Training in progress, step 1600, checkpoint
Browse files- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +2243 -3
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 936544523
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:170ec0688a7be1f7f2a5aa23ea0e5a60abb655c988cf1f0505588635b428e5aa
|
| 3 |
size 936544523
|
last-checkpoint/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:c18ab5d0441c18232854c58bbfbc8248c2a6b3ba6d26d20fbc7784a942883b62
|
| 3 |
size 1465
|
last-checkpoint/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,
|
|
@@ -8968,6 +8968,2246 @@
|
|
| 8968 |
"learning_rate": 2.8428855621867327e-07,
|
| 8969 |
"loss": 3.2265,
|
| 8970 |
"step": 1280
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8971 |
}
|
| 8972 |
],
|
| 8973 |
"logging_steps": 1,
|
|
@@ -8987,7 +11227,7 @@
|
|
| 8987 |
"attributes": {}
|
| 8988 |
}
|
| 8989 |
},
|
| 8990 |
-
"total_flos":
|
| 8991 |
"train_batch_size": 1,
|
| 8992 |
"trial_name": null,
|
| 8993 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.007112707518798664,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1600,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 8968 |
"learning_rate": 2.8428855621867327e-07,
|
| 8969 |
"loss": 3.2265,
|
| 8970 |
"step": 1280
|
| 8971 |
+
},
|
| 8972 |
+
{
|
| 8973 |
+
"epoch": 0.00569461145723818,
|
| 8974 |
+
"grad_norm": 9.125,
|
| 8975 |
+
"learning_rate": 2.8451083030484895e-07,
|
| 8976 |
+
"loss": 3.1246,
|
| 8977 |
+
"step": 1281
|
| 8978 |
+
},
|
| 8979 |
+
{
|
| 8980 |
+
"epoch": 0.005699056899437429,
|
| 8981 |
+
"grad_norm": 8.875,
|
| 8982 |
+
"learning_rate": 2.847331043910246e-07,
|
| 8983 |
+
"loss": 3.1274,
|
| 8984 |
+
"step": 1282
|
| 8985 |
+
},
|
| 8986 |
+
{
|
| 8987 |
+
"epoch": 0.005703502341636678,
|
| 8988 |
+
"grad_norm": 11.9375,
|
| 8989 |
+
"learning_rate": 2.8495537847720025e-07,
|
| 8990 |
+
"loss": 2.8214,
|
| 8991 |
+
"step": 1283
|
| 8992 |
+
},
|
| 8993 |
+
{
|
| 8994 |
+
"epoch": 0.005707947783835927,
|
| 8995 |
+
"grad_norm": 9.6875,
|
| 8996 |
+
"learning_rate": 2.8517765256337593e-07,
|
| 8997 |
+
"loss": 3.0174,
|
| 8998 |
+
"step": 1284
|
| 8999 |
+
},
|
| 9000 |
+
{
|
| 9001 |
+
"epoch": 0.005712393226035177,
|
| 9002 |
+
"grad_norm": 8.625,
|
| 9003 |
+
"learning_rate": 2.853999266495516e-07,
|
| 9004 |
+
"loss": 3.0763,
|
| 9005 |
+
"step": 1285
|
| 9006 |
+
},
|
| 9007 |
+
{
|
| 9008 |
+
"epoch": 0.005716838668234426,
|
| 9009 |
+
"grad_norm": 8.25,
|
| 9010 |
+
"learning_rate": 2.856222007357273e-07,
|
| 9011 |
+
"loss": 3.2222,
|
| 9012 |
+
"step": 1286
|
| 9013 |
+
},
|
| 9014 |
+
{
|
| 9015 |
+
"epoch": 0.005721284110433675,
|
| 9016 |
+
"grad_norm": 12.5625,
|
| 9017 |
+
"learning_rate": 2.858444748219029e-07,
|
| 9018 |
+
"loss": 2.7932,
|
| 9019 |
+
"step": 1287
|
| 9020 |
+
},
|
| 9021 |
+
{
|
| 9022 |
+
"epoch": 0.005725729552632924,
|
| 9023 |
+
"grad_norm": 10.75,
|
| 9024 |
+
"learning_rate": 2.860667489080786e-07,
|
| 9025 |
+
"loss": 2.8552,
|
| 9026 |
+
"step": 1288
|
| 9027 |
+
},
|
| 9028 |
+
{
|
| 9029 |
+
"epoch": 0.0057301749948321734,
|
| 9030 |
+
"grad_norm": 11.1875,
|
| 9031 |
+
"learning_rate": 2.8628902299425427e-07,
|
| 9032 |
+
"loss": 2.8399,
|
| 9033 |
+
"step": 1289
|
| 9034 |
+
},
|
| 9035 |
+
{
|
| 9036 |
+
"epoch": 0.0057346204370314225,
|
| 9037 |
+
"grad_norm": 9.1875,
|
| 9038 |
+
"learning_rate": 2.865112970804299e-07,
|
| 9039 |
+
"loss": 3.0231,
|
| 9040 |
+
"step": 1290
|
| 9041 |
+
},
|
| 9042 |
+
{
|
| 9043 |
+
"epoch": 0.005739065879230672,
|
| 9044 |
+
"grad_norm": 10.5625,
|
| 9045 |
+
"learning_rate": 2.8673357116660557e-07,
|
| 9046 |
+
"loss": 2.9536,
|
| 9047 |
+
"step": 1291
|
| 9048 |
+
},
|
| 9049 |
+
{
|
| 9050 |
+
"epoch": 0.005743511321429921,
|
| 9051 |
+
"grad_norm": 9.25,
|
| 9052 |
+
"learning_rate": 2.869558452527812e-07,
|
| 9053 |
+
"loss": 3.0822,
|
| 9054 |
+
"step": 1292
|
| 9055 |
+
},
|
| 9056 |
+
{
|
| 9057 |
+
"epoch": 0.00574795676362917,
|
| 9058 |
+
"grad_norm": 11.9375,
|
| 9059 |
+
"learning_rate": 2.8717811933895687e-07,
|
| 9060 |
+
"loss": 2.8038,
|
| 9061 |
+
"step": 1293
|
| 9062 |
+
},
|
| 9063 |
+
{
|
| 9064 |
+
"epoch": 0.00575240220582842,
|
| 9065 |
+
"grad_norm": 9.5,
|
| 9066 |
+
"learning_rate": 2.8740039342513255e-07,
|
| 9067 |
+
"loss": 3.0759,
|
| 9068 |
+
"step": 1294
|
| 9069 |
+
},
|
| 9070 |
+
{
|
| 9071 |
+
"epoch": 0.005756847648027669,
|
| 9072 |
+
"grad_norm": 9.3125,
|
| 9073 |
+
"learning_rate": 2.876226675113082e-07,
|
| 9074 |
+
"loss": 3.1649,
|
| 9075 |
+
"step": 1295
|
| 9076 |
+
},
|
| 9077 |
+
{
|
| 9078 |
+
"epoch": 0.005761293090226918,
|
| 9079 |
+
"grad_norm": 11.5625,
|
| 9080 |
+
"learning_rate": 2.8784494159748385e-07,
|
| 9081 |
+
"loss": 2.8971,
|
| 9082 |
+
"step": 1296
|
| 9083 |
+
},
|
| 9084 |
+
{
|
| 9085 |
+
"epoch": 0.005765738532426167,
|
| 9086 |
+
"grad_norm": 9.5625,
|
| 9087 |
+
"learning_rate": 2.8806721568365953e-07,
|
| 9088 |
+
"loss": 3.1151,
|
| 9089 |
+
"step": 1297
|
| 9090 |
+
},
|
| 9091 |
+
{
|
| 9092 |
+
"epoch": 0.005770183974625416,
|
| 9093 |
+
"grad_norm": 9.6875,
|
| 9094 |
+
"learning_rate": 2.882894897698352e-07,
|
| 9095 |
+
"loss": 3.0496,
|
| 9096 |
+
"step": 1298
|
| 9097 |
+
},
|
| 9098 |
+
{
|
| 9099 |
+
"epoch": 0.005774629416824665,
|
| 9100 |
+
"grad_norm": 9.3125,
|
| 9101 |
+
"learning_rate": 2.885117638560109e-07,
|
| 9102 |
+
"loss": 3.0914,
|
| 9103 |
+
"step": 1299
|
| 9104 |
+
},
|
| 9105 |
+
{
|
| 9106 |
+
"epoch": 0.005779074859023914,
|
| 9107 |
+
"grad_norm": 10.625,
|
| 9108 |
+
"learning_rate": 2.8873403794218656e-07,
|
| 9109 |
+
"loss": 2.9635,
|
| 9110 |
+
"step": 1300
|
| 9111 |
+
},
|
| 9112 |
+
{
|
| 9113 |
+
"epoch": 0.005783520301223163,
|
| 9114 |
+
"grad_norm": 9.6875,
|
| 9115 |
+
"learning_rate": 2.889563120283622e-07,
|
| 9116 |
+
"loss": 3.0395,
|
| 9117 |
+
"step": 1301
|
| 9118 |
+
},
|
| 9119 |
+
{
|
| 9120 |
+
"epoch": 0.005787965743422413,
|
| 9121 |
+
"grad_norm": 10.3125,
|
| 9122 |
+
"learning_rate": 2.8917858611453787e-07,
|
| 9123 |
+
"loss": 2.8086,
|
| 9124 |
+
"step": 1302
|
| 9125 |
+
},
|
| 9126 |
+
{
|
| 9127 |
+
"epoch": 0.005792411185621662,
|
| 9128 |
+
"grad_norm": 9.6875,
|
| 9129 |
+
"learning_rate": 2.8940086020071354e-07,
|
| 9130 |
+
"loss": 3.0765,
|
| 9131 |
+
"step": 1303
|
| 9132 |
+
},
|
| 9133 |
+
{
|
| 9134 |
+
"epoch": 0.005796856627820911,
|
| 9135 |
+
"grad_norm": 11.75,
|
| 9136 |
+
"learning_rate": 2.896231342868892e-07,
|
| 9137 |
+
"loss": 2.7715,
|
| 9138 |
+
"step": 1304
|
| 9139 |
+
},
|
| 9140 |
+
{
|
| 9141 |
+
"epoch": 0.00580130207002016,
|
| 9142 |
+
"grad_norm": 8.4375,
|
| 9143 |
+
"learning_rate": 2.8984540837306485e-07,
|
| 9144 |
+
"loss": 3.1307,
|
| 9145 |
+
"step": 1305
|
| 9146 |
+
},
|
| 9147 |
+
{
|
| 9148 |
+
"epoch": 0.005805747512219409,
|
| 9149 |
+
"grad_norm": 10.125,
|
| 9150 |
+
"learning_rate": 2.9006768245924047e-07,
|
| 9151 |
+
"loss": 2.9209,
|
| 9152 |
+
"step": 1306
|
| 9153 |
+
},
|
| 9154 |
+
{
|
| 9155 |
+
"epoch": 0.005810192954418658,
|
| 9156 |
+
"grad_norm": 10.6875,
|
| 9157 |
+
"learning_rate": 2.9028995654541615e-07,
|
| 9158 |
+
"loss": 2.926,
|
| 9159 |
+
"step": 1307
|
| 9160 |
+
},
|
| 9161 |
+
{
|
| 9162 |
+
"epoch": 0.005814638396617907,
|
| 9163 |
+
"grad_norm": 10.9375,
|
| 9164 |
+
"learning_rate": 2.905122306315918e-07,
|
| 9165 |
+
"loss": 2.8759,
|
| 9166 |
+
"step": 1308
|
| 9167 |
+
},
|
| 9168 |
+
{
|
| 9169 |
+
"epoch": 0.005819083838817156,
|
| 9170 |
+
"grad_norm": 10.25,
|
| 9171 |
+
"learning_rate": 2.907345047177675e-07,
|
| 9172 |
+
"loss": 2.9456,
|
| 9173 |
+
"step": 1309
|
| 9174 |
+
},
|
| 9175 |
+
{
|
| 9176 |
+
"epoch": 0.0058235292810164055,
|
| 9177 |
+
"grad_norm": 11.875,
|
| 9178 |
+
"learning_rate": 2.909567788039432e-07,
|
| 9179 |
+
"loss": 2.874,
|
| 9180 |
+
"step": 1310
|
| 9181 |
+
},
|
| 9182 |
+
{
|
| 9183 |
+
"epoch": 0.005827974723215655,
|
| 9184 |
+
"grad_norm": 9.9375,
|
| 9185 |
+
"learning_rate": 2.911790528901188e-07,
|
| 9186 |
+
"loss": 2.9463,
|
| 9187 |
+
"step": 1311
|
| 9188 |
+
},
|
| 9189 |
+
{
|
| 9190 |
+
"epoch": 0.0058324201654149045,
|
| 9191 |
+
"grad_norm": 9.8125,
|
| 9192 |
+
"learning_rate": 2.914013269762945e-07,
|
| 9193 |
+
"loss": 3.0364,
|
| 9194 |
+
"step": 1312
|
| 9195 |
+
},
|
| 9196 |
+
{
|
| 9197 |
+
"epoch": 0.0058368656076141535,
|
| 9198 |
+
"grad_norm": 7.78125,
|
| 9199 |
+
"learning_rate": 2.9162360106247016e-07,
|
| 9200 |
+
"loss": 3.1732,
|
| 9201 |
+
"step": 1313
|
| 9202 |
+
},
|
| 9203 |
+
{
|
| 9204 |
+
"epoch": 0.005841311049813403,
|
| 9205 |
+
"grad_norm": 8.625,
|
| 9206 |
+
"learning_rate": 2.9184587514864584e-07,
|
| 9207 |
+
"loss": 3.236,
|
| 9208 |
+
"step": 1314
|
| 9209 |
+
},
|
| 9210 |
+
{
|
| 9211 |
+
"epoch": 0.005845756492012652,
|
| 9212 |
+
"grad_norm": 7.71875,
|
| 9213 |
+
"learning_rate": 2.9206814923482146e-07,
|
| 9214 |
+
"loss": 3.3732,
|
| 9215 |
+
"step": 1315
|
| 9216 |
+
},
|
| 9217 |
+
{
|
| 9218 |
+
"epoch": 0.005850201934211901,
|
| 9219 |
+
"grad_norm": 7.0,
|
| 9220 |
+
"learning_rate": 2.9229042332099714e-07,
|
| 9221 |
+
"loss": 3.3333,
|
| 9222 |
+
"step": 1316
|
| 9223 |
+
},
|
| 9224 |
+
{
|
| 9225 |
+
"epoch": 0.00585464737641115,
|
| 9226 |
+
"grad_norm": 7.375,
|
| 9227 |
+
"learning_rate": 2.925126974071728e-07,
|
| 9228 |
+
"loss": 3.2613,
|
| 9229 |
+
"step": 1317
|
| 9230 |
+
},
|
| 9231 |
+
{
|
| 9232 |
+
"epoch": 0.005859092818610399,
|
| 9233 |
+
"grad_norm": 7.59375,
|
| 9234 |
+
"learning_rate": 2.927349714933485e-07,
|
| 9235 |
+
"loss": 3.2561,
|
| 9236 |
+
"step": 1318
|
| 9237 |
+
},
|
| 9238 |
+
{
|
| 9239 |
+
"epoch": 0.005863538260809649,
|
| 9240 |
+
"grad_norm": 9.25,
|
| 9241 |
+
"learning_rate": 2.929572455795241e-07,
|
| 9242 |
+
"loss": 3.1236,
|
| 9243 |
+
"step": 1319
|
| 9244 |
+
},
|
| 9245 |
+
{
|
| 9246 |
+
"epoch": 0.005867983703008898,
|
| 9247 |
+
"grad_norm": 7.40625,
|
| 9248 |
+
"learning_rate": 2.931795196656998e-07,
|
| 9249 |
+
"loss": 3.4843,
|
| 9250 |
+
"step": 1320
|
| 9251 |
+
},
|
| 9252 |
+
{
|
| 9253 |
+
"epoch": 0.005872429145208147,
|
| 9254 |
+
"grad_norm": 10.3125,
|
| 9255 |
+
"learning_rate": 2.934017937518755e-07,
|
| 9256 |
+
"loss": 3.0469,
|
| 9257 |
+
"step": 1321
|
| 9258 |
+
},
|
| 9259 |
+
{
|
| 9260 |
+
"epoch": 0.005876874587407396,
|
| 9261 |
+
"grad_norm": 10.1875,
|
| 9262 |
+
"learning_rate": 2.9362406783805116e-07,
|
| 9263 |
+
"loss": 2.9779,
|
| 9264 |
+
"step": 1322
|
| 9265 |
+
},
|
| 9266 |
+
{
|
| 9267 |
+
"epoch": 0.005881320029606645,
|
| 9268 |
+
"grad_norm": 9.8125,
|
| 9269 |
+
"learning_rate": 2.938463419242268e-07,
|
| 9270 |
+
"loss": 3.0385,
|
| 9271 |
+
"step": 1323
|
| 9272 |
+
},
|
| 9273 |
+
{
|
| 9274 |
+
"epoch": 0.005885765471805894,
|
| 9275 |
+
"grad_norm": 8.5,
|
| 9276 |
+
"learning_rate": 2.9406861601040246e-07,
|
| 9277 |
+
"loss": 3.2974,
|
| 9278 |
+
"step": 1324
|
| 9279 |
+
},
|
| 9280 |
+
{
|
| 9281 |
+
"epoch": 0.005890210914005143,
|
| 9282 |
+
"grad_norm": 10.1875,
|
| 9283 |
+
"learning_rate": 2.942908900965781e-07,
|
| 9284 |
+
"loss": 2.9625,
|
| 9285 |
+
"step": 1325
|
| 9286 |
+
},
|
| 9287 |
+
{
|
| 9288 |
+
"epoch": 0.005894656356204392,
|
| 9289 |
+
"grad_norm": 10.1875,
|
| 9290 |
+
"learning_rate": 2.9451316418275376e-07,
|
| 9291 |
+
"loss": 3.0334,
|
| 9292 |
+
"step": 1326
|
| 9293 |
+
},
|
| 9294 |
+
{
|
| 9295 |
+
"epoch": 0.005899101798403641,
|
| 9296 |
+
"grad_norm": 11.4375,
|
| 9297 |
+
"learning_rate": 2.9473543826892944e-07,
|
| 9298 |
+
"loss": 2.8511,
|
| 9299 |
+
"step": 1327
|
| 9300 |
+
},
|
| 9301 |
+
{
|
| 9302 |
+
"epoch": 0.005903547240602891,
|
| 9303 |
+
"grad_norm": 11.625,
|
| 9304 |
+
"learning_rate": 2.949577123551051e-07,
|
| 9305 |
+
"loss": 2.8895,
|
| 9306 |
+
"step": 1328
|
| 9307 |
+
},
|
| 9308 |
+
{
|
| 9309 |
+
"epoch": 0.00590799268280214,
|
| 9310 |
+
"grad_norm": 9.4375,
|
| 9311 |
+
"learning_rate": 2.9517998644128074e-07,
|
| 9312 |
+
"loss": 3.0653,
|
| 9313 |
+
"step": 1329
|
| 9314 |
+
},
|
| 9315 |
+
{
|
| 9316 |
+
"epoch": 0.005912438125001389,
|
| 9317 |
+
"grad_norm": 8.375,
|
| 9318 |
+
"learning_rate": 2.954022605274564e-07,
|
| 9319 |
+
"loss": 3.1183,
|
| 9320 |
+
"step": 1330
|
| 9321 |
+
},
|
| 9322 |
+
{
|
| 9323 |
+
"epoch": 0.005916883567200638,
|
| 9324 |
+
"grad_norm": 9.4375,
|
| 9325 |
+
"learning_rate": 2.956245346136321e-07,
|
| 9326 |
+
"loss": 3.0215,
|
| 9327 |
+
"step": 1331
|
| 9328 |
+
},
|
| 9329 |
+
{
|
| 9330 |
+
"epoch": 0.0059213290093998875,
|
| 9331 |
+
"grad_norm": 11.0,
|
| 9332 |
+
"learning_rate": 2.958468086998078e-07,
|
| 9333 |
+
"loss": 3.0208,
|
| 9334 |
+
"step": 1332
|
| 9335 |
+
},
|
| 9336 |
+
{
|
| 9337 |
+
"epoch": 0.0059257744515991365,
|
| 9338 |
+
"grad_norm": 10.25,
|
| 9339 |
+
"learning_rate": 2.960690827859834e-07,
|
| 9340 |
+
"loss": 3.0723,
|
| 9341 |
+
"step": 1333
|
| 9342 |
+
},
|
| 9343 |
+
{
|
| 9344 |
+
"epoch": 0.005930219893798386,
|
| 9345 |
+
"grad_norm": 9.5625,
|
| 9346 |
+
"learning_rate": 2.962913568721591e-07,
|
| 9347 |
+
"loss": 3.0629,
|
| 9348 |
+
"step": 1334
|
| 9349 |
+
},
|
| 9350 |
+
{
|
| 9351 |
+
"epoch": 0.005934665335997635,
|
| 9352 |
+
"grad_norm": 11.3125,
|
| 9353 |
+
"learning_rate": 2.9651363095833476e-07,
|
| 9354 |
+
"loss": 2.9396,
|
| 9355 |
+
"step": 1335
|
| 9356 |
+
},
|
| 9357 |
+
{
|
| 9358 |
+
"epoch": 0.005939110778196885,
|
| 9359 |
+
"grad_norm": 8.125,
|
| 9360 |
+
"learning_rate": 2.9673590504451043e-07,
|
| 9361 |
+
"loss": 3.3736,
|
| 9362 |
+
"step": 1336
|
| 9363 |
+
},
|
| 9364 |
+
{
|
| 9365 |
+
"epoch": 0.005943556220396134,
|
| 9366 |
+
"grad_norm": 8.8125,
|
| 9367 |
+
"learning_rate": 2.969581791306861e-07,
|
| 9368 |
+
"loss": 3.0242,
|
| 9369 |
+
"step": 1337
|
| 9370 |
+
},
|
| 9371 |
+
{
|
| 9372 |
+
"epoch": 0.005948001662595383,
|
| 9373 |
+
"grad_norm": 8.4375,
|
| 9374 |
+
"learning_rate": 2.9718045321686174e-07,
|
| 9375 |
+
"loss": 3.1464,
|
| 9376 |
+
"step": 1338
|
| 9377 |
+
},
|
| 9378 |
+
{
|
| 9379 |
+
"epoch": 0.005952447104794632,
|
| 9380 |
+
"grad_norm": 9.0625,
|
| 9381 |
+
"learning_rate": 2.974027273030374e-07,
|
| 9382 |
+
"loss": 3.1056,
|
| 9383 |
+
"step": 1339
|
| 9384 |
+
},
|
| 9385 |
+
{
|
| 9386 |
+
"epoch": 0.005956892546993881,
|
| 9387 |
+
"grad_norm": 9.875,
|
| 9388 |
+
"learning_rate": 2.9762500138921304e-07,
|
| 9389 |
+
"loss": 2.9336,
|
| 9390 |
+
"step": 1340
|
| 9391 |
+
},
|
| 9392 |
+
{
|
| 9393 |
+
"epoch": 0.00596133798919313,
|
| 9394 |
+
"grad_norm": 9.3125,
|
| 9395 |
+
"learning_rate": 2.978472754753887e-07,
|
| 9396 |
+
"loss": 2.9919,
|
| 9397 |
+
"step": 1341
|
| 9398 |
+
},
|
| 9399 |
+
{
|
| 9400 |
+
"epoch": 0.005965783431392379,
|
| 9401 |
+
"grad_norm": 9.25,
|
| 9402 |
+
"learning_rate": 2.980695495615644e-07,
|
| 9403 |
+
"loss": 3.1238,
|
| 9404 |
+
"step": 1342
|
| 9405 |
+
},
|
| 9406 |
+
{
|
| 9407 |
+
"epoch": 0.005970228873591628,
|
| 9408 |
+
"grad_norm": 11.0625,
|
| 9409 |
+
"learning_rate": 2.9829182364774e-07,
|
| 9410 |
+
"loss": 2.907,
|
| 9411 |
+
"step": 1343
|
| 9412 |
+
},
|
| 9413 |
+
{
|
| 9414 |
+
"epoch": 0.005974674315790877,
|
| 9415 |
+
"grad_norm": 11.0625,
|
| 9416 |
+
"learning_rate": 2.985140977339157e-07,
|
| 9417 |
+
"loss": 2.8599,
|
| 9418 |
+
"step": 1344
|
| 9419 |
+
},
|
| 9420 |
+
{
|
| 9421 |
+
"epoch": 0.005979119757990127,
|
| 9422 |
+
"grad_norm": 9.4375,
|
| 9423 |
+
"learning_rate": 2.987363718200914e-07,
|
| 9424 |
+
"loss": 3.0859,
|
| 9425 |
+
"step": 1345
|
| 9426 |
+
},
|
| 9427 |
+
{
|
| 9428 |
+
"epoch": 0.005983565200189376,
|
| 9429 |
+
"grad_norm": 9.0,
|
| 9430 |
+
"learning_rate": 2.9895864590626705e-07,
|
| 9431 |
+
"loss": 3.0753,
|
| 9432 |
+
"step": 1346
|
| 9433 |
+
},
|
| 9434 |
+
{
|
| 9435 |
+
"epoch": 0.005988010642388625,
|
| 9436 |
+
"grad_norm": 9.0625,
|
| 9437 |
+
"learning_rate": 2.9918091999244273e-07,
|
| 9438 |
+
"loss": 3.0776,
|
| 9439 |
+
"step": 1347
|
| 9440 |
+
},
|
| 9441 |
+
{
|
| 9442 |
+
"epoch": 0.005992456084587874,
|
| 9443 |
+
"grad_norm": 10.75,
|
| 9444 |
+
"learning_rate": 2.9940319407861836e-07,
|
| 9445 |
+
"loss": 2.9689,
|
| 9446 |
+
"step": 1348
|
| 9447 |
+
},
|
| 9448 |
+
{
|
| 9449 |
+
"epoch": 0.005996901526787123,
|
| 9450 |
+
"grad_norm": 10.6875,
|
| 9451 |
+
"learning_rate": 2.9962546816479403e-07,
|
| 9452 |
+
"loss": 2.9394,
|
| 9453 |
+
"step": 1349
|
| 9454 |
+
},
|
| 9455 |
+
{
|
| 9456 |
+
"epoch": 0.006001346968986372,
|
| 9457 |
+
"grad_norm": 9.5,
|
| 9458 |
+
"learning_rate": 2.998477422509697e-07,
|
| 9459 |
+
"loss": 2.9483,
|
| 9460 |
+
"step": 1350
|
| 9461 |
+
},
|
| 9462 |
+
{
|
| 9463 |
+
"epoch": 0.006005792411185621,
|
| 9464 |
+
"grad_norm": 7.25,
|
| 9465 |
+
"learning_rate": 3.000700163371454e-07,
|
| 9466 |
+
"loss": 3.2638,
|
| 9467 |
+
"step": 1351
|
| 9468 |
+
},
|
| 9469 |
+
{
|
| 9470 |
+
"epoch": 0.0060102378533848705,
|
| 9471 |
+
"grad_norm": 11.125,
|
| 9472 |
+
"learning_rate": 3.00292290423321e-07,
|
| 9473 |
+
"loss": 2.8421,
|
| 9474 |
+
"step": 1352
|
| 9475 |
+
},
|
| 9476 |
+
{
|
| 9477 |
+
"epoch": 0.00601468329558412,
|
| 9478 |
+
"grad_norm": 10.3125,
|
| 9479 |
+
"learning_rate": 3.005145645094967e-07,
|
| 9480 |
+
"loss": 3.0807,
|
| 9481 |
+
"step": 1353
|
| 9482 |
+
},
|
| 9483 |
+
{
|
| 9484 |
+
"epoch": 0.0060191287377833694,
|
| 9485 |
+
"grad_norm": 10.625,
|
| 9486 |
+
"learning_rate": 3.0073683859567237e-07,
|
| 9487 |
+
"loss": 3.0317,
|
| 9488 |
+
"step": 1354
|
| 9489 |
+
},
|
| 9490 |
+
{
|
| 9491 |
+
"epoch": 0.0060235741799826185,
|
| 9492 |
+
"grad_norm": 10.3125,
|
| 9493 |
+
"learning_rate": 3.0095911268184805e-07,
|
| 9494 |
+
"loss": 2.9453,
|
| 9495 |
+
"step": 1355
|
| 9496 |
+
},
|
| 9497 |
+
{
|
| 9498 |
+
"epoch": 0.006028019622181868,
|
| 9499 |
+
"grad_norm": 10.9375,
|
| 9500 |
+
"learning_rate": 3.0118138676802367e-07,
|
| 9501 |
+
"loss": 2.9334,
|
| 9502 |
+
"step": 1356
|
| 9503 |
+
},
|
| 9504 |
+
{
|
| 9505 |
+
"epoch": 0.006032465064381117,
|
| 9506 |
+
"grad_norm": 10.0625,
|
| 9507 |
+
"learning_rate": 3.014036608541993e-07,
|
| 9508 |
+
"loss": 2.9875,
|
| 9509 |
+
"step": 1357
|
| 9510 |
+
},
|
| 9511 |
+
{
|
| 9512 |
+
"epoch": 0.006036910506580366,
|
| 9513 |
+
"grad_norm": 10.25,
|
| 9514 |
+
"learning_rate": 3.01625934940375e-07,
|
| 9515 |
+
"loss": 2.9972,
|
| 9516 |
+
"step": 1358
|
| 9517 |
+
},
|
| 9518 |
+
{
|
| 9519 |
+
"epoch": 0.006041355948779615,
|
| 9520 |
+
"grad_norm": 11.0625,
|
| 9521 |
+
"learning_rate": 3.0184820902655065e-07,
|
| 9522 |
+
"loss": 2.9373,
|
| 9523 |
+
"step": 1359
|
| 9524 |
+
},
|
| 9525 |
+
{
|
| 9526 |
+
"epoch": 0.006045801390978864,
|
| 9527 |
+
"grad_norm": 11.9375,
|
| 9528 |
+
"learning_rate": 3.0207048311272633e-07,
|
| 9529 |
+
"loss": 2.8333,
|
| 9530 |
+
"step": 1360
|
| 9531 |
+
},
|
| 9532 |
+
{
|
| 9533 |
+
"epoch": 0.006050246833178113,
|
| 9534 |
+
"grad_norm": 10.5625,
|
| 9535 |
+
"learning_rate": 3.02292757198902e-07,
|
| 9536 |
+
"loss": 2.9671,
|
| 9537 |
+
"step": 1361
|
| 9538 |
+
},
|
| 9539 |
+
{
|
| 9540 |
+
"epoch": 0.006054692275377363,
|
| 9541 |
+
"grad_norm": 9.9375,
|
| 9542 |
+
"learning_rate": 3.0251503128507763e-07,
|
| 9543 |
+
"loss": 2.8971,
|
| 9544 |
+
"step": 1362
|
| 9545 |
+
},
|
| 9546 |
+
{
|
| 9547 |
+
"epoch": 0.006059137717576612,
|
| 9548 |
+
"grad_norm": 11.9375,
|
| 9549 |
+
"learning_rate": 3.027373053712533e-07,
|
| 9550 |
+
"loss": 2.6949,
|
| 9551 |
+
"step": 1363
|
| 9552 |
+
},
|
| 9553 |
+
{
|
| 9554 |
+
"epoch": 0.006063583159775861,
|
| 9555 |
+
"grad_norm": 10.75,
|
| 9556 |
+
"learning_rate": 3.02959579457429e-07,
|
| 9557 |
+
"loss": 2.9037,
|
| 9558 |
+
"step": 1364
|
| 9559 |
+
},
|
| 9560 |
+
{
|
| 9561 |
+
"epoch": 0.00606802860197511,
|
| 9562 |
+
"grad_norm": 9.25,
|
| 9563 |
+
"learning_rate": 3.0318185354360467e-07,
|
| 9564 |
+
"loss": 3.2175,
|
| 9565 |
+
"step": 1365
|
| 9566 |
+
},
|
| 9567 |
+
{
|
| 9568 |
+
"epoch": 0.006072474044174359,
|
| 9569 |
+
"grad_norm": 13.6875,
|
| 9570 |
+
"learning_rate": 3.034041276297803e-07,
|
| 9571 |
+
"loss": 2.669,
|
| 9572 |
+
"step": 1366
|
| 9573 |
+
},
|
| 9574 |
+
{
|
| 9575 |
+
"epoch": 0.006076919486373608,
|
| 9576 |
+
"grad_norm": 7.96875,
|
| 9577 |
+
"learning_rate": 3.0362640171595597e-07,
|
| 9578 |
+
"loss": 3.303,
|
| 9579 |
+
"step": 1367
|
| 9580 |
+
},
|
| 9581 |
+
{
|
| 9582 |
+
"epoch": 0.006081364928572857,
|
| 9583 |
+
"grad_norm": 7.0,
|
| 9584 |
+
"learning_rate": 3.0384867580213165e-07,
|
| 9585 |
+
"loss": 3.3917,
|
| 9586 |
+
"step": 1368
|
| 9587 |
+
},
|
| 9588 |
+
{
|
| 9589 |
+
"epoch": 0.006085810370772106,
|
| 9590 |
+
"grad_norm": 10.25,
|
| 9591 |
+
"learning_rate": 3.040709498883073e-07,
|
| 9592 |
+
"loss": 2.9504,
|
| 9593 |
+
"step": 1369
|
| 9594 |
+
},
|
| 9595 |
+
{
|
| 9596 |
+
"epoch": 0.006090255812971356,
|
| 9597 |
+
"grad_norm": 12.125,
|
| 9598 |
+
"learning_rate": 3.0429322397448295e-07,
|
| 9599 |
+
"loss": 2.7856,
|
| 9600 |
+
"step": 1370
|
| 9601 |
+
},
|
| 9602 |
+
{
|
| 9603 |
+
"epoch": 0.006094701255170605,
|
| 9604 |
+
"grad_norm": 8.875,
|
| 9605 |
+
"learning_rate": 3.0451549806065863e-07,
|
| 9606 |
+
"loss": 3.1407,
|
| 9607 |
+
"step": 1371
|
| 9608 |
+
},
|
| 9609 |
+
{
|
| 9610 |
+
"epoch": 0.006099146697369854,
|
| 9611 |
+
"grad_norm": 9.25,
|
| 9612 |
+
"learning_rate": 3.047377721468343e-07,
|
| 9613 |
+
"loss": 3.1882,
|
| 9614 |
+
"step": 1372
|
| 9615 |
+
},
|
| 9616 |
+
{
|
| 9617 |
+
"epoch": 0.006103592139569103,
|
| 9618 |
+
"grad_norm": 10.0,
|
| 9619 |
+
"learning_rate": 3.0496004623300993e-07,
|
| 9620 |
+
"loss": 2.9092,
|
| 9621 |
+
"step": 1373
|
| 9622 |
+
},
|
| 9623 |
+
{
|
| 9624 |
+
"epoch": 0.006108037581768352,
|
| 9625 |
+
"grad_norm": 11.3125,
|
| 9626 |
+
"learning_rate": 3.051823203191856e-07,
|
| 9627 |
+
"loss": 2.9245,
|
| 9628 |
+
"step": 1374
|
| 9629 |
+
},
|
| 9630 |
+
{
|
| 9631 |
+
"epoch": 0.0061124830239676015,
|
| 9632 |
+
"grad_norm": 10.125,
|
| 9633 |
+
"learning_rate": 3.054045944053613e-07,
|
| 9634 |
+
"loss": 2.9332,
|
| 9635 |
+
"step": 1375
|
| 9636 |
+
},
|
| 9637 |
+
{
|
| 9638 |
+
"epoch": 0.0061169284661668506,
|
| 9639 |
+
"grad_norm": 10.4375,
|
| 9640 |
+
"learning_rate": 3.056268684915369e-07,
|
| 9641 |
+
"loss": 2.942,
|
| 9642 |
+
"step": 1376
|
| 9643 |
+
},
|
| 9644 |
+
{
|
| 9645 |
+
"epoch": 0.0061213739083661,
|
| 9646 |
+
"grad_norm": 8.75,
|
| 9647 |
+
"learning_rate": 3.058491425777126e-07,
|
| 9648 |
+
"loss": 3.152,
|
| 9649 |
+
"step": 1377
|
| 9650 |
+
},
|
| 9651 |
+
{
|
| 9652 |
+
"epoch": 0.006125819350565349,
|
| 9653 |
+
"grad_norm": 10.3125,
|
| 9654 |
+
"learning_rate": 3.0607141666388827e-07,
|
| 9655 |
+
"loss": 2.9159,
|
| 9656 |
+
"step": 1378
|
| 9657 |
+
},
|
| 9658 |
+
{
|
| 9659 |
+
"epoch": 0.006130264792764599,
|
| 9660 |
+
"grad_norm": 8.3125,
|
| 9661 |
+
"learning_rate": 3.0629369075006394e-07,
|
| 9662 |
+
"loss": 3.1517,
|
| 9663 |
+
"step": 1379
|
| 9664 |
+
},
|
| 9665 |
+
{
|
| 9666 |
+
"epoch": 0.006134710234963848,
|
| 9667 |
+
"grad_norm": 11.0,
|
| 9668 |
+
"learning_rate": 3.0651596483623957e-07,
|
| 9669 |
+
"loss": 2.9173,
|
| 9670 |
+
"step": 1380
|
| 9671 |
+
},
|
| 9672 |
+
{
|
| 9673 |
+
"epoch": 0.006139155677163097,
|
| 9674 |
+
"grad_norm": 9.0,
|
| 9675 |
+
"learning_rate": 3.0673823892241525e-07,
|
| 9676 |
+
"loss": 3.0811,
|
| 9677 |
+
"step": 1381
|
| 9678 |
+
},
|
| 9679 |
+
{
|
| 9680 |
+
"epoch": 0.006143601119362346,
|
| 9681 |
+
"grad_norm": 10.75,
|
| 9682 |
+
"learning_rate": 3.069605130085909e-07,
|
| 9683 |
+
"loss": 2.8594,
|
| 9684 |
+
"step": 1382
|
| 9685 |
+
},
|
| 9686 |
+
{
|
| 9687 |
+
"epoch": 0.006148046561561595,
|
| 9688 |
+
"grad_norm": 11.0625,
|
| 9689 |
+
"learning_rate": 3.071827870947666e-07,
|
| 9690 |
+
"loss": 3.0104,
|
| 9691 |
+
"step": 1383
|
| 9692 |
+
},
|
| 9693 |
+
{
|
| 9694 |
+
"epoch": 0.006152492003760844,
|
| 9695 |
+
"grad_norm": 11.125,
|
| 9696 |
+
"learning_rate": 3.0740506118094223e-07,
|
| 9697 |
+
"loss": 2.8202,
|
| 9698 |
+
"step": 1384
|
| 9699 |
+
},
|
| 9700 |
+
{
|
| 9701 |
+
"epoch": 0.006156937445960093,
|
| 9702 |
+
"grad_norm": 9.25,
|
| 9703 |
+
"learning_rate": 3.076273352671179e-07,
|
| 9704 |
+
"loss": 3.0002,
|
| 9705 |
+
"step": 1385
|
| 9706 |
+
},
|
| 9707 |
+
{
|
| 9708 |
+
"epoch": 0.006161382888159342,
|
| 9709 |
+
"grad_norm": 10.625,
|
| 9710 |
+
"learning_rate": 3.078496093532936e-07,
|
| 9711 |
+
"loss": 2.8458,
|
| 9712 |
+
"step": 1386
|
| 9713 |
+
},
|
| 9714 |
+
{
|
| 9715 |
+
"epoch": 0.006165828330358592,
|
| 9716 |
+
"grad_norm": 11.0625,
|
| 9717 |
+
"learning_rate": 3.0807188343946926e-07,
|
| 9718 |
+
"loss": 2.915,
|
| 9719 |
+
"step": 1387
|
| 9720 |
+
},
|
| 9721 |
+
{
|
| 9722 |
+
"epoch": 0.006170273772557841,
|
| 9723 |
+
"grad_norm": 10.5,
|
| 9724 |
+
"learning_rate": 3.0829415752564494e-07,
|
| 9725 |
+
"loss": 3.0639,
|
| 9726 |
+
"step": 1388
|
| 9727 |
+
},
|
| 9728 |
+
{
|
| 9729 |
+
"epoch": 0.00617471921475709,
|
| 9730 |
+
"grad_norm": 8.5,
|
| 9731 |
+
"learning_rate": 3.0851643161182056e-07,
|
| 9732 |
+
"loss": 3.1389,
|
| 9733 |
+
"step": 1389
|
| 9734 |
+
},
|
| 9735 |
+
{
|
| 9736 |
+
"epoch": 0.006179164656956339,
|
| 9737 |
+
"grad_norm": 9.625,
|
| 9738 |
+
"learning_rate": 3.087387056979962e-07,
|
| 9739 |
+
"loss": 3.3029,
|
| 9740 |
+
"step": 1390
|
| 9741 |
+
},
|
| 9742 |
+
{
|
| 9743 |
+
"epoch": 0.006183610099155588,
|
| 9744 |
+
"grad_norm": 9.6875,
|
| 9745 |
+
"learning_rate": 3.0896097978417187e-07,
|
| 9746 |
+
"loss": 3.0344,
|
| 9747 |
+
"step": 1391
|
| 9748 |
+
},
|
| 9749 |
+
{
|
| 9750 |
+
"epoch": 0.006188055541354837,
|
| 9751 |
+
"grad_norm": 11.5625,
|
| 9752 |
+
"learning_rate": 3.0918325387034754e-07,
|
| 9753 |
+
"loss": 2.8049,
|
| 9754 |
+
"step": 1392
|
| 9755 |
+
},
|
| 9756 |
+
{
|
| 9757 |
+
"epoch": 0.006192500983554086,
|
| 9758 |
+
"grad_norm": 10.6875,
|
| 9759 |
+
"learning_rate": 3.094055279565232e-07,
|
| 9760 |
+
"loss": 2.912,
|
| 9761 |
+
"step": 1393
|
| 9762 |
+
},
|
| 9763 |
+
{
|
| 9764 |
+
"epoch": 0.006196946425753335,
|
| 9765 |
+
"grad_norm": 10.75,
|
| 9766 |
+
"learning_rate": 3.0962780204269885e-07,
|
| 9767 |
+
"loss": 2.9791,
|
| 9768 |
+
"step": 1394
|
| 9769 |
+
},
|
| 9770 |
+
{
|
| 9771 |
+
"epoch": 0.006201391867952585,
|
| 9772 |
+
"grad_norm": 9.4375,
|
| 9773 |
+
"learning_rate": 3.098500761288745e-07,
|
| 9774 |
+
"loss": 2.98,
|
| 9775 |
+
"step": 1395
|
| 9776 |
+
},
|
| 9777 |
+
{
|
| 9778 |
+
"epoch": 0.006205837310151834,
|
| 9779 |
+
"grad_norm": 11.5625,
|
| 9780 |
+
"learning_rate": 3.100723502150502e-07,
|
| 9781 |
+
"loss": 2.9625,
|
| 9782 |
+
"step": 1396
|
| 9783 |
+
},
|
| 9784 |
+
{
|
| 9785 |
+
"epoch": 0.0062102827523510835,
|
| 9786 |
+
"grad_norm": 10.3125,
|
| 9787 |
+
"learning_rate": 3.102946243012259e-07,
|
| 9788 |
+
"loss": 2.9074,
|
| 9789 |
+
"step": 1397
|
| 9790 |
+
},
|
| 9791 |
+
{
|
| 9792 |
+
"epoch": 0.0062147281945503325,
|
| 9793 |
+
"grad_norm": 8.9375,
|
| 9794 |
+
"learning_rate": 3.1051689838740156e-07,
|
| 9795 |
+
"loss": 3.0771,
|
| 9796 |
+
"step": 1398
|
| 9797 |
+
},
|
| 9798 |
+
{
|
| 9799 |
+
"epoch": 0.006219173636749582,
|
| 9800 |
+
"grad_norm": 8.4375,
|
| 9801 |
+
"learning_rate": 3.107391724735772e-07,
|
| 9802 |
+
"loss": 3.1583,
|
| 9803 |
+
"step": 1399
|
| 9804 |
+
},
|
| 9805 |
+
{
|
| 9806 |
+
"epoch": 0.006223619078948831,
|
| 9807 |
+
"grad_norm": 8.4375,
|
| 9808 |
+
"learning_rate": 3.1096144655975286e-07,
|
| 9809 |
+
"loss": 3.1964,
|
| 9810 |
+
"step": 1400
|
| 9811 |
+
},
|
| 9812 |
+
{
|
| 9813 |
+
"epoch": 0.00622806452114808,
|
| 9814 |
+
"grad_norm": 12.4375,
|
| 9815 |
+
"learning_rate": 3.1118372064592854e-07,
|
| 9816 |
+
"loss": 2.7624,
|
| 9817 |
+
"step": 1401
|
| 9818 |
+
},
|
| 9819 |
+
{
|
| 9820 |
+
"epoch": 0.006232509963347329,
|
| 9821 |
+
"grad_norm": 9.0,
|
| 9822 |
+
"learning_rate": 3.114059947321042e-07,
|
| 9823 |
+
"loss": 3.0169,
|
| 9824 |
+
"step": 1402
|
| 9825 |
+
},
|
| 9826 |
+
{
|
| 9827 |
+
"epoch": 0.006236955405546578,
|
| 9828 |
+
"grad_norm": 8.875,
|
| 9829 |
+
"learning_rate": 3.1162826881827984e-07,
|
| 9830 |
+
"loss": 3.0511,
|
| 9831 |
+
"step": 1403
|
| 9832 |
+
},
|
| 9833 |
+
{
|
| 9834 |
+
"epoch": 0.006241400847745828,
|
| 9835 |
+
"grad_norm": 8.25,
|
| 9836 |
+
"learning_rate": 3.118505429044555e-07,
|
| 9837 |
+
"loss": 3.3416,
|
| 9838 |
+
"step": 1404
|
| 9839 |
+
},
|
| 9840 |
+
{
|
| 9841 |
+
"epoch": 0.006245846289945077,
|
| 9842 |
+
"grad_norm": 7.375,
|
| 9843 |
+
"learning_rate": 3.120728169906312e-07,
|
| 9844 |
+
"loss": 3.1489,
|
| 9845 |
+
"step": 1405
|
| 9846 |
+
},
|
| 9847 |
+
{
|
| 9848 |
+
"epoch": 0.006250291732144326,
|
| 9849 |
+
"grad_norm": 9.0,
|
| 9850 |
+
"learning_rate": 3.122950910768068e-07,
|
| 9851 |
+
"loss": 3.1389,
|
| 9852 |
+
"step": 1406
|
| 9853 |
+
},
|
| 9854 |
+
{
|
| 9855 |
+
"epoch": 0.006254737174343575,
|
| 9856 |
+
"grad_norm": 8.6875,
|
| 9857 |
+
"learning_rate": 3.125173651629825e-07,
|
| 9858 |
+
"loss": 3.1529,
|
| 9859 |
+
"step": 1407
|
| 9860 |
+
},
|
| 9861 |
+
{
|
| 9862 |
+
"epoch": 0.006259182616542824,
|
| 9863 |
+
"grad_norm": 8.25,
|
| 9864 |
+
"learning_rate": 3.127396392491582e-07,
|
| 9865 |
+
"loss": 3.3158,
|
| 9866 |
+
"step": 1408
|
| 9867 |
+
},
|
| 9868 |
+
{
|
| 9869 |
+
"epoch": 0.006263628058742073,
|
| 9870 |
+
"grad_norm": 10.1875,
|
| 9871 |
+
"learning_rate": 3.1296191333533385e-07,
|
| 9872 |
+
"loss": 3.0265,
|
| 9873 |
+
"step": 1409
|
| 9874 |
+
},
|
| 9875 |
+
{
|
| 9876 |
+
"epoch": 0.006268073500941322,
|
| 9877 |
+
"grad_norm": 8.0625,
|
| 9878 |
+
"learning_rate": 3.131841874215095e-07,
|
| 9879 |
+
"loss": 3.1497,
|
| 9880 |
+
"step": 1410
|
| 9881 |
+
},
|
| 9882 |
+
{
|
| 9883 |
+
"epoch": 0.006272518943140571,
|
| 9884 |
+
"grad_norm": 9.9375,
|
| 9885 |
+
"learning_rate": 3.1340646150768516e-07,
|
| 9886 |
+
"loss": 2.8999,
|
| 9887 |
+
"step": 1411
|
| 9888 |
+
},
|
| 9889 |
+
{
|
| 9890 |
+
"epoch": 0.006276964385339821,
|
| 9891 |
+
"grad_norm": 9.625,
|
| 9892 |
+
"learning_rate": 3.1362873559386083e-07,
|
| 9893 |
+
"loss": 3.1192,
|
| 9894 |
+
"step": 1412
|
| 9895 |
+
},
|
| 9896 |
+
{
|
| 9897 |
+
"epoch": 0.00628140982753907,
|
| 9898 |
+
"grad_norm": 11.25,
|
| 9899 |
+
"learning_rate": 3.138510096800365e-07,
|
| 9900 |
+
"loss": 2.8486,
|
| 9901 |
+
"step": 1413
|
| 9902 |
+
},
|
| 9903 |
+
{
|
| 9904 |
+
"epoch": 0.006285855269738319,
|
| 9905 |
+
"grad_norm": 7.96875,
|
| 9906 |
+
"learning_rate": 3.1407328376621214e-07,
|
| 9907 |
+
"loss": 3.3656,
|
| 9908 |
+
"step": 1414
|
| 9909 |
+
},
|
| 9910 |
+
{
|
| 9911 |
+
"epoch": 0.006290300711937568,
|
| 9912 |
+
"grad_norm": 9.5,
|
| 9913 |
+
"learning_rate": 3.1429555785238776e-07,
|
| 9914 |
+
"loss": 3.0808,
|
| 9915 |
+
"step": 1415
|
| 9916 |
+
},
|
| 9917 |
+
{
|
| 9918 |
+
"epoch": 0.006294746154136817,
|
| 9919 |
+
"grad_norm": 11.375,
|
| 9920 |
+
"learning_rate": 3.145178319385635e-07,
|
| 9921 |
+
"loss": 2.8574,
|
| 9922 |
+
"step": 1416
|
| 9923 |
+
},
|
| 9924 |
+
{
|
| 9925 |
+
"epoch": 0.0062991915963360665,
|
| 9926 |
+
"grad_norm": 6.75,
|
| 9927 |
+
"learning_rate": 3.147401060247391e-07,
|
| 9928 |
+
"loss": 3.3653,
|
| 9929 |
+
"step": 1417
|
| 9930 |
+
},
|
| 9931 |
+
{
|
| 9932 |
+
"epoch": 0.0063036370385353155,
|
| 9933 |
+
"grad_norm": 10.0625,
|
| 9934 |
+
"learning_rate": 3.149623801109148e-07,
|
| 9935 |
+
"loss": 2.9331,
|
| 9936 |
+
"step": 1418
|
| 9937 |
+
},
|
| 9938 |
+
{
|
| 9939 |
+
"epoch": 0.006308082480734565,
|
| 9940 |
+
"grad_norm": 11.75,
|
| 9941 |
+
"learning_rate": 3.151846541970904e-07,
|
| 9942 |
+
"loss": 2.87,
|
| 9943 |
+
"step": 1419
|
| 9944 |
+
},
|
| 9945 |
+
{
|
| 9946 |
+
"epoch": 0.006312527922933814,
|
| 9947 |
+
"grad_norm": 8.625,
|
| 9948 |
+
"learning_rate": 3.1540692828326615e-07,
|
| 9949 |
+
"loss": 3.054,
|
| 9950 |
+
"step": 1420
|
| 9951 |
+
},
|
| 9952 |
+
{
|
| 9953 |
+
"epoch": 0.006316973365133064,
|
| 9954 |
+
"grad_norm": 10.4375,
|
| 9955 |
+
"learning_rate": 3.156292023694418e-07,
|
| 9956 |
+
"loss": 2.9734,
|
| 9957 |
+
"step": 1421
|
| 9958 |
+
},
|
| 9959 |
+
{
|
| 9960 |
+
"epoch": 0.006321418807332313,
|
| 9961 |
+
"grad_norm": 9.8125,
|
| 9962 |
+
"learning_rate": 3.1585147645561745e-07,
|
| 9963 |
+
"loss": 3.0222,
|
| 9964 |
+
"step": 1422
|
| 9965 |
+
},
|
| 9966 |
+
{
|
| 9967 |
+
"epoch": 0.006325864249531562,
|
| 9968 |
+
"grad_norm": 8.75,
|
| 9969 |
+
"learning_rate": 3.160737505417931e-07,
|
| 9970 |
+
"loss": 3.1698,
|
| 9971 |
+
"step": 1423
|
| 9972 |
+
},
|
| 9973 |
+
{
|
| 9974 |
+
"epoch": 0.006330309691730811,
|
| 9975 |
+
"grad_norm": 10.3125,
|
| 9976 |
+
"learning_rate": 3.162960246279688e-07,
|
| 9977 |
+
"loss": 3.1181,
|
| 9978 |
+
"step": 1424
|
| 9979 |
+
},
|
| 9980 |
+
{
|
| 9981 |
+
"epoch": 0.00633475513393006,
|
| 9982 |
+
"grad_norm": 11.8125,
|
| 9983 |
+
"learning_rate": 3.1651829871414443e-07,
|
| 9984 |
+
"loss": 2.8102,
|
| 9985 |
+
"step": 1425
|
| 9986 |
+
},
|
| 9987 |
+
{
|
| 9988 |
+
"epoch": 0.006339200576129309,
|
| 9989 |
+
"grad_norm": 9.375,
|
| 9990 |
+
"learning_rate": 3.167405728003201e-07,
|
| 9991 |
+
"loss": 3.0335,
|
| 9992 |
+
"step": 1426
|
| 9993 |
+
},
|
| 9994 |
+
{
|
| 9995 |
+
"epoch": 0.006343646018328558,
|
| 9996 |
+
"grad_norm": 8.8125,
|
| 9997 |
+
"learning_rate": 3.1696284688649574e-07,
|
| 9998 |
+
"loss": 3.0745,
|
| 9999 |
+
"step": 1427
|
| 10000 |
+
},
|
| 10001 |
+
{
|
| 10002 |
+
"epoch": 0.006348091460527807,
|
| 10003 |
+
"grad_norm": 7.0625,
|
| 10004 |
+
"learning_rate": 3.1718512097267147e-07,
|
| 10005 |
+
"loss": 3.4504,
|
| 10006 |
+
"step": 1428
|
| 10007 |
+
},
|
| 10008 |
+
{
|
| 10009 |
+
"epoch": 0.006352536902727057,
|
| 10010 |
+
"grad_norm": 10.0,
|
| 10011 |
+
"learning_rate": 3.174073950588471e-07,
|
| 10012 |
+
"loss": 2.9273,
|
| 10013 |
+
"step": 1429
|
| 10014 |
+
},
|
| 10015 |
+
{
|
| 10016 |
+
"epoch": 0.006356982344926306,
|
| 10017 |
+
"grad_norm": 9.25,
|
| 10018 |
+
"learning_rate": 3.1762966914502277e-07,
|
| 10019 |
+
"loss": 2.9806,
|
| 10020 |
+
"step": 1430
|
| 10021 |
+
},
|
| 10022 |
+
{
|
| 10023 |
+
"epoch": 0.006361427787125555,
|
| 10024 |
+
"grad_norm": 8.1875,
|
| 10025 |
+
"learning_rate": 3.178519432311984e-07,
|
| 10026 |
+
"loss": 3.1518,
|
| 10027 |
+
"step": 1431
|
| 10028 |
+
},
|
| 10029 |
+
{
|
| 10030 |
+
"epoch": 0.006365873229324804,
|
| 10031 |
+
"grad_norm": 9.25,
|
| 10032 |
+
"learning_rate": 3.1807421731737407e-07,
|
| 10033 |
+
"loss": 3.1738,
|
| 10034 |
+
"step": 1432
|
| 10035 |
+
},
|
| 10036 |
+
{
|
| 10037 |
+
"epoch": 0.006370318671524053,
|
| 10038 |
+
"grad_norm": 10.9375,
|
| 10039 |
+
"learning_rate": 3.1829649140354975e-07,
|
| 10040 |
+
"loss": 2.9154,
|
| 10041 |
+
"step": 1433
|
| 10042 |
+
},
|
| 10043 |
+
{
|
| 10044 |
+
"epoch": 0.006374764113723302,
|
| 10045 |
+
"grad_norm": 9.125,
|
| 10046 |
+
"learning_rate": 3.185187654897254e-07,
|
| 10047 |
+
"loss": 3.1293,
|
| 10048 |
+
"step": 1434
|
| 10049 |
+
},
|
| 10050 |
+
{
|
| 10051 |
+
"epoch": 0.006379209555922551,
|
| 10052 |
+
"grad_norm": 11.9375,
|
| 10053 |
+
"learning_rate": 3.1874103957590105e-07,
|
| 10054 |
+
"loss": 2.8604,
|
| 10055 |
+
"step": 1435
|
| 10056 |
+
},
|
| 10057 |
+
{
|
| 10058 |
+
"epoch": 0.0063836549981218,
|
| 10059 |
+
"grad_norm": 9.8125,
|
| 10060 |
+
"learning_rate": 3.1896331366207673e-07,
|
| 10061 |
+
"loss": 3.0019,
|
| 10062 |
+
"step": 1436
|
| 10063 |
+
},
|
| 10064 |
+
{
|
| 10065 |
+
"epoch": 0.0063881004403210494,
|
| 10066 |
+
"grad_norm": 7.5625,
|
| 10067 |
+
"learning_rate": 3.191855877482524e-07,
|
| 10068 |
+
"loss": 3.3028,
|
| 10069 |
+
"step": 1437
|
| 10070 |
+
},
|
| 10071 |
+
{
|
| 10072 |
+
"epoch": 0.006392545882520299,
|
| 10073 |
+
"grad_norm": 9.875,
|
| 10074 |
+
"learning_rate": 3.1940786183442803e-07,
|
| 10075 |
+
"loss": 3.109,
|
| 10076 |
+
"step": 1438
|
| 10077 |
+
},
|
| 10078 |
+
{
|
| 10079 |
+
"epoch": 0.0063969913247195484,
|
| 10080 |
+
"grad_norm": 10.4375,
|
| 10081 |
+
"learning_rate": 3.1963013592060376e-07,
|
| 10082 |
+
"loss": 2.8731,
|
| 10083 |
+
"step": 1439
|
| 10084 |
+
},
|
| 10085 |
+
{
|
| 10086 |
+
"epoch": 0.0064014367669187975,
|
| 10087 |
+
"grad_norm": 10.5625,
|
| 10088 |
+
"learning_rate": 3.198524100067794e-07,
|
| 10089 |
+
"loss": 2.9233,
|
| 10090 |
+
"step": 1440
|
| 10091 |
+
},
|
| 10092 |
+
{
|
| 10093 |
+
"epoch": 0.0064058822091180466,
|
| 10094 |
+
"grad_norm": 8.9375,
|
| 10095 |
+
"learning_rate": 3.2007468409295507e-07,
|
| 10096 |
+
"loss": 3.174,
|
| 10097 |
+
"step": 1441
|
| 10098 |
+
},
|
| 10099 |
+
{
|
| 10100 |
+
"epoch": 0.006410327651317296,
|
| 10101 |
+
"grad_norm": 11.75,
|
| 10102 |
+
"learning_rate": 3.202969581791307e-07,
|
| 10103 |
+
"loss": 2.8942,
|
| 10104 |
+
"step": 1442
|
| 10105 |
+
},
|
| 10106 |
+
{
|
| 10107 |
+
"epoch": 0.006414773093516545,
|
| 10108 |
+
"grad_norm": 6.46875,
|
| 10109 |
+
"learning_rate": 3.205192322653064e-07,
|
| 10110 |
+
"loss": 3.3406,
|
| 10111 |
+
"step": 1443
|
| 10112 |
+
},
|
| 10113 |
+
{
|
| 10114 |
+
"epoch": 0.006419218535715794,
|
| 10115 |
+
"grad_norm": 9.9375,
|
| 10116 |
+
"learning_rate": 3.2074150635148205e-07,
|
| 10117 |
+
"loss": 2.9487,
|
| 10118 |
+
"step": 1444
|
| 10119 |
+
},
|
| 10120 |
+
{
|
| 10121 |
+
"epoch": 0.006423663977915043,
|
| 10122 |
+
"grad_norm": 11.4375,
|
| 10123 |
+
"learning_rate": 3.209637804376577e-07,
|
| 10124 |
+
"loss": 2.8494,
|
| 10125 |
+
"step": 1445
|
| 10126 |
+
},
|
| 10127 |
+
{
|
| 10128 |
+
"epoch": 0.006428109420114293,
|
| 10129 |
+
"grad_norm": 7.65625,
|
| 10130 |
+
"learning_rate": 3.2118605452383335e-07,
|
| 10131 |
+
"loss": 3.29,
|
| 10132 |
+
"step": 1446
|
| 10133 |
+
},
|
| 10134 |
+
{
|
| 10135 |
+
"epoch": 0.006432554862313542,
|
| 10136 |
+
"grad_norm": 10.3125,
|
| 10137 |
+
"learning_rate": 3.214083286100091e-07,
|
| 10138 |
+
"loss": 2.9935,
|
| 10139 |
+
"step": 1447
|
| 10140 |
+
},
|
| 10141 |
+
{
|
| 10142 |
+
"epoch": 0.006437000304512791,
|
| 10143 |
+
"grad_norm": 8.9375,
|
| 10144 |
+
"learning_rate": 3.216306026961847e-07,
|
| 10145 |
+
"loss": 3.0299,
|
| 10146 |
+
"step": 1448
|
| 10147 |
+
},
|
| 10148 |
+
{
|
| 10149 |
+
"epoch": 0.00644144574671204,
|
| 10150 |
+
"grad_norm": 9.1875,
|
| 10151 |
+
"learning_rate": 3.2185287678236033e-07,
|
| 10152 |
+
"loss": 3.094,
|
| 10153 |
+
"step": 1449
|
| 10154 |
+
},
|
| 10155 |
+
{
|
| 10156 |
+
"epoch": 0.006445891188911289,
|
| 10157 |
+
"grad_norm": 9.75,
|
| 10158 |
+
"learning_rate": 3.22075150868536e-07,
|
| 10159 |
+
"loss": 3.1475,
|
| 10160 |
+
"step": 1450
|
| 10161 |
+
},
|
| 10162 |
+
{
|
| 10163 |
+
"epoch": 0.006450336631110538,
|
| 10164 |
+
"grad_norm": 6.9375,
|
| 10165 |
+
"learning_rate": 3.2229742495471163e-07,
|
| 10166 |
+
"loss": 3.4116,
|
| 10167 |
+
"step": 1451
|
| 10168 |
+
},
|
| 10169 |
+
{
|
| 10170 |
+
"epoch": 0.006454782073309787,
|
| 10171 |
+
"grad_norm": 12.125,
|
| 10172 |
+
"learning_rate": 3.2251969904088736e-07,
|
| 10173 |
+
"loss": 2.7524,
|
| 10174 |
+
"step": 1452
|
| 10175 |
+
},
|
| 10176 |
+
{
|
| 10177 |
+
"epoch": 0.006459227515509036,
|
| 10178 |
+
"grad_norm": 9.5625,
|
| 10179 |
+
"learning_rate": 3.22741973127063e-07,
|
| 10180 |
+
"loss": 2.9845,
|
| 10181 |
+
"step": 1453
|
| 10182 |
+
},
|
| 10183 |
+
{
|
| 10184 |
+
"epoch": 0.006463672957708285,
|
| 10185 |
+
"grad_norm": 8.8125,
|
| 10186 |
+
"learning_rate": 3.2296424721323867e-07,
|
| 10187 |
+
"loss": 3.3037,
|
| 10188 |
+
"step": 1454
|
| 10189 |
+
},
|
| 10190 |
+
{
|
| 10191 |
+
"epoch": 0.006468118399907535,
|
| 10192 |
+
"grad_norm": 10.9375,
|
| 10193 |
+
"learning_rate": 3.231865212994143e-07,
|
| 10194 |
+
"loss": 2.8459,
|
| 10195 |
+
"step": 1455
|
| 10196 |
+
},
|
| 10197 |
+
{
|
| 10198 |
+
"epoch": 0.006472563842106784,
|
| 10199 |
+
"grad_norm": 9.3125,
|
| 10200 |
+
"learning_rate": 3.2340879538559e-07,
|
| 10201 |
+
"loss": 3.1079,
|
| 10202 |
+
"step": 1456
|
| 10203 |
+
},
|
| 10204 |
+
{
|
| 10205 |
+
"epoch": 0.006477009284306033,
|
| 10206 |
+
"grad_norm": 10.3125,
|
| 10207 |
+
"learning_rate": 3.2363106947176565e-07,
|
| 10208 |
+
"loss": 2.9778,
|
| 10209 |
+
"step": 1457
|
| 10210 |
+
},
|
| 10211 |
+
{
|
| 10212 |
+
"epoch": 0.006481454726505282,
|
| 10213 |
+
"grad_norm": 10.4375,
|
| 10214 |
+
"learning_rate": 3.238533435579413e-07,
|
| 10215 |
+
"loss": 2.9249,
|
| 10216 |
+
"step": 1458
|
| 10217 |
+
},
|
| 10218 |
+
{
|
| 10219 |
+
"epoch": 0.006485900168704531,
|
| 10220 |
+
"grad_norm": 8.1875,
|
| 10221 |
+
"learning_rate": 3.24075617644117e-07,
|
| 10222 |
+
"loss": 3.2607,
|
| 10223 |
+
"step": 1459
|
| 10224 |
+
},
|
| 10225 |
+
{
|
| 10226 |
+
"epoch": 0.0064903456109037805,
|
| 10227 |
+
"grad_norm": 11.625,
|
| 10228 |
+
"learning_rate": 3.242978917302927e-07,
|
| 10229 |
+
"loss": 2.9025,
|
| 10230 |
+
"step": 1460
|
| 10231 |
+
},
|
| 10232 |
+
{
|
| 10233 |
+
"epoch": 0.0064947910531030295,
|
| 10234 |
+
"grad_norm": 10.0625,
|
| 10235 |
+
"learning_rate": 3.245201658164683e-07,
|
| 10236 |
+
"loss": 3.0274,
|
| 10237 |
+
"step": 1461
|
| 10238 |
+
},
|
| 10239 |
+
{
|
| 10240 |
+
"epoch": 0.006499236495302279,
|
| 10241 |
+
"grad_norm": 9.3125,
|
| 10242 |
+
"learning_rate": 3.24742439902644e-07,
|
| 10243 |
+
"loss": 3.0148,
|
| 10244 |
+
"step": 1462
|
| 10245 |
+
},
|
| 10246 |
+
{
|
| 10247 |
+
"epoch": 0.0065036819375015285,
|
| 10248 |
+
"grad_norm": 8.0625,
|
| 10249 |
+
"learning_rate": 3.2496471398881966e-07,
|
| 10250 |
+
"loss": 3.232,
|
| 10251 |
+
"step": 1463
|
| 10252 |
+
},
|
| 10253 |
+
{
|
| 10254 |
+
"epoch": 0.006508127379700778,
|
| 10255 |
+
"grad_norm": 8.75,
|
| 10256 |
+
"learning_rate": 3.251869880749953e-07,
|
| 10257 |
+
"loss": 3.2344,
|
| 10258 |
+
"step": 1464
|
| 10259 |
+
},
|
| 10260 |
+
{
|
| 10261 |
+
"epoch": 0.006512572821900027,
|
| 10262 |
+
"grad_norm": 11.375,
|
| 10263 |
+
"learning_rate": 3.2540926216117096e-07,
|
| 10264 |
+
"loss": 2.8236,
|
| 10265 |
+
"step": 1465
|
| 10266 |
+
},
|
| 10267 |
+
{
|
| 10268 |
+
"epoch": 0.006517018264099276,
|
| 10269 |
+
"grad_norm": 7.84375,
|
| 10270 |
+
"learning_rate": 3.256315362473466e-07,
|
| 10271 |
+
"loss": 3.5294,
|
| 10272 |
+
"step": 1466
|
| 10273 |
+
},
|
| 10274 |
+
{
|
| 10275 |
+
"epoch": 0.006521463706298525,
|
| 10276 |
+
"grad_norm": 7.34375,
|
| 10277 |
+
"learning_rate": 3.258538103335223e-07,
|
| 10278 |
+
"loss": 3.533,
|
| 10279 |
+
"step": 1467
|
| 10280 |
+
},
|
| 10281 |
+
{
|
| 10282 |
+
"epoch": 0.006525909148497774,
|
| 10283 |
+
"grad_norm": 10.875,
|
| 10284 |
+
"learning_rate": 3.2607608441969794e-07,
|
| 10285 |
+
"loss": 2.8747,
|
| 10286 |
+
"step": 1468
|
| 10287 |
+
},
|
| 10288 |
+
{
|
| 10289 |
+
"epoch": 0.006530354590697023,
|
| 10290 |
+
"grad_norm": 10.75,
|
| 10291 |
+
"learning_rate": 3.262983585058736e-07,
|
| 10292 |
+
"loss": 2.8983,
|
| 10293 |
+
"step": 1469
|
| 10294 |
+
},
|
| 10295 |
+
{
|
| 10296 |
+
"epoch": 0.006534800032896272,
|
| 10297 |
+
"grad_norm": 11.875,
|
| 10298 |
+
"learning_rate": 3.2652063259204925e-07,
|
| 10299 |
+
"loss": 2.7522,
|
| 10300 |
+
"step": 1470
|
| 10301 |
+
},
|
| 10302 |
+
{
|
| 10303 |
+
"epoch": 0.006539245475095521,
|
| 10304 |
+
"grad_norm": 9.625,
|
| 10305 |
+
"learning_rate": 3.26742906678225e-07,
|
| 10306 |
+
"loss": 3.1023,
|
| 10307 |
+
"step": 1471
|
| 10308 |
+
},
|
| 10309 |
+
{
|
| 10310 |
+
"epoch": 0.006543690917294771,
|
| 10311 |
+
"grad_norm": 9.6875,
|
| 10312 |
+
"learning_rate": 3.269651807644006e-07,
|
| 10313 |
+
"loss": 3.0412,
|
| 10314 |
+
"step": 1472
|
| 10315 |
+
},
|
| 10316 |
+
{
|
| 10317 |
+
"epoch": 0.00654813635949402,
|
| 10318 |
+
"grad_norm": 10.25,
|
| 10319 |
+
"learning_rate": 3.271874548505763e-07,
|
| 10320 |
+
"loss": 2.9,
|
| 10321 |
+
"step": 1473
|
| 10322 |
+
},
|
| 10323 |
+
{
|
| 10324 |
+
"epoch": 0.006552581801693269,
|
| 10325 |
+
"grad_norm": 9.3125,
|
| 10326 |
+
"learning_rate": 3.274097289367519e-07,
|
| 10327 |
+
"loss": 3.0553,
|
| 10328 |
+
"step": 1474
|
| 10329 |
+
},
|
| 10330 |
+
{
|
| 10331 |
+
"epoch": 0.006557027243892518,
|
| 10332 |
+
"grad_norm": 11.0,
|
| 10333 |
+
"learning_rate": 3.2763200302292764e-07,
|
| 10334 |
+
"loss": 3.0155,
|
| 10335 |
+
"step": 1475
|
| 10336 |
+
},
|
| 10337 |
+
{
|
| 10338 |
+
"epoch": 0.006561472686091767,
|
| 10339 |
+
"grad_norm": 11.1875,
|
| 10340 |
+
"learning_rate": 3.2785427710910326e-07,
|
| 10341 |
+
"loss": 2.9113,
|
| 10342 |
+
"step": 1476
|
| 10343 |
+
},
|
| 10344 |
+
{
|
| 10345 |
+
"epoch": 0.006565918128291016,
|
| 10346 |
+
"grad_norm": 9.6875,
|
| 10347 |
+
"learning_rate": 3.2807655119527894e-07,
|
| 10348 |
+
"loss": 2.9571,
|
| 10349 |
+
"step": 1477
|
| 10350 |
+
},
|
| 10351 |
+
{
|
| 10352 |
+
"epoch": 0.006570363570490265,
|
| 10353 |
+
"grad_norm": 9.9375,
|
| 10354 |
+
"learning_rate": 3.2829882528145456e-07,
|
| 10355 |
+
"loss": 3.1552,
|
| 10356 |
+
"step": 1478
|
| 10357 |
+
},
|
| 10358 |
+
{
|
| 10359 |
+
"epoch": 0.006574809012689514,
|
| 10360 |
+
"grad_norm": 7.4375,
|
| 10361 |
+
"learning_rate": 3.285210993676303e-07,
|
| 10362 |
+
"loss": 3.3457,
|
| 10363 |
+
"step": 1479
|
| 10364 |
+
},
|
| 10365 |
+
{
|
| 10366 |
+
"epoch": 0.006579254454888764,
|
| 10367 |
+
"grad_norm": 11.6875,
|
| 10368 |
+
"learning_rate": 3.287433734538059e-07,
|
| 10369 |
+
"loss": 2.91,
|
| 10370 |
+
"step": 1480
|
| 10371 |
+
},
|
| 10372 |
+
{
|
| 10373 |
+
"epoch": 0.006583699897088013,
|
| 10374 |
+
"grad_norm": 8.25,
|
| 10375 |
+
"learning_rate": 3.2896564753998154e-07,
|
| 10376 |
+
"loss": 3.1658,
|
| 10377 |
+
"step": 1481
|
| 10378 |
+
},
|
| 10379 |
+
{
|
| 10380 |
+
"epoch": 0.0065881453392872625,
|
| 10381 |
+
"grad_norm": 10.375,
|
| 10382 |
+
"learning_rate": 3.291879216261572e-07,
|
| 10383 |
+
"loss": 2.9511,
|
| 10384 |
+
"step": 1482
|
| 10385 |
+
},
|
| 10386 |
+
{
|
| 10387 |
+
"epoch": 0.0065925907814865115,
|
| 10388 |
+
"grad_norm": 9.8125,
|
| 10389 |
+
"learning_rate": 3.294101957123329e-07,
|
| 10390 |
+
"loss": 2.9954,
|
| 10391 |
+
"step": 1483
|
| 10392 |
+
},
|
| 10393 |
+
{
|
| 10394 |
+
"epoch": 0.006597036223685761,
|
| 10395 |
+
"grad_norm": 10.375,
|
| 10396 |
+
"learning_rate": 3.296324697985086e-07,
|
| 10397 |
+
"loss": 2.8514,
|
| 10398 |
+
"step": 1484
|
| 10399 |
+
},
|
| 10400 |
+
{
|
| 10401 |
+
"epoch": 0.00660148166588501,
|
| 10402 |
+
"grad_norm": 8.9375,
|
| 10403 |
+
"learning_rate": 3.298547438846842e-07,
|
| 10404 |
+
"loss": 3.0513,
|
| 10405 |
+
"step": 1485
|
| 10406 |
+
},
|
| 10407 |
+
{
|
| 10408 |
+
"epoch": 0.006605927108084259,
|
| 10409 |
+
"grad_norm": 5.59375,
|
| 10410 |
+
"learning_rate": 3.300770179708599e-07,
|
| 10411 |
+
"loss": 3.5142,
|
| 10412 |
+
"step": 1486
|
| 10413 |
+
},
|
| 10414 |
+
{
|
| 10415 |
+
"epoch": 0.006610372550283508,
|
| 10416 |
+
"grad_norm": 9.0625,
|
| 10417 |
+
"learning_rate": 3.3029929205703556e-07,
|
| 10418 |
+
"loss": 3.1915,
|
| 10419 |
+
"step": 1487
|
| 10420 |
+
},
|
| 10421 |
+
{
|
| 10422 |
+
"epoch": 0.006614817992482757,
|
| 10423 |
+
"grad_norm": 8.25,
|
| 10424 |
+
"learning_rate": 3.3052156614321124e-07,
|
| 10425 |
+
"loss": 3.1742,
|
| 10426 |
+
"step": 1488
|
| 10427 |
+
},
|
| 10428 |
+
{
|
| 10429 |
+
"epoch": 0.006619263434682007,
|
| 10430 |
+
"grad_norm": 8.125,
|
| 10431 |
+
"learning_rate": 3.3074384022938686e-07,
|
| 10432 |
+
"loss": 3.2412,
|
| 10433 |
+
"step": 1489
|
| 10434 |
+
},
|
| 10435 |
+
{
|
| 10436 |
+
"epoch": 0.006623708876881256,
|
| 10437 |
+
"grad_norm": 9.9375,
|
| 10438 |
+
"learning_rate": 3.309661143155626e-07,
|
| 10439 |
+
"loss": 3.0882,
|
| 10440 |
+
"step": 1490
|
| 10441 |
+
},
|
| 10442 |
+
{
|
| 10443 |
+
"epoch": 0.006628154319080505,
|
| 10444 |
+
"grad_norm": 10.4375,
|
| 10445 |
+
"learning_rate": 3.311883884017382e-07,
|
| 10446 |
+
"loss": 2.9808,
|
| 10447 |
+
"step": 1491
|
| 10448 |
+
},
|
| 10449 |
+
{
|
| 10450 |
+
"epoch": 0.006632599761279754,
|
| 10451 |
+
"grad_norm": 11.1875,
|
| 10452 |
+
"learning_rate": 3.314106624879139e-07,
|
| 10453 |
+
"loss": 2.9024,
|
| 10454 |
+
"step": 1492
|
| 10455 |
+
},
|
| 10456 |
+
{
|
| 10457 |
+
"epoch": 0.006637045203479003,
|
| 10458 |
+
"grad_norm": 11.25,
|
| 10459 |
+
"learning_rate": 3.316329365740895e-07,
|
| 10460 |
+
"loss": 2.8403,
|
| 10461 |
+
"step": 1493
|
| 10462 |
+
},
|
| 10463 |
+
{
|
| 10464 |
+
"epoch": 0.006641490645678252,
|
| 10465 |
+
"grad_norm": 10.375,
|
| 10466 |
+
"learning_rate": 3.3185521066026525e-07,
|
| 10467 |
+
"loss": 2.984,
|
| 10468 |
+
"step": 1494
|
| 10469 |
+
},
|
| 10470 |
+
{
|
| 10471 |
+
"epoch": 0.006645936087877501,
|
| 10472 |
+
"grad_norm": 10.0625,
|
| 10473 |
+
"learning_rate": 3.320774847464409e-07,
|
| 10474 |
+
"loss": 3.0457,
|
| 10475 |
+
"step": 1495
|
| 10476 |
+
},
|
| 10477 |
+
{
|
| 10478 |
+
"epoch": 0.00665038153007675,
|
| 10479 |
+
"grad_norm": 8.5,
|
| 10480 |
+
"learning_rate": 3.3229975883261655e-07,
|
| 10481 |
+
"loss": 3.0898,
|
| 10482 |
+
"step": 1496
|
| 10483 |
+
},
|
| 10484 |
+
{
|
| 10485 |
+
"epoch": 0.006654826972276,
|
| 10486 |
+
"grad_norm": 8.5625,
|
| 10487 |
+
"learning_rate": 3.325220329187922e-07,
|
| 10488 |
+
"loss": 3.1635,
|
| 10489 |
+
"step": 1497
|
| 10490 |
+
},
|
| 10491 |
+
{
|
| 10492 |
+
"epoch": 0.006659272414475249,
|
| 10493 |
+
"grad_norm": 9.9375,
|
| 10494 |
+
"learning_rate": 3.327443070049678e-07,
|
| 10495 |
+
"loss": 3.1107,
|
| 10496 |
+
"step": 1498
|
| 10497 |
+
},
|
| 10498 |
+
{
|
| 10499 |
+
"epoch": 0.006663717856674498,
|
| 10500 |
+
"grad_norm": 8.3125,
|
| 10501 |
+
"learning_rate": 3.3296658109114353e-07,
|
| 10502 |
+
"loss": 3.2937,
|
| 10503 |
+
"step": 1499
|
| 10504 |
+
},
|
| 10505 |
+
{
|
| 10506 |
+
"epoch": 0.006668163298873747,
|
| 10507 |
+
"grad_norm": 10.0,
|
| 10508 |
+
"learning_rate": 3.3318885517731916e-07,
|
| 10509 |
+
"loss": 2.9609,
|
| 10510 |
+
"step": 1500
|
| 10511 |
+
},
|
| 10512 |
+
{
|
| 10513 |
+
"epoch": 0.006672608741072996,
|
| 10514 |
+
"grad_norm": 9.0,
|
| 10515 |
+
"learning_rate": 3.3341112926349484e-07,
|
| 10516 |
+
"loss": 3.1486,
|
| 10517 |
+
"step": 1501
|
| 10518 |
+
},
|
| 10519 |
+
{
|
| 10520 |
+
"epoch": 0.0066770541832722454,
|
| 10521 |
+
"grad_norm": 11.5625,
|
| 10522 |
+
"learning_rate": 3.3363340334967046e-07,
|
| 10523 |
+
"loss": 2.9105,
|
| 10524 |
+
"step": 1502
|
| 10525 |
+
},
|
| 10526 |
+
{
|
| 10527 |
+
"epoch": 0.0066814996254714945,
|
| 10528 |
+
"grad_norm": 8.3125,
|
| 10529 |
+
"learning_rate": 3.338556774358462e-07,
|
| 10530 |
+
"loss": 3.233,
|
| 10531 |
+
"step": 1503
|
| 10532 |
+
},
|
| 10533 |
+
{
|
| 10534 |
+
"epoch": 0.006685945067670744,
|
| 10535 |
+
"grad_norm": 9.0,
|
| 10536 |
+
"learning_rate": 3.340779515220218e-07,
|
| 10537 |
+
"loss": 3.1901,
|
| 10538 |
+
"step": 1504
|
| 10539 |
+
},
|
| 10540 |
+
{
|
| 10541 |
+
"epoch": 0.006690390509869993,
|
| 10542 |
+
"grad_norm": 11.5625,
|
| 10543 |
+
"learning_rate": 3.343002256081975e-07,
|
| 10544 |
+
"loss": 2.9103,
|
| 10545 |
+
"step": 1505
|
| 10546 |
+
},
|
| 10547 |
+
{
|
| 10548 |
+
"epoch": 0.0066948359520692426,
|
| 10549 |
+
"grad_norm": 12.125,
|
| 10550 |
+
"learning_rate": 3.345224996943731e-07,
|
| 10551 |
+
"loss": 2.8905,
|
| 10552 |
+
"step": 1506
|
| 10553 |
+
},
|
| 10554 |
+
{
|
| 10555 |
+
"epoch": 0.006699281394268492,
|
| 10556 |
+
"grad_norm": 10.5625,
|
| 10557 |
+
"learning_rate": 3.3474477378054885e-07,
|
| 10558 |
+
"loss": 2.8387,
|
| 10559 |
+
"step": 1507
|
| 10560 |
+
},
|
| 10561 |
+
{
|
| 10562 |
+
"epoch": 0.006703726836467741,
|
| 10563 |
+
"grad_norm": 11.9375,
|
| 10564 |
+
"learning_rate": 3.349670478667245e-07,
|
| 10565 |
+
"loss": 2.7865,
|
| 10566 |
+
"step": 1508
|
| 10567 |
+
},
|
| 10568 |
+
{
|
| 10569 |
+
"epoch": 0.00670817227866699,
|
| 10570 |
+
"grad_norm": 9.125,
|
| 10571 |
+
"learning_rate": 3.3518932195290015e-07,
|
| 10572 |
+
"loss": 3.1315,
|
| 10573 |
+
"step": 1509
|
| 10574 |
+
},
|
| 10575 |
+
{
|
| 10576 |
+
"epoch": 0.006712617720866239,
|
| 10577 |
+
"grad_norm": 9.0625,
|
| 10578 |
+
"learning_rate": 3.3541159603907583e-07,
|
| 10579 |
+
"loss": 3.0736,
|
| 10580 |
+
"step": 1510
|
| 10581 |
+
},
|
| 10582 |
+
{
|
| 10583 |
+
"epoch": 0.006717063163065488,
|
| 10584 |
+
"grad_norm": 8.625,
|
| 10585 |
+
"learning_rate": 3.356338701252515e-07,
|
| 10586 |
+
"loss": 3.1138,
|
| 10587 |
+
"step": 1511
|
| 10588 |
+
},
|
| 10589 |
+
{
|
| 10590 |
+
"epoch": 0.006721508605264737,
|
| 10591 |
+
"grad_norm": 10.6875,
|
| 10592 |
+
"learning_rate": 3.3585614421142713e-07,
|
| 10593 |
+
"loss": 3.0507,
|
| 10594 |
+
"step": 1512
|
| 10595 |
+
},
|
| 10596 |
+
{
|
| 10597 |
+
"epoch": 0.006725954047463986,
|
| 10598 |
+
"grad_norm": 10.375,
|
| 10599 |
+
"learning_rate": 3.360784182976028e-07,
|
| 10600 |
+
"loss": 3.0222,
|
| 10601 |
+
"step": 1513
|
| 10602 |
+
},
|
| 10603 |
+
{
|
| 10604 |
+
"epoch": 0.006730399489663236,
|
| 10605 |
+
"grad_norm": 12.0,
|
| 10606 |
+
"learning_rate": 3.363006923837785e-07,
|
| 10607 |
+
"loss": 2.7891,
|
| 10608 |
+
"step": 1514
|
| 10609 |
+
},
|
| 10610 |
+
{
|
| 10611 |
+
"epoch": 0.006734844931862485,
|
| 10612 |
+
"grad_norm": 8.6875,
|
| 10613 |
+
"learning_rate": 3.365229664699541e-07,
|
| 10614 |
+
"loss": 3.027,
|
| 10615 |
+
"step": 1515
|
| 10616 |
+
},
|
| 10617 |
+
{
|
| 10618 |
+
"epoch": 0.006739290374061734,
|
| 10619 |
+
"grad_norm": 9.875,
|
| 10620 |
+
"learning_rate": 3.367452405561298e-07,
|
| 10621 |
+
"loss": 2.9934,
|
| 10622 |
+
"step": 1516
|
| 10623 |
+
},
|
| 10624 |
+
{
|
| 10625 |
+
"epoch": 0.006743735816260983,
|
| 10626 |
+
"grad_norm": 10.3125,
|
| 10627 |
+
"learning_rate": 3.369675146423054e-07,
|
| 10628 |
+
"loss": 2.9754,
|
| 10629 |
+
"step": 1517
|
| 10630 |
+
},
|
| 10631 |
+
{
|
| 10632 |
+
"epoch": 0.006748181258460232,
|
| 10633 |
+
"grad_norm": 9.25,
|
| 10634 |
+
"learning_rate": 3.3718978872848115e-07,
|
| 10635 |
+
"loss": 3.003,
|
| 10636 |
+
"step": 1518
|
| 10637 |
+
},
|
| 10638 |
+
{
|
| 10639 |
+
"epoch": 0.006752626700659481,
|
| 10640 |
+
"grad_norm": 10.875,
|
| 10641 |
+
"learning_rate": 3.3741206281465677e-07,
|
| 10642 |
+
"loss": 2.863,
|
| 10643 |
+
"step": 1519
|
| 10644 |
+
},
|
| 10645 |
+
{
|
| 10646 |
+
"epoch": 0.00675707214285873,
|
| 10647 |
+
"grad_norm": 11.5,
|
| 10648 |
+
"learning_rate": 3.3763433690083245e-07,
|
| 10649 |
+
"loss": 2.8394,
|
| 10650 |
+
"step": 1520
|
| 10651 |
+
},
|
| 10652 |
+
{
|
| 10653 |
+
"epoch": 0.006761517585057979,
|
| 10654 |
+
"grad_norm": 11.3125,
|
| 10655 |
+
"learning_rate": 3.378566109870081e-07,
|
| 10656 |
+
"loss": 2.8357,
|
| 10657 |
+
"step": 1521
|
| 10658 |
+
},
|
| 10659 |
+
{
|
| 10660 |
+
"epoch": 0.006765963027257228,
|
| 10661 |
+
"grad_norm": 9.625,
|
| 10662 |
+
"learning_rate": 3.380788850731838e-07,
|
| 10663 |
+
"loss": 3.0925,
|
| 10664 |
+
"step": 1522
|
| 10665 |
+
},
|
| 10666 |
+
{
|
| 10667 |
+
"epoch": 0.006770408469456478,
|
| 10668 |
+
"grad_norm": 9.375,
|
| 10669 |
+
"learning_rate": 3.3830115915935943e-07,
|
| 10670 |
+
"loss": 3.2608,
|
| 10671 |
+
"step": 1523
|
| 10672 |
+
},
|
| 10673 |
+
{
|
| 10674 |
+
"epoch": 0.006774853911655727,
|
| 10675 |
+
"grad_norm": 9.5625,
|
| 10676 |
+
"learning_rate": 3.385234332455351e-07,
|
| 10677 |
+
"loss": 3.16,
|
| 10678 |
+
"step": 1524
|
| 10679 |
+
},
|
| 10680 |
+
{
|
| 10681 |
+
"epoch": 0.0067792993538549765,
|
| 10682 |
+
"grad_norm": 8.625,
|
| 10683 |
+
"learning_rate": 3.3874570733171073e-07,
|
| 10684 |
+
"loss": 3.0973,
|
| 10685 |
+
"step": 1525
|
| 10686 |
+
},
|
| 10687 |
+
{
|
| 10688 |
+
"epoch": 0.0067837447960542255,
|
| 10689 |
+
"grad_norm": 9.8125,
|
| 10690 |
+
"learning_rate": 3.3896798141788646e-07,
|
| 10691 |
+
"loss": 2.9419,
|
| 10692 |
+
"step": 1526
|
| 10693 |
+
},
|
| 10694 |
+
{
|
| 10695 |
+
"epoch": 0.006788190238253475,
|
| 10696 |
+
"grad_norm": 11.625,
|
| 10697 |
+
"learning_rate": 3.391902555040621e-07,
|
| 10698 |
+
"loss": 2.8104,
|
| 10699 |
+
"step": 1527
|
| 10700 |
+
},
|
| 10701 |
+
{
|
| 10702 |
+
"epoch": 0.006792635680452724,
|
| 10703 |
+
"grad_norm": 10.5,
|
| 10704 |
+
"learning_rate": 3.3941252959023777e-07,
|
| 10705 |
+
"loss": 2.9048,
|
| 10706 |
+
"step": 1528
|
| 10707 |
+
},
|
| 10708 |
+
{
|
| 10709 |
+
"epoch": 0.006797081122651973,
|
| 10710 |
+
"grad_norm": 10.4375,
|
| 10711 |
+
"learning_rate": 3.396348036764134e-07,
|
| 10712 |
+
"loss": 3.0021,
|
| 10713 |
+
"step": 1529
|
| 10714 |
+
},
|
| 10715 |
+
{
|
| 10716 |
+
"epoch": 0.006801526564851222,
|
| 10717 |
+
"grad_norm": 9.25,
|
| 10718 |
+
"learning_rate": 3.398570777625891e-07,
|
| 10719 |
+
"loss": 3.0744,
|
| 10720 |
+
"step": 1530
|
| 10721 |
+
},
|
| 10722 |
+
{
|
| 10723 |
+
"epoch": 0.006805972007050472,
|
| 10724 |
+
"grad_norm": 8.0,
|
| 10725 |
+
"learning_rate": 3.4007935184876475e-07,
|
| 10726 |
+
"loss": 3.2261,
|
| 10727 |
+
"step": 1531
|
| 10728 |
+
},
|
| 10729 |
+
{
|
| 10730 |
+
"epoch": 0.006810417449249721,
|
| 10731 |
+
"grad_norm": 11.75,
|
| 10732 |
+
"learning_rate": 3.4030162593494037e-07,
|
| 10733 |
+
"loss": 2.8506,
|
| 10734 |
+
"step": 1532
|
| 10735 |
+
},
|
| 10736 |
+
{
|
| 10737 |
+
"epoch": 0.00681486289144897,
|
| 10738 |
+
"grad_norm": 9.1875,
|
| 10739 |
+
"learning_rate": 3.4052390002111605e-07,
|
| 10740 |
+
"loss": 3.1766,
|
| 10741 |
+
"step": 1533
|
| 10742 |
+
},
|
| 10743 |
+
{
|
| 10744 |
+
"epoch": 0.006819308333648219,
|
| 10745 |
+
"grad_norm": 9.3125,
|
| 10746 |
+
"learning_rate": 3.407461741072917e-07,
|
| 10747 |
+
"loss": 3.0462,
|
| 10748 |
+
"step": 1534
|
| 10749 |
+
},
|
| 10750 |
+
{
|
| 10751 |
+
"epoch": 0.006823753775847468,
|
| 10752 |
+
"grad_norm": 7.3125,
|
| 10753 |
+
"learning_rate": 3.409684481934674e-07,
|
| 10754 |
+
"loss": 3.1243,
|
| 10755 |
+
"step": 1535
|
| 10756 |
+
},
|
| 10757 |
+
{
|
| 10758 |
+
"epoch": 0.006828199218046717,
|
| 10759 |
+
"grad_norm": 8.75,
|
| 10760 |
+
"learning_rate": 3.4119072227964303e-07,
|
| 10761 |
+
"loss": 3.0206,
|
| 10762 |
+
"step": 1536
|
| 10763 |
+
},
|
| 10764 |
+
{
|
| 10765 |
+
"epoch": 0.006832644660245966,
|
| 10766 |
+
"grad_norm": 11.8125,
|
| 10767 |
+
"learning_rate": 3.414129963658187e-07,
|
| 10768 |
+
"loss": 2.8674,
|
| 10769 |
+
"step": 1537
|
| 10770 |
+
},
|
| 10771 |
+
{
|
| 10772 |
+
"epoch": 0.006837090102445215,
|
| 10773 |
+
"grad_norm": 9.0625,
|
| 10774 |
+
"learning_rate": 3.416352704519944e-07,
|
| 10775 |
+
"loss": 3.0717,
|
| 10776 |
+
"step": 1538
|
| 10777 |
+
},
|
| 10778 |
+
{
|
| 10779 |
+
"epoch": 0.006841535544644464,
|
| 10780 |
+
"grad_norm": 12.0625,
|
| 10781 |
+
"learning_rate": 3.4185754453817006e-07,
|
| 10782 |
+
"loss": 2.7562,
|
| 10783 |
+
"step": 1539
|
| 10784 |
+
},
|
| 10785 |
+
{
|
| 10786 |
+
"epoch": 0.006845980986843714,
|
| 10787 |
+
"grad_norm": 9.5,
|
| 10788 |
+
"learning_rate": 3.420798186243457e-07,
|
| 10789 |
+
"loss": 3.0367,
|
| 10790 |
+
"step": 1540
|
| 10791 |
+
},
|
| 10792 |
+
{
|
| 10793 |
+
"epoch": 0.006850426429042963,
|
| 10794 |
+
"grad_norm": 11.1875,
|
| 10795 |
+
"learning_rate": 3.423020927105214e-07,
|
| 10796 |
+
"loss": 2.8669,
|
| 10797 |
+
"step": 1541
|
| 10798 |
+
},
|
| 10799 |
+
{
|
| 10800 |
+
"epoch": 0.006854871871242212,
|
| 10801 |
+
"grad_norm": 12.4375,
|
| 10802 |
+
"learning_rate": 3.4252436679669704e-07,
|
| 10803 |
+
"loss": 2.7301,
|
| 10804 |
+
"step": 1542
|
| 10805 |
+
},
|
| 10806 |
+
{
|
| 10807 |
+
"epoch": 0.006859317313441461,
|
| 10808 |
+
"grad_norm": 11.1875,
|
| 10809 |
+
"learning_rate": 3.427466408828727e-07,
|
| 10810 |
+
"loss": 2.866,
|
| 10811 |
+
"step": 1543
|
| 10812 |
+
},
|
| 10813 |
+
{
|
| 10814 |
+
"epoch": 0.00686376275564071,
|
| 10815 |
+
"grad_norm": 7.59375,
|
| 10816 |
+
"learning_rate": 3.4296891496904835e-07,
|
| 10817 |
+
"loss": 3.1486,
|
| 10818 |
+
"step": 1544
|
| 10819 |
+
},
|
| 10820 |
+
{
|
| 10821 |
+
"epoch": 0.0068682081978399595,
|
| 10822 |
+
"grad_norm": 12.0,
|
| 10823 |
+
"learning_rate": 3.431911890552241e-07,
|
| 10824 |
+
"loss": 2.8157,
|
| 10825 |
+
"step": 1545
|
| 10826 |
+
},
|
| 10827 |
+
{
|
| 10828 |
+
"epoch": 0.0068726536400392085,
|
| 10829 |
+
"grad_norm": 9.375,
|
| 10830 |
+
"learning_rate": 3.434134631413997e-07,
|
| 10831 |
+
"loss": 2.9621,
|
| 10832 |
+
"step": 1546
|
| 10833 |
+
},
|
| 10834 |
+
{
|
| 10835 |
+
"epoch": 0.006877099082238458,
|
| 10836 |
+
"grad_norm": 8.1875,
|
| 10837 |
+
"learning_rate": 3.436357372275753e-07,
|
| 10838 |
+
"loss": 3.2594,
|
| 10839 |
+
"step": 1547
|
| 10840 |
+
},
|
| 10841 |
+
{
|
| 10842 |
+
"epoch": 0.0068815445244377075,
|
| 10843 |
+
"grad_norm": 8.75,
|
| 10844 |
+
"learning_rate": 3.43858011313751e-07,
|
| 10845 |
+
"loss": 3.2303,
|
| 10846 |
+
"step": 1548
|
| 10847 |
+
},
|
| 10848 |
+
{
|
| 10849 |
+
"epoch": 0.006885989966636957,
|
| 10850 |
+
"grad_norm": 9.8125,
|
| 10851 |
+
"learning_rate": 3.4408028539992663e-07,
|
| 10852 |
+
"loss": 2.9867,
|
| 10853 |
+
"step": 1549
|
| 10854 |
+
},
|
| 10855 |
+
{
|
| 10856 |
+
"epoch": 0.006890435408836206,
|
| 10857 |
+
"grad_norm": 10.375,
|
| 10858 |
+
"learning_rate": 3.4430255948610236e-07,
|
| 10859 |
+
"loss": 3.0437,
|
| 10860 |
+
"step": 1550
|
| 10861 |
+
},
|
| 10862 |
+
{
|
| 10863 |
+
"epoch": 0.006894880851035455,
|
| 10864 |
+
"grad_norm": 10.0625,
|
| 10865 |
+
"learning_rate": 3.44524833572278e-07,
|
| 10866 |
+
"loss": 3.0103,
|
| 10867 |
+
"step": 1551
|
| 10868 |
+
},
|
| 10869 |
+
{
|
| 10870 |
+
"epoch": 0.006899326293234704,
|
| 10871 |
+
"grad_norm": 11.5,
|
| 10872 |
+
"learning_rate": 3.4474710765845366e-07,
|
| 10873 |
+
"loss": 2.8495,
|
| 10874 |
+
"step": 1552
|
| 10875 |
+
},
|
| 10876 |
+
{
|
| 10877 |
+
"epoch": 0.006903771735433953,
|
| 10878 |
+
"grad_norm": 7.15625,
|
| 10879 |
+
"learning_rate": 3.449693817446293e-07,
|
| 10880 |
+
"loss": 3.3484,
|
| 10881 |
+
"step": 1553
|
| 10882 |
+
},
|
| 10883 |
+
{
|
| 10884 |
+
"epoch": 0.006908217177633202,
|
| 10885 |
+
"grad_norm": 11.125,
|
| 10886 |
+
"learning_rate": 3.45191655830805e-07,
|
| 10887 |
+
"loss": 2.9013,
|
| 10888 |
+
"step": 1554
|
| 10889 |
+
},
|
| 10890 |
+
{
|
| 10891 |
+
"epoch": 0.006912662619832451,
|
| 10892 |
+
"grad_norm": 8.25,
|
| 10893 |
+
"learning_rate": 3.4541392991698064e-07,
|
| 10894 |
+
"loss": 3.1855,
|
| 10895 |
+
"step": 1555
|
| 10896 |
+
},
|
| 10897 |
+
{
|
| 10898 |
+
"epoch": 0.0069171080620317,
|
| 10899 |
+
"grad_norm": 11.6875,
|
| 10900 |
+
"learning_rate": 3.456362040031563e-07,
|
| 10901 |
+
"loss": 2.904,
|
| 10902 |
+
"step": 1556
|
| 10903 |
+
},
|
| 10904 |
+
{
|
| 10905 |
+
"epoch": 0.00692155350423095,
|
| 10906 |
+
"grad_norm": 9.0625,
|
| 10907 |
+
"learning_rate": 3.4585847808933195e-07,
|
| 10908 |
+
"loss": 3.1332,
|
| 10909 |
+
"step": 1557
|
| 10910 |
+
},
|
| 10911 |
+
{
|
| 10912 |
+
"epoch": 0.006925998946430199,
|
| 10913 |
+
"grad_norm": 8.6875,
|
| 10914 |
+
"learning_rate": 3.460807521755077e-07,
|
| 10915 |
+
"loss": 3.0808,
|
| 10916 |
+
"step": 1558
|
| 10917 |
+
},
|
| 10918 |
+
{
|
| 10919 |
+
"epoch": 0.006930444388629448,
|
| 10920 |
+
"grad_norm": 8.25,
|
| 10921 |
+
"learning_rate": 3.463030262616833e-07,
|
| 10922 |
+
"loss": 3.4512,
|
| 10923 |
+
"step": 1559
|
| 10924 |
+
},
|
| 10925 |
+
{
|
| 10926 |
+
"epoch": 0.006934889830828697,
|
| 10927 |
+
"grad_norm": 9.25,
|
| 10928 |
+
"learning_rate": 3.46525300347859e-07,
|
| 10929 |
+
"loss": 3.0079,
|
| 10930 |
+
"step": 1560
|
| 10931 |
+
},
|
| 10932 |
+
{
|
| 10933 |
+
"epoch": 0.006939335273027946,
|
| 10934 |
+
"grad_norm": 8.8125,
|
| 10935 |
+
"learning_rate": 3.4674757443403466e-07,
|
| 10936 |
+
"loss": 3.2632,
|
| 10937 |
+
"step": 1561
|
| 10938 |
+
},
|
| 10939 |
+
{
|
| 10940 |
+
"epoch": 0.006943780715227195,
|
| 10941 |
+
"grad_norm": 11.625,
|
| 10942 |
+
"learning_rate": 3.4696984852021033e-07,
|
| 10943 |
+
"loss": 2.7714,
|
| 10944 |
+
"step": 1562
|
| 10945 |
+
},
|
| 10946 |
+
{
|
| 10947 |
+
"epoch": 0.006948226157426444,
|
| 10948 |
+
"grad_norm": 9.375,
|
| 10949 |
+
"learning_rate": 3.4719212260638596e-07,
|
| 10950 |
+
"loss": 3.0542,
|
| 10951 |
+
"step": 1563
|
| 10952 |
+
},
|
| 10953 |
+
{
|
| 10954 |
+
"epoch": 0.006952671599625693,
|
| 10955 |
+
"grad_norm": 9.0625,
|
| 10956 |
+
"learning_rate": 3.474143966925616e-07,
|
| 10957 |
+
"loss": 3.121,
|
| 10958 |
+
"step": 1564
|
| 10959 |
+
},
|
| 10960 |
+
{
|
| 10961 |
+
"epoch": 0.006957117041824943,
|
| 10962 |
+
"grad_norm": 9.1875,
|
| 10963 |
+
"learning_rate": 3.476366707787373e-07,
|
| 10964 |
+
"loss": 2.9955,
|
| 10965 |
+
"step": 1565
|
| 10966 |
+
},
|
| 10967 |
+
{
|
| 10968 |
+
"epoch": 0.006961562484024192,
|
| 10969 |
+
"grad_norm": 10.1875,
|
| 10970 |
+
"learning_rate": 3.4785894486491294e-07,
|
| 10971 |
+
"loss": 3.0429,
|
| 10972 |
+
"step": 1566
|
| 10973 |
+
},
|
| 10974 |
+
{
|
| 10975 |
+
"epoch": 0.0069660079262234414,
|
| 10976 |
+
"grad_norm": 10.875,
|
| 10977 |
+
"learning_rate": 3.480812189510886e-07,
|
| 10978 |
+
"loss": 2.9506,
|
| 10979 |
+
"step": 1567
|
| 10980 |
+
},
|
| 10981 |
+
{
|
| 10982 |
+
"epoch": 0.0069704533684226905,
|
| 10983 |
+
"grad_norm": 11.75,
|
| 10984 |
+
"learning_rate": 3.4830349303726424e-07,
|
| 10985 |
+
"loss": 2.8532,
|
| 10986 |
+
"step": 1568
|
| 10987 |
+
},
|
| 10988 |
+
{
|
| 10989 |
+
"epoch": 0.00697489881062194,
|
| 10990 |
+
"grad_norm": 8.5625,
|
| 10991 |
+
"learning_rate": 3.4852576712343997e-07,
|
| 10992 |
+
"loss": 3.1654,
|
| 10993 |
+
"step": 1569
|
| 10994 |
+
},
|
| 10995 |
+
{
|
| 10996 |
+
"epoch": 0.006979344252821189,
|
| 10997 |
+
"grad_norm": 9.5,
|
| 10998 |
+
"learning_rate": 3.487480412096156e-07,
|
| 10999 |
+
"loss": 2.9531,
|
| 11000 |
+
"step": 1570
|
| 11001 |
+
},
|
| 11002 |
+
{
|
| 11003 |
+
"epoch": 0.006983789695020438,
|
| 11004 |
+
"grad_norm": 12.25,
|
| 11005 |
+
"learning_rate": 3.489703152957913e-07,
|
| 11006 |
+
"loss": 2.7171,
|
| 11007 |
+
"step": 1571
|
| 11008 |
+
},
|
| 11009 |
+
{
|
| 11010 |
+
"epoch": 0.006988235137219687,
|
| 11011 |
+
"grad_norm": 9.625,
|
| 11012 |
+
"learning_rate": 3.491925893819669e-07,
|
| 11013 |
+
"loss": 2.9861,
|
| 11014 |
+
"step": 1572
|
| 11015 |
+
},
|
| 11016 |
+
{
|
| 11017 |
+
"epoch": 0.006992680579418936,
|
| 11018 |
+
"grad_norm": 9.8125,
|
| 11019 |
+
"learning_rate": 3.4941486346814263e-07,
|
| 11020 |
+
"loss": 3.0001,
|
| 11021 |
+
"step": 1573
|
| 11022 |
+
},
|
| 11023 |
+
{
|
| 11024 |
+
"epoch": 0.006997126021618186,
|
| 11025 |
+
"grad_norm": 9.875,
|
| 11026 |
+
"learning_rate": 3.4963713755431826e-07,
|
| 11027 |
+
"loss": 2.9464,
|
| 11028 |
+
"step": 1574
|
| 11029 |
+
},
|
| 11030 |
+
{
|
| 11031 |
+
"epoch": 0.007001571463817435,
|
| 11032 |
+
"grad_norm": 8.6875,
|
| 11033 |
+
"learning_rate": 3.4985941164049393e-07,
|
| 11034 |
+
"loss": 3.1696,
|
| 11035 |
+
"step": 1575
|
| 11036 |
+
},
|
| 11037 |
+
{
|
| 11038 |
+
"epoch": 0.007006016906016684,
|
| 11039 |
+
"grad_norm": 9.8125,
|
| 11040 |
+
"learning_rate": 3.5008168572666956e-07,
|
| 11041 |
+
"loss": 2.996,
|
| 11042 |
+
"step": 1576
|
| 11043 |
+
},
|
| 11044 |
+
{
|
| 11045 |
+
"epoch": 0.007010462348215933,
|
| 11046 |
+
"grad_norm": 11.5,
|
| 11047 |
+
"learning_rate": 3.503039598128453e-07,
|
| 11048 |
+
"loss": 2.8594,
|
| 11049 |
+
"step": 1577
|
| 11050 |
+
},
|
| 11051 |
+
{
|
| 11052 |
+
"epoch": 0.007014907790415182,
|
| 11053 |
+
"grad_norm": 7.5,
|
| 11054 |
+
"learning_rate": 3.505262338990209e-07,
|
| 11055 |
+
"loss": 3.3521,
|
| 11056 |
+
"step": 1578
|
| 11057 |
+
},
|
| 11058 |
+
{
|
| 11059 |
+
"epoch": 0.007019353232614431,
|
| 11060 |
+
"grad_norm": 7.90625,
|
| 11061 |
+
"learning_rate": 3.507485079851966e-07,
|
| 11062 |
+
"loss": 3.2562,
|
| 11063 |
+
"step": 1579
|
| 11064 |
+
},
|
| 11065 |
+
{
|
| 11066 |
+
"epoch": 0.00702379867481368,
|
| 11067 |
+
"grad_norm": 8.25,
|
| 11068 |
+
"learning_rate": 3.509707820713722e-07,
|
| 11069 |
+
"loss": 3.3022,
|
| 11070 |
+
"step": 1580
|
| 11071 |
+
},
|
| 11072 |
+
{
|
| 11073 |
+
"epoch": 0.007028244117012929,
|
| 11074 |
+
"grad_norm": 9.625,
|
| 11075 |
+
"learning_rate": 3.5119305615754784e-07,
|
| 11076 |
+
"loss": 2.986,
|
| 11077 |
+
"step": 1581
|
| 11078 |
+
},
|
| 11079 |
+
{
|
| 11080 |
+
"epoch": 0.007032689559212179,
|
| 11081 |
+
"grad_norm": 8.125,
|
| 11082 |
+
"learning_rate": 3.5141533024372357e-07,
|
| 11083 |
+
"loss": 3.4736,
|
| 11084 |
+
"step": 1582
|
| 11085 |
+
},
|
| 11086 |
+
{
|
| 11087 |
+
"epoch": 0.007037135001411428,
|
| 11088 |
+
"grad_norm": 12.3125,
|
| 11089 |
+
"learning_rate": 3.516376043298992e-07,
|
| 11090 |
+
"loss": 2.797,
|
| 11091 |
+
"step": 1583
|
| 11092 |
+
},
|
| 11093 |
+
{
|
| 11094 |
+
"epoch": 0.007041580443610677,
|
| 11095 |
+
"grad_norm": 10.3125,
|
| 11096 |
+
"learning_rate": 3.518598784160749e-07,
|
| 11097 |
+
"loss": 3.002,
|
| 11098 |
+
"step": 1584
|
| 11099 |
+
},
|
| 11100 |
+
{
|
| 11101 |
+
"epoch": 0.007046025885809926,
|
| 11102 |
+
"grad_norm": 9.625,
|
| 11103 |
+
"learning_rate": 3.5208215250225055e-07,
|
| 11104 |
+
"loss": 2.9461,
|
| 11105 |
+
"step": 1585
|
| 11106 |
+
},
|
| 11107 |
+
{
|
| 11108 |
+
"epoch": 0.007050471328009175,
|
| 11109 |
+
"grad_norm": 10.25,
|
| 11110 |
+
"learning_rate": 3.5230442658842623e-07,
|
| 11111 |
+
"loss": 2.9823,
|
| 11112 |
+
"step": 1586
|
| 11113 |
+
},
|
| 11114 |
+
{
|
| 11115 |
+
"epoch": 0.0070549167702084244,
|
| 11116 |
+
"grad_norm": 10.4375,
|
| 11117 |
+
"learning_rate": 3.5252670067460186e-07,
|
| 11118 |
+
"loss": 2.9964,
|
| 11119 |
+
"step": 1587
|
| 11120 |
+
},
|
| 11121 |
+
{
|
| 11122 |
+
"epoch": 0.0070593622124076735,
|
| 11123 |
+
"grad_norm": 9.1875,
|
| 11124 |
+
"learning_rate": 3.5274897476077753e-07,
|
| 11125 |
+
"loss": 3.0419,
|
| 11126 |
+
"step": 1588
|
| 11127 |
+
},
|
| 11128 |
+
{
|
| 11129 |
+
"epoch": 0.0070638076546069226,
|
| 11130 |
+
"grad_norm": 10.5,
|
| 11131 |
+
"learning_rate": 3.529712488469532e-07,
|
| 11132 |
+
"loss": 2.9279,
|
| 11133 |
+
"step": 1589
|
| 11134 |
+
},
|
| 11135 |
+
{
|
| 11136 |
+
"epoch": 0.007068253096806172,
|
| 11137 |
+
"grad_norm": 9.375,
|
| 11138 |
+
"learning_rate": 3.531935229331289e-07,
|
| 11139 |
+
"loss": 2.9769,
|
| 11140 |
+
"step": 1590
|
| 11141 |
+
},
|
| 11142 |
+
{
|
| 11143 |
+
"epoch": 0.0070726985390054216,
|
| 11144 |
+
"grad_norm": 8.5625,
|
| 11145 |
+
"learning_rate": 3.534157970193045e-07,
|
| 11146 |
+
"loss": 3.1274,
|
| 11147 |
+
"step": 1591
|
| 11148 |
+
},
|
| 11149 |
+
{
|
| 11150 |
+
"epoch": 0.007077143981204671,
|
| 11151 |
+
"grad_norm": 10.125,
|
| 11152 |
+
"learning_rate": 3.5363807110548024e-07,
|
| 11153 |
+
"loss": 3.0245,
|
| 11154 |
+
"step": 1592
|
| 11155 |
+
},
|
| 11156 |
+
{
|
| 11157 |
+
"epoch": 0.00708158942340392,
|
| 11158 |
+
"grad_norm": 10.375,
|
| 11159 |
+
"learning_rate": 3.5386034519165587e-07,
|
| 11160 |
+
"loss": 2.9498,
|
| 11161 |
+
"step": 1593
|
| 11162 |
+
},
|
| 11163 |
+
{
|
| 11164 |
+
"epoch": 0.007086034865603169,
|
| 11165 |
+
"grad_norm": 11.125,
|
| 11166 |
+
"learning_rate": 3.5408261927783155e-07,
|
| 11167 |
+
"loss": 2.9692,
|
| 11168 |
+
"step": 1594
|
| 11169 |
+
},
|
| 11170 |
+
{
|
| 11171 |
+
"epoch": 0.007090480307802418,
|
| 11172 |
+
"grad_norm": 10.8125,
|
| 11173 |
+
"learning_rate": 3.5430489336400717e-07,
|
| 11174 |
+
"loss": 3.0287,
|
| 11175 |
+
"step": 1595
|
| 11176 |
+
},
|
| 11177 |
+
{
|
| 11178 |
+
"epoch": 0.007094925750001667,
|
| 11179 |
+
"grad_norm": 12.3125,
|
| 11180 |
+
"learning_rate": 3.545271674501829e-07,
|
| 11181 |
+
"loss": 2.7552,
|
| 11182 |
+
"step": 1596
|
| 11183 |
+
},
|
| 11184 |
+
{
|
| 11185 |
+
"epoch": 0.007099371192200916,
|
| 11186 |
+
"grad_norm": 10.375,
|
| 11187 |
+
"learning_rate": 3.5474944153635853e-07,
|
| 11188 |
+
"loss": 2.9677,
|
| 11189 |
+
"step": 1597
|
| 11190 |
+
},
|
| 11191 |
+
{
|
| 11192 |
+
"epoch": 0.007103816634400165,
|
| 11193 |
+
"grad_norm": 10.4375,
|
| 11194 |
+
"learning_rate": 3.5497171562253415e-07,
|
| 11195 |
+
"loss": 2.9863,
|
| 11196 |
+
"step": 1598
|
| 11197 |
+
},
|
| 11198 |
+
{
|
| 11199 |
+
"epoch": 0.007108262076599415,
|
| 11200 |
+
"grad_norm": 11.3125,
|
| 11201 |
+
"learning_rate": 3.5519398970870983e-07,
|
| 11202 |
+
"loss": 2.8698,
|
| 11203 |
+
"step": 1599
|
| 11204 |
+
},
|
| 11205 |
+
{
|
| 11206 |
+
"epoch": 0.007112707518798664,
|
| 11207 |
+
"grad_norm": 12.1875,
|
| 11208 |
+
"learning_rate": 3.5541626379488546e-07,
|
| 11209 |
+
"loss": 2.7243,
|
| 11210 |
+
"step": 1600
|
| 11211 |
}
|
| 11212 |
],
|
| 11213 |
"logging_steps": 1,
|
|
|
|
| 11227 |
"attributes": {}
|
| 11228 |
}
|
| 11229 |
},
|
| 11230 |
+
"total_flos": 6.3339761762304e+16,
|
| 11231 |
"train_batch_size": 1,
|
| 11232 |
"trial_name": null,
|
| 11233 |
"trial_params": null
|