 0.11626250296831131,
|
| 12617 |
+
"reward_std": 0.2122942954301834,
|
| 12618 |
+
"rewards/reward_func/mean": 0.11626250296831131,
|
| 12619 |
+
"rewards/reward_func/std": 0.28399014472961426,
|
| 12620 |
+
"step": 485
|
| 12621 |
+
},
|
| 12622 |
+
{
|
| 12623 |
+
"clip_ratio/high_max": 0.0,
|
| 12624 |
+
"clip_ratio/high_mean": 0.0,
|
| 12625 |
+
"clip_ratio/low_mean": 0.0,
|
| 12626 |
+
"clip_ratio/low_min": 0.0,
|
| 12627 |
+
"clip_ratio/region_mean": 0.0,
|
| 12628 |
+
"completions/clipped_ratio": 1.0,
|
| 12629 |
+
"completions/max_length": 384.0,
|
| 12630 |
+
"completions/max_terminated_length": 0.0,
|
| 12631 |
+
"completions/mean_length": 384.0,
|
| 12632 |
+
"completions/mean_terminated_length": 0.0,
|
| 12633 |
+
"completions/min_length": 384.0,
|
| 12634 |
+
"completions/min_terminated_length": 0.0,
|
| 12635 |
+
"entropy": 0.5729491971433163,
|
| 12636 |
+
"epoch": 0.94921875,
|
| 12637 |
+
"frac_reward_zero_std": 0.0,
|
| 12638 |
+
"grad_norm": 0.22905918955802917,
|
| 12639 |
+
"learning_rate": 2.447368421052632e-06,
|
| 12640 |
+
"loss": 0.0,
|
| 12641 |
+
"num_tokens": 2287904.0,
|
| 12642 |
+
"reward": 0.49886250495910645,
|
| 12643 |
+
"reward_std": 0.4972279667854309,
|
| 12644 |
+
"rewards/reward_func/mean": 0.49886250495910645,
|
| 12645 |
+
"rewards/reward_func/std": 0.5300286412239075,
|
| 12646 |
+
"step": 486
|
| 12647 |
+
},
|
| 12648 |
+
{
|
| 12649 |
+
"clip_ratio/high_max": 0.0,
|
| 12650 |
+
"clip_ratio/high_mean": 0.0,
|
| 12651 |
+
"clip_ratio/low_mean": 0.0,
|
| 12652 |
+
"clip_ratio/low_min": 0.0,
|
| 12653 |
+
"clip_ratio/region_mean": 0.0,
|
| 12654 |
+
"completions/clipped_ratio": 1.0,
|
| 12655 |
+
"completions/max_length": 384.0,
|
| 12656 |
+
"completions/max_terminated_length": 0.0,
|
| 12657 |
+
"completions/mean_length": 384.0,
|
| 12658 |
+
"completions/mean_terminated_length": 0.0,
|
| 12659 |
+
"completions/min_length": 384.0,
|
| 12660 |
+
"completions/min_terminated_length": 0.0,
|
| 12661 |
+
"entropy": 0.6615194007754326,
|
| 12662 |
+
"epoch": 0.951171875,
|
| 12663 |
+
"frac_reward_zero_std": 0.0,
|
| 12664 |
+
"grad_norm": 0.26248008012771606,
|
| 12665 |
+
"learning_rate": 2.442105263157895e-06,
|
| 12666 |
+
"loss": 0.0,
|
| 12667 |
+
"num_tokens": 2292752.0,
|
| 12668 |
+
"reward": 0.75,
|
| 12669 |
+
"reward_std": 0.5,
|
| 12670 |
+
"rewards/reward_func/mean": 0.75,
|
| 12671 |
+
"rewards/reward_func/std": 0.4629100561141968,
|
| 12672 |
+
"step": 487
|
| 12673 |
+
},
|
| 12674 |
+
{
|
| 12675 |
+
"clip_ratio/high_max": 0.0,
|
| 12676 |
+
"clip_ratio/high_mean": 0.0,
|
| 12677 |
+
"clip_ratio/low_mean": 0.0,
|
| 12678 |
+
"clip_ratio/low_min": 0.0,
|
| 12679 |
+
"clip_ratio/region_mean": 0.0,
|
| 12680 |
+
"completions/clipped_ratio": 1.0,
|
| 12681 |
+
"completions/max_length": 384.0,
|
| 12682 |
+
"completions/max_terminated_length": 0.0,
|
| 12683 |
+
"completions/mean_length": 384.0,
|
| 12684 |
+
"completions/mean_terminated_length": 0.0,
|
| 12685 |
+
"completions/min_length": 384.0,
|
| 12686 |
+
"completions/min_terminated_length": 0.0,
|
| 12687 |
+
"entropy": 0.5113859847187996,
|
| 12688 |
+
"epoch": 0.953125,
|
| 12689 |
+
"frac_reward_zero_std": 0.5,
|
| 12690 |
+
"grad_norm": 0.14094099402427673,
|
| 12691 |
+
"learning_rate": 2.436842105263158e-06,
|
| 12692 |
+
"loss": -0.0,
|
| 12693 |
+
"num_tokens": 2297416.0,
|
| 12694 |
+
"reward": 0.754687488079071,
|
| 12695 |
+
"reward_std": 0.2833658754825592,
|
| 12696 |
+
"rewards/reward_func/mean": 0.754687488079071,
|
| 12697 |
+
"rewards/reward_func/std": 0.45434102416038513,
|
| 12698 |
+
"step": 488
|
| 12699 |
+
},
|
| 12700 |
+
{
|
| 12701 |
+
"clip_ratio/high_max": 0.0,
|
| 12702 |
+
"clip_ratio/high_mean": 0.0,
|
| 12703 |
+
"clip_ratio/low_mean": 0.0,
|
| 12704 |
+
"clip_ratio/low_min": 0.0,
|
| 12705 |
+
"clip_ratio/region_mean": 0.0,
|
| 12706 |
+
"completions/clipped_ratio": 1.0,
|
| 12707 |
+
"completions/max_length": 384.0,
|
| 12708 |
+
"completions/max_terminated_length": 0.0,
|
| 12709 |
+
"completions/mean_length": 384.0,
|
| 12710 |
+
"completions/mean_terminated_length": 0.0,
|
| 12711 |
+
"completions/min_length": 384.0,
|
| 12712 |
+
"completions/min_terminated_length": 0.0,
|
| 12713 |
+
"entropy": 0.6214317940175533,
|
| 12714 |
+
"epoch": 0.955078125,
|
| 12715 |
+
"frac_reward_zero_std": 0.5,
|
| 12716 |
+
"grad_norm": 0.16864041984081268,
|
| 12717 |
+
"learning_rate": 2.4315789473684213e-06,
|
| 12718 |
+
"loss": 0.0,
|
| 12719 |
+
"num_tokens": 2301776.0,
|
| 12720 |
+
"reward": 0.6328125,
|
| 12721 |
+
"reward_std": 0.24523451924324036,
|
| 12722 |
+
"rewards/reward_func/mean": 0.6328125,
|
| 12723 |
+
"rewards/reward_func/std": 0.5071338415145874,
|
| 12724 |
+
"step": 489
|
| 12725 |
+
},
|
| 12726 |
+
{
|
| 12727 |
+
"clip_ratio/high_max": 0.0,
|
| 12728 |
+
"clip_ratio/high_mean": 0.0,
|
| 12729 |
+
"clip_ratio/low_mean": 0.0,
|
| 12730 |
+
"clip_ratio/low_min": 0.0,
|
| 12731 |
+
"clip_ratio/region_mean": 0.0,
|
| 12732 |
+
"completions/clipped_ratio": 1.0,
|
| 12733 |
+
"completions/max_length": 384.0,
|
| 12734 |
+
"completions/max_terminated_length": 0.0,
|
| 12735 |
+
"completions/mean_length": 384.0,
|
| 12736 |
+
"completions/mean_terminated_length": 0.0,
|
| 12737 |
+
"completions/min_length": 384.0,
|
| 12738 |
+
"completions/min_terminated_length": 0.0,
|
| 12739 |
+
"entropy": 0.5048791384324431,
|
| 12740 |
+
"epoch": 0.95703125,
|
| 12741 |
+
"frac_reward_zero_std": 0.5,
|
| 12742 |
+
"grad_norm": 0.13176527619361877,
|
| 12743 |
+
"learning_rate": 2.4263157894736845e-06,
|
| 12744 |
+
"loss": -0.0,
|
| 12745 |
+
"num_tokens": 2306272.0,
|
| 12746 |
+
"reward": 0.8134000301361084,
|
| 12747 |
+
"reward_std": 0.23240000009536743,
|
| 12748 |
+
"rewards/reward_func/mean": 0.8134000301361084,
|
| 12749 |
+
"rewards/reward_func/std": 0.3286632299423218,
|
| 12750 |
+
"step": 490
|
| 12751 |
+
},
|
| 12752 |
+
{
|
| 12753 |
+
"clip_ratio/high_max": 0.0,
|
| 12754 |
+
"clip_ratio/high_mean": 0.0,
|
| 12755 |
+
"clip_ratio/low_mean": 0.0,
|
| 12756 |
+
"clip_ratio/low_min": 0.0,
|
| 12757 |
+
"clip_ratio/region_mean": 0.0,
|
| 12758 |
+
"completions/clipped_ratio": 1.0,
|
| 12759 |
+
"completions/max_length": 384.0,
|
| 12760 |
+
"completions/max_terminated_length": 0.0,
|
| 12761 |
+
"completions/mean_length": 384.0,
|
| 12762 |
+
"completions/mean_terminated_length": 0.0,
|
| 12763 |
+
"completions/min_length": 384.0,
|
| 12764 |
+
"completions/min_terminated_length": 0.0,
|
| 12765 |
+
"entropy": 0.8380111455917358,
|
| 12766 |
+
"epoch": 0.958984375,
|
| 12767 |
+
"frac_reward_zero_std": 0.5,
|
| 12768 |
+
"grad_norm": 0.1738596111536026,
|
| 12769 |
+
"learning_rate": 2.4210526315789477e-06,
|
| 12770 |
+
"loss": -0.0,
|
| 12771 |
+
"num_tokens": 2310752.0,
|
| 12772 |
+
"reward": 0.754687488079071,
|
| 12773 |
+
"reward_std": 0.2833658754825592,
|
| 12774 |
+
"rewards/reward_func/mean": 0.754687488079071,
|
| 12775 |
+
"rewards/reward_func/std": 0.45434102416038513,
|
| 12776 |
+
"step": 491
|
| 12777 |
+
},
|
| 12778 |
+
{
|
| 12779 |
+
"clip_ratio/high_max": 0.0,
|
| 12780 |
+
"clip_ratio/high_mean": 0.0,
|
| 12781 |
+
"clip_ratio/low_mean": 0.0,
|
| 12782 |
+
"clip_ratio/low_min": 0.0,
|
| 12783 |
+
"clip_ratio/region_mean": 0.0,
|
| 12784 |
+
"completions/clipped_ratio": 1.0,
|
| 12785 |
+
"completions/max_length": 384.0,
|
| 12786 |
+
"completions/max_terminated_length": 0.0,
|
| 12787 |
+
"completions/mean_length": 384.0,
|
| 12788 |
+
"completions/mean_terminated_length": 0.0,
|
| 12789 |
+
"completions/min_length": 384.0,
|
| 12790 |
+
"completions/min_terminated_length": 0.0,
|
| 12791 |
+
"entropy": 0.3626231402158737,
|
| 12792 |
+
"epoch": 0.9609375,
|
| 12793 |
+
"frac_reward_zero_std": 0.0,
|
| 12794 |
+
"grad_norm": 0.1840285211801529,
|
| 12795 |
+
"learning_rate": 2.415789473684211e-06,
|
| 12796 |
+
"loss": -0.0,
|
| 12797 |
+
"num_tokens": 2315256.0,
|
| 12798 |
+
"reward": 0.6972000002861023,
|
| 12799 |
+
"reward_std": 0.46480000019073486,
|
| 12800 |
+
"rewards/reward_func/mean": 0.6972000002861023,
|
| 12801 |
+
"rewards/reward_func/std": 0.43032118678092957,
|
| 12802 |
+
"step": 492
|
| 12803 |
+
},
|
| 12804 |
+
{
|
| 12805 |
+
"clip_ratio/high_max": 0.0,
|
| 12806 |
+
"clip_ratio/high_mean": 0.0,
|
| 12807 |
+
"clip_ratio/low_mean": 0.0,
|
| 12808 |
+
"clip_ratio/low_min": 0.0,
|
| 12809 |
+
"clip_ratio/region_mean": 0.0,
|
| 12810 |
+
"completions/clipped_ratio": 1.0,
|
| 12811 |
+
"completions/max_length": 384.0,
|
| 12812 |
+
"completions/max_terminated_length": 0.0,
|
| 12813 |
+
"completions/mean_length": 384.0,
|
| 12814 |
+
"completions/mean_terminated_length": 0.0,
|
| 12815 |
+
"completions/min_length": 384.0,
|
| 12816 |
+
"completions/min_terminated_length": 0.0,
|
| 12817 |
+
"entropy": 0.7078562341630459,
|
| 12818 |
+
"epoch": 0.962890625,
|
| 12819 |
+
"frac_reward_zero_std": 0.0,
|
| 12820 |
+
"grad_norm": 0.20519906282424927,
|
| 12821 |
+
"learning_rate": 2.410526315789474e-06,
|
| 12822 |
+
"loss": -0.0,
|
| 12823 |
+
"num_tokens": 2320008.0,
|
| 12824 |
+
"reward": 0.488349974155426,
|
| 12825 |
+
"reward_std": 0.4877285361289978,
|
| 12826 |
+
"rewards/reward_func/mean": 0.4883500039577484,
|
| 12827 |
+
"rewards/reward_func/std": 0.52220219373703,
|
| 12828 |
+
"step": 493
|
| 12829 |
+
},
|
| 12830 |
+
{
|
| 12831 |
+
"clip_ratio/high_max": 0.0,
|
| 12832 |
+
"clip_ratio/high_mean": 0.0,
|
| 12833 |
+
"clip_ratio/low_mean": 0.0,
|
| 12834 |
+
"clip_ratio/low_min": 0.0,
|
| 12835 |
+
"clip_ratio/region_mean": 0.0,
|
| 12836 |
+
"completions/clipped_ratio": 1.0,
|
| 12837 |
+
"completions/max_length": 384.0,
|
| 12838 |
+
"completions/max_terminated_length": 0.0,
|
| 12839 |
+
"completions/mean_length": 384.0,
|
| 12840 |
+
"completions/mean_terminated_length": 0.0,
|
| 12841 |
+
"completions/min_length": 384.0,
|
| 12842 |
+
"completions/min_terminated_length": 0.0,
|
| 12843 |
+
"entropy": 0.5425986461341381,
|
| 12844 |
+
"epoch": 0.96484375,
|
| 12845 |
+
"frac_reward_zero_std": 0.0,
|
| 12846 |
+
"grad_norm": 0.2026311159133911,
|
| 12847 |
+
"learning_rate": 2.4052631578947367e-06,
|
| 12848 |
+
"loss": 0.0,
|
| 12849 |
+
"num_tokens": 2324960.0,
|
| 12850 |
+
"reward": 0.24543750286102295,
|
| 12851 |
+
"reward_std": 0.2619554400444031,
|
| 12852 |
+
"rewards/reward_func/mean": 0.24543750286102295,
|
| 12853 |
+
"rewards/reward_func/std": 0.3972027599811554,
|
| 12854 |
+
"step": 494
|
| 12855 |
+
},
|
| 12856 |
+
{
|
| 12857 |
+
"clip_ratio/high_max": 0.0,
|
| 12858 |
+
"clip_ratio/high_mean": 0.0,
|
| 12859 |
+
"clip_ratio/low_mean": 0.0,
|
| 12860 |
+
"clip_ratio/low_min": 0.0,
|
| 12861 |
+
"clip_ratio/region_mean": 0.0,
|
| 12862 |
+
"completions/clipped_ratio": 1.0,
|
| 12863 |
+
"completions/max_length": 384.0,
|
| 12864 |
+
"completions/max_terminated_length": 0.0,
|
| 12865 |
+
"completions/mean_length": 384.0,
|
| 12866 |
+
"completions/mean_terminated_length": 0.0,
|
| 12867 |
+
"completions/min_length": 384.0,
|
| 12868 |
+
"completions/min_terminated_length": 0.0,
|
| 12869 |
+
"entropy": 0.54656021296978,
|
| 12870 |
+
"epoch": 0.966796875,
|
| 12871 |
+
"frac_reward_zero_std": 0.0,
|
| 12872 |
+
"grad_norm": 0.2986944913864136,
|
| 12873 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 12874 |
+
"loss": 0.0,
|
| 12875 |
+
"num_tokens": 2329616.0,
|
| 12876 |
+
"reward": 0.26292499899864197,
|
| 12877 |
+
"reward_std": 0.4851202964782715,
|
| 12878 |
+
"rewards/reward_func/mean": 0.26292499899864197,
|
| 12879 |
+
"rewards/reward_func/std": 0.4491825997829437,
|
| 12880 |
+
"step": 495
|
| 12881 |
+
},
|
| 12882 |
+
{
|
| 12883 |
+
"clip_ratio/high_max": 0.0,
|
| 12884 |
+
"clip_ratio/high_mean": 0.0,
|
| 12885 |
+
"clip_ratio/low_mean": 0.0,
|
| 12886 |
+
"clip_ratio/low_min": 0.0,
|
| 12887 |
+
"clip_ratio/region_mean": 0.0,
|
| 12888 |
+
"completions/clipped_ratio": 1.0,
|
| 12889 |
+
"completions/max_length": 384.0,
|
| 12890 |
+
"completions/max_terminated_length": 0.0,
|
| 12891 |
+
"completions/mean_length": 384.0,
|
| 12892 |
+
"completions/mean_terminated_length": 0.0,
|
| 12893 |
+
"completions/min_length": 384.0,
|
| 12894 |
+
"completions/min_terminated_length": 0.0,
|
| 12895 |
+
"entropy": 0.8450591489672661,
|
| 12896 |
+
"epoch": 0.96875,
|
| 12897 |
+
"frac_reward_zero_std": 0.0,
|
| 12898 |
+
"grad_norm": 0.36952802538871765,
|
| 12899 |
+
"learning_rate": 2.3947368421052635e-06,
|
| 12900 |
+
"loss": 0.0,
|
| 12901 |
+
"num_tokens": 2334088.0,
|
| 12902 |
+
"reward": 0.6332374811172485,
|
| 12903 |
+
"reward_std": 0.5205224752426147,
|
| 12904 |
+
"rewards/reward_func/mean": 0.6332374811172485,
|
| 12905 |
+
"rewards/reward_func/std": 0.5004647970199585,
|
| 12906 |
+
"step": 496
|
| 12907 |
+
},
|
| 12908 |
+
{
|
| 12909 |
+
"clip_ratio/high_max": 0.0,
|
| 12910 |
+
"clip_ratio/high_mean": 0.0,
|
| 12911 |
+
"clip_ratio/low_mean": 0.0,
|
| 12912 |
+
"clip_ratio/low_min": 0.0,
|
| 12913 |
+
"clip_ratio/region_mean": 0.0,
|
| 12914 |
+
"completions/clipped_ratio": 1.0,
|
| 12915 |
+
"completions/max_length": 384.0,
|
| 12916 |
+
"completions/max_terminated_length": 0.0,
|
| 12917 |
+
"completions/mean_length": 384.0,
|
| 12918 |
+
"completions/mean_terminated_length": 0.0,
|
| 12919 |
+
"completions/min_length": 384.0,
|
| 12920 |
+
"completions/min_terminated_length": 0.0,
|
| 12921 |
+
"entropy": 0.6656617224216461,
|
| 12922 |
+
"epoch": 0.970703125,
|
| 12923 |
+
"frac_reward_zero_std": 0.0,
|
| 12924 |
+
"grad_norm": 0.19458124041557312,
|
| 12925 |
+
"learning_rate": 2.3894736842105266e-06,
|
| 12926 |
+
"loss": 0.0,
|
| 12927 |
+
"num_tokens": 2339056.0,
|
| 12928 |
+
"reward": 0.3577499985694885,
|
| 12929 |
+
"reward_std": 0.46434903144836426,
|
| 12930 |
+
"rewards/reward_func/mean": 0.3577499985694885,
|
| 12931 |
+
"rewards/reward_func/std": 0.4540364146232605,
|
| 12932 |
+
"step": 497
|
| 12933 |
+
},
|
| 12934 |
+
{
|
| 12935 |
+
"clip_ratio/high_max": 0.0,
|
| 12936 |
+
"clip_ratio/high_mean": 0.0,
|
| 12937 |
+
"clip_ratio/low_mean": 0.0,
|
| 12938 |
+
"clip_ratio/low_min": 0.0,
|
| 12939 |
+
"clip_ratio/region_mean": 0.0,
|
| 12940 |
+
"completions/clipped_ratio": 1.0,
|
| 12941 |
+
"completions/max_length": 384.0,
|
| 12942 |
+
"completions/max_terminated_length": 0.0,
|
| 12943 |
+
"completions/mean_length": 384.0,
|
| 12944 |
+
"completions/mean_terminated_length": 0.0,
|
| 12945 |
+
"completions/min_length": 384.0,
|
| 12946 |
+
"completions/min_terminated_length": 0.0,
|
| 12947 |
+
"entropy": 0.6396152228116989,
|
| 12948 |
+
"epoch": 0.97265625,
|
| 12949 |
+
"frac_reward_zero_std": 0.0,
|
| 12950 |
+
"grad_norm": 0.25116077065467834,
|
| 12951 |
+
"learning_rate": 2.38421052631579e-06,
|
| 12952 |
+
"loss": -0.0,
|
| 12953 |
+
"num_tokens": 2343704.0,
|
| 12954 |
+
"reward": 0.37542498111724854,
|
| 12955 |
+
"reward_std": 0.5320296287536621,
|
| 12956 |
+
"rewards/reward_func/mean": 0.3754250109195709,
|
| 12957 |
+
"rewards/reward_func/std": 0.5113484263420105,
|
| 12958 |
+
"step": 498
|
| 12959 |
+
},
|
| 12960 |
+
{
|
| 12961 |
+
"clip_ratio/high_max": 0.0,
|
| 12962 |
+
"clip_ratio/high_mean": 0.0,
|
| 12963 |
+
"clip_ratio/low_mean": 0.0,
|
| 12964 |
+
"clip_ratio/low_min": 0.0,
|
| 12965 |
+
"clip_ratio/region_mean": 0.0,
|
| 12966 |
+
"completions/clipped_ratio": 1.0,
|
| 12967 |
+
"completions/max_length": 384.0,
|
| 12968 |
+
"completions/max_terminated_length": 0.0,
|
| 12969 |
+
"completions/mean_length": 384.0,
|
| 12970 |
+
"completions/mean_terminated_length": 0.0,
|
| 12971 |
+
"completions/min_length": 384.0,
|
| 12972 |
+
"completions/min_terminated_length": 0.0,
|
| 12973 |
+
"entropy": 0.4765377175062895,
|
| 12974 |
+
"epoch": 0.974609375,
|
| 12975 |
+
"frac_reward_zero_std": 0.5,
|
| 12976 |
+
"grad_norm": 0.15191896259784698,
|
| 12977 |
+
"learning_rate": 2.378947368421053e-06,
|
| 12978 |
+
"loss": 0.0,
|
| 12979 |
+
"num_tokens": 2348496.0,
|
| 12980 |
+
"reward": 0.23672500252723694,
|
| 12981 |
+
"reward_std": 0.26632454991340637,
|
| 12982 |
+
"rewards/reward_func/mean": 0.23672500252723694,
|
| 12983 |
+
"rewards/reward_func/std": 0.43085548281669617,
|
| 12984 |
+
"step": 499
|
| 12985 |
+
},
|
| 12986 |
+
{
|
| 12987 |
+
"clip_ratio/high_max": 0.0,
|
| 12988 |
+
"clip_ratio/high_mean": 0.0,
|
| 12989 |
+
"clip_ratio/low_mean": 0.0,
|
| 12990 |
+
"clip_ratio/low_min": 0.0,
|
| 12991 |
+
"clip_ratio/region_mean": 0.0,
|
| 12992 |
+
"completions/clipped_ratio": 1.0,
|
| 12993 |
+
"completions/max_length": 384.0,
|
| 12994 |
+
"completions/max_terminated_length": 0.0,
|
| 12995 |
+
"completions/mean_length": 384.0,
|
| 12996 |
+
"completions/mean_terminated_length": 0.0,
|
| 12997 |
+
"completions/min_length": 384.0,
|
| 12998 |
+
"completions/min_terminated_length": 0.0,
|
| 12999 |
+
"entropy": 0.7641369961202145,
|
| 13000 |
+
"epoch": 0.9765625,
|
| 13001 |
+
"frac_reward_zero_std": 0.0,
|
| 13002 |
+
"grad_norm": 0.2522798776626587,
|
| 13003 |
+
"learning_rate": 2.373684210526316e-06,
|
| 13004 |
+
"loss": 0.0,
|
| 13005 |
+
"num_tokens": 2353504.0,
|
| 13006 |
+
"reward": 0.475600004196167,
|
| 13007 |
+
"reward_std": 0.5505199432373047,
|
| 13008 |
+
"rewards/reward_func/mean": 0.475600004196167,
|
| 13009 |
+
"rewards/reward_func/std": 0.5098142623901367,
|
| 13010 |
+
"step": 500
|
| 13011 |
}
|
| 13012 |
],
|
| 13013 |
"logging_steps": 1,
|
| 13014 |
"max_steps": 950,
|
| 13015 |
+
"num_input_tokens_seen": 2353504,
|
| 13016 |
"num_train_epochs": 2,
|
| 13017 |
"save_steps": 50,
|
| 13018 |
"stateful_callbacks": {
|