Training in progress, step 1270, 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:b0143bc91b3ad815b4f13a42aea5c7975d4f31dd640aa131f2349d4581586ad7
|
| 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 sha256:14c3e525ca05e179ddf164d7eead6f8f3351de1c8dbf833b0398766f590b5299
|
| 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,
|
|
@@ -6658,6 +6658,2246 @@
|
|
| 6658 |
"learning_rate": 1.999998627690604e-05,
|
| 6659 |
"loss": 3.2958,
|
| 6660 |
"step": 950
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6661 |
}
|
| 6662 |
],
|
| 6663 |
"logging_steps": 1,
|
|
@@ -6677,7 +8917,7 @@
|
|
| 6677 |
"attributes": {}
|
| 6678 |
}
|
| 6679 |
},
|
| 6680 |
-
"total_flos":
|
| 6681 |
"train_batch_size": 1,
|
| 6682 |
"trial_name": null,
|
| 6683 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.0056457115930464395,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 1270,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 6658 |
"learning_rate": 1.999998627690604e-05,
|
| 6659 |
"loss": 3.2958,
|
| 6660 |
"step": 950
|
| 6661 |
+
},
|
| 6662 |
+
{
|
| 6663 |
+
"epoch": 0.004227615531485956,
|
| 6664 |
+
"grad_norm": 10.0625,
|
| 6665 |
+
"learning_rate": 1.9999986247969644e-05,
|
| 6666 |
+
"loss": 2.955,
|
| 6667 |
+
"step": 951
|
| 6668 |
+
},
|
| 6669 |
+
{
|
| 6670 |
+
"epoch": 0.004232060973685205,
|
| 6671 |
+
"grad_norm": 10.5625,
|
| 6672 |
+
"learning_rate": 2.1138265595305574e-07,
|
| 6673 |
+
"loss": 3.0113,
|
| 6674 |
+
"step": 952
|
| 6675 |
+
},
|
| 6676 |
+
{
|
| 6677 |
+
"epoch": 0.004236506415884454,
|
| 6678 |
+
"grad_norm": 10.5625,
|
| 6679 |
+
"learning_rate": 2.116049300392314e-07,
|
| 6680 |
+
"loss": 2.9137,
|
| 6681 |
+
"step": 953
|
| 6682 |
+
},
|
| 6683 |
+
{
|
| 6684 |
+
"epoch": 0.004240951858083703,
|
| 6685 |
+
"grad_norm": 9.4375,
|
| 6686 |
+
"learning_rate": 2.1182720412540707e-07,
|
| 6687 |
+
"loss": 3.0932,
|
| 6688 |
+
"step": 954
|
| 6689 |
+
},
|
| 6690 |
+
{
|
| 6691 |
+
"epoch": 0.004245397300282953,
|
| 6692 |
+
"grad_norm": 9.9375,
|
| 6693 |
+
"learning_rate": 2.1204947821158272e-07,
|
| 6694 |
+
"loss": 3.0056,
|
| 6695 |
+
"step": 955
|
| 6696 |
+
},
|
| 6697 |
+
{
|
| 6698 |
+
"epoch": 0.004249842742482202,
|
| 6699 |
+
"grad_norm": 7.625,
|
| 6700 |
+
"learning_rate": 2.122717522977584e-07,
|
| 6701 |
+
"loss": 3.1488,
|
| 6702 |
+
"step": 956
|
| 6703 |
+
},
|
| 6704 |
+
{
|
| 6705 |
+
"epoch": 0.004254288184681451,
|
| 6706 |
+
"grad_norm": 10.125,
|
| 6707 |
+
"learning_rate": 2.1249402638393403e-07,
|
| 6708 |
+
"loss": 2.9632,
|
| 6709 |
+
"step": 957
|
| 6710 |
+
},
|
| 6711 |
+
{
|
| 6712 |
+
"epoch": 0.0042587336268807,
|
| 6713 |
+
"grad_norm": 7.5,
|
| 6714 |
+
"learning_rate": 2.127163004701097e-07,
|
| 6715 |
+
"loss": 3.2703,
|
| 6716 |
+
"step": 958
|
| 6717 |
+
},
|
| 6718 |
+
{
|
| 6719 |
+
"epoch": 0.004263179069079949,
|
| 6720 |
+
"grad_norm": 8.75,
|
| 6721 |
+
"learning_rate": 2.1293857455628536e-07,
|
| 6722 |
+
"loss": 3.1646,
|
| 6723 |
+
"step": 959
|
| 6724 |
+
},
|
| 6725 |
+
{
|
| 6726 |
+
"epoch": 0.004267624511279198,
|
| 6727 |
+
"grad_norm": 8.1875,
|
| 6728 |
+
"learning_rate": 2.1316084864246103e-07,
|
| 6729 |
+
"loss": 3.1636,
|
| 6730 |
+
"step": 960
|
| 6731 |
+
},
|
| 6732 |
+
{
|
| 6733 |
+
"epoch": 0.004272069953478447,
|
| 6734 |
+
"grad_norm": 8.9375,
|
| 6735 |
+
"learning_rate": 2.1338312272863669e-07,
|
| 6736 |
+
"loss": 3.0689,
|
| 6737 |
+
"step": 961
|
| 6738 |
+
},
|
| 6739 |
+
{
|
| 6740 |
+
"epoch": 0.004276515395677696,
|
| 6741 |
+
"grad_norm": 9.25,
|
| 6742 |
+
"learning_rate": 2.1360539681481236e-07,
|
| 6743 |
+
"loss": 3.0252,
|
| 6744 |
+
"step": 962
|
| 6745 |
+
},
|
| 6746 |
+
{
|
| 6747 |
+
"epoch": 0.004280960837876946,
|
| 6748 |
+
"grad_norm": 11.75,
|
| 6749 |
+
"learning_rate": 2.1382767090098801e-07,
|
| 6750 |
+
"loss": 2.8206,
|
| 6751 |
+
"step": 963
|
| 6752 |
+
},
|
| 6753 |
+
{
|
| 6754 |
+
"epoch": 0.004285406280076195,
|
| 6755 |
+
"grad_norm": 8.6875,
|
| 6756 |
+
"learning_rate": 2.140499449871637e-07,
|
| 6757 |
+
"loss": 3.0941,
|
| 6758 |
+
"step": 964
|
| 6759 |
+
},
|
| 6760 |
+
{
|
| 6761 |
+
"epoch": 0.004289851722275444,
|
| 6762 |
+
"grad_norm": 10.875,
|
| 6763 |
+
"learning_rate": 2.1427221907333934e-07,
|
| 6764 |
+
"loss": 2.9032,
|
| 6765 |
+
"step": 965
|
| 6766 |
+
},
|
| 6767 |
+
{
|
| 6768 |
+
"epoch": 0.004294297164474693,
|
| 6769 |
+
"grad_norm": 8.5625,
|
| 6770 |
+
"learning_rate": 2.1449449315951502e-07,
|
| 6771 |
+
"loss": 3.1731,
|
| 6772 |
+
"step": 966
|
| 6773 |
+
},
|
| 6774 |
+
{
|
| 6775 |
+
"epoch": 0.004298742606673942,
|
| 6776 |
+
"grad_norm": 9.1875,
|
| 6777 |
+
"learning_rate": 2.1471676724569067e-07,
|
| 6778 |
+
"loss": 3.0915,
|
| 6779 |
+
"step": 967
|
| 6780 |
+
},
|
| 6781 |
+
{
|
| 6782 |
+
"epoch": 0.004303188048873191,
|
| 6783 |
+
"grad_norm": 11.0625,
|
| 6784 |
+
"learning_rate": 2.1493904133186635e-07,
|
| 6785 |
+
"loss": 2.8839,
|
| 6786 |
+
"step": 968
|
| 6787 |
+
},
|
| 6788 |
+
{
|
| 6789 |
+
"epoch": 0.0043076334910724405,
|
| 6790 |
+
"grad_norm": 9.5625,
|
| 6791 |
+
"learning_rate": 2.15161315418042e-07,
|
| 6792 |
+
"loss": 3.082,
|
| 6793 |
+
"step": 969
|
| 6794 |
+
},
|
| 6795 |
+
{
|
| 6796 |
+
"epoch": 0.0043120789332716895,
|
| 6797 |
+
"grad_norm": 10.125,
|
| 6798 |
+
"learning_rate": 2.1538358950421768e-07,
|
| 6799 |
+
"loss": 3.0645,
|
| 6800 |
+
"step": 970
|
| 6801 |
+
},
|
| 6802 |
+
{
|
| 6803 |
+
"epoch": 0.0043165243754709395,
|
| 6804 |
+
"grad_norm": 9.0625,
|
| 6805 |
+
"learning_rate": 2.1560586359039333e-07,
|
| 6806 |
+
"loss": 3.3608,
|
| 6807 |
+
"step": 971
|
| 6808 |
+
},
|
| 6809 |
+
{
|
| 6810 |
+
"epoch": 0.0043209698176701885,
|
| 6811 |
+
"grad_norm": 8.6875,
|
| 6812 |
+
"learning_rate": 2.15828137676569e-07,
|
| 6813 |
+
"loss": 3.0053,
|
| 6814 |
+
"step": 972
|
| 6815 |
+
},
|
| 6816 |
+
{
|
| 6817 |
+
"epoch": 0.004325415259869438,
|
| 6818 |
+
"grad_norm": 10.375,
|
| 6819 |
+
"learning_rate": 2.1605041176274469e-07,
|
| 6820 |
+
"loss": 3.0382,
|
| 6821 |
+
"step": 973
|
| 6822 |
+
},
|
| 6823 |
+
{
|
| 6824 |
+
"epoch": 0.004329860702068687,
|
| 6825 |
+
"grad_norm": 11.1875,
|
| 6826 |
+
"learning_rate": 2.162726858489203e-07,
|
| 6827 |
+
"loss": 2.9776,
|
| 6828 |
+
"step": 974
|
| 6829 |
+
},
|
| 6830 |
+
{
|
| 6831 |
+
"epoch": 0.004334306144267936,
|
| 6832 |
+
"grad_norm": 10.75,
|
| 6833 |
+
"learning_rate": 2.1649495993509596e-07,
|
| 6834 |
+
"loss": 2.9646,
|
| 6835 |
+
"step": 975
|
| 6836 |
+
},
|
| 6837 |
+
{
|
| 6838 |
+
"epoch": 0.004338751586467185,
|
| 6839 |
+
"grad_norm": 10.3125,
|
| 6840 |
+
"learning_rate": 2.1671723402127164e-07,
|
| 6841 |
+
"loss": 3.0247,
|
| 6842 |
+
"step": 976
|
| 6843 |
+
},
|
| 6844 |
+
{
|
| 6845 |
+
"epoch": 0.004343197028666434,
|
| 6846 |
+
"grad_norm": 9.875,
|
| 6847 |
+
"learning_rate": 2.169395081074473e-07,
|
| 6848 |
+
"loss": 3.0667,
|
| 6849 |
+
"step": 977
|
| 6850 |
+
},
|
| 6851 |
+
{
|
| 6852 |
+
"epoch": 0.004347642470865683,
|
| 6853 |
+
"grad_norm": 11.4375,
|
| 6854 |
+
"learning_rate": 2.1716178219362297e-07,
|
| 6855 |
+
"loss": 2.8984,
|
| 6856 |
+
"step": 978
|
| 6857 |
+
},
|
| 6858 |
+
{
|
| 6859 |
+
"epoch": 0.004352087913064932,
|
| 6860 |
+
"grad_norm": 8.6875,
|
| 6861 |
+
"learning_rate": 2.1738405627979862e-07,
|
| 6862 |
+
"loss": 3.1805,
|
| 6863 |
+
"step": 979
|
| 6864 |
+
},
|
| 6865 |
+
{
|
| 6866 |
+
"epoch": 0.004356533355264182,
|
| 6867 |
+
"grad_norm": 7.34375,
|
| 6868 |
+
"learning_rate": 2.176063303659743e-07,
|
| 6869 |
+
"loss": 3.2741,
|
| 6870 |
+
"step": 980
|
| 6871 |
+
},
|
| 6872 |
+
{
|
| 6873 |
+
"epoch": 0.004360978797463431,
|
| 6874 |
+
"grad_norm": 9.25,
|
| 6875 |
+
"learning_rate": 2.1782860445214995e-07,
|
| 6876 |
+
"loss": 3.1854,
|
| 6877 |
+
"step": 981
|
| 6878 |
+
},
|
| 6879 |
+
{
|
| 6880 |
+
"epoch": 0.00436542423966268,
|
| 6881 |
+
"grad_norm": 12.5625,
|
| 6882 |
+
"learning_rate": 2.1805087853832563e-07,
|
| 6883 |
+
"loss": 2.7744,
|
| 6884 |
+
"step": 982
|
| 6885 |
+
},
|
| 6886 |
+
{
|
| 6887 |
+
"epoch": 0.004369869681861929,
|
| 6888 |
+
"grad_norm": 10.9375,
|
| 6889 |
+
"learning_rate": 2.182731526245013e-07,
|
| 6890 |
+
"loss": 2.8225,
|
| 6891 |
+
"step": 983
|
| 6892 |
+
},
|
| 6893 |
+
{
|
| 6894 |
+
"epoch": 0.004374315124061178,
|
| 6895 |
+
"grad_norm": 8.25,
|
| 6896 |
+
"learning_rate": 2.1849542671067696e-07,
|
| 6897 |
+
"loss": 3.277,
|
| 6898 |
+
"step": 984
|
| 6899 |
+
},
|
| 6900 |
+
{
|
| 6901 |
+
"epoch": 0.004378760566260427,
|
| 6902 |
+
"grad_norm": 9.75,
|
| 6903 |
+
"learning_rate": 2.1871770079685263e-07,
|
| 6904 |
+
"loss": 3.0862,
|
| 6905 |
+
"step": 985
|
| 6906 |
+
},
|
| 6907 |
+
{
|
| 6908 |
+
"epoch": 0.004383206008459676,
|
| 6909 |
+
"grad_norm": 6.9375,
|
| 6910 |
+
"learning_rate": 2.1893997488302829e-07,
|
| 6911 |
+
"loss": 3.2672,
|
| 6912 |
+
"step": 986
|
| 6913 |
+
},
|
| 6914 |
+
{
|
| 6915 |
+
"epoch": 0.004387651450658925,
|
| 6916 |
+
"grad_norm": 8.8125,
|
| 6917 |
+
"learning_rate": 2.1916224896920396e-07,
|
| 6918 |
+
"loss": 3.1132,
|
| 6919 |
+
"step": 987
|
| 6920 |
+
},
|
| 6921 |
+
{
|
| 6922 |
+
"epoch": 0.004392096892858175,
|
| 6923 |
+
"grad_norm": 7.46875,
|
| 6924 |
+
"learning_rate": 2.1938452305537962e-07,
|
| 6925 |
+
"loss": 3.1818,
|
| 6926 |
+
"step": 988
|
| 6927 |
+
},
|
| 6928 |
+
{
|
| 6929 |
+
"epoch": 0.004396542335057424,
|
| 6930 |
+
"grad_norm": 11.3125,
|
| 6931 |
+
"learning_rate": 2.196067971415553e-07,
|
| 6932 |
+
"loss": 2.8631,
|
| 6933 |
+
"step": 989
|
| 6934 |
+
},
|
| 6935 |
+
{
|
| 6936 |
+
"epoch": 0.004400987777256673,
|
| 6937 |
+
"grad_norm": 11.5,
|
| 6938 |
+
"learning_rate": 2.1982907122773094e-07,
|
| 6939 |
+
"loss": 2.8712,
|
| 6940 |
+
"step": 990
|
| 6941 |
+
},
|
| 6942 |
+
{
|
| 6943 |
+
"epoch": 0.0044054332194559225,
|
| 6944 |
+
"grad_norm": 10.125,
|
| 6945 |
+
"learning_rate": 2.2005134531390657e-07,
|
| 6946 |
+
"loss": 2.9481,
|
| 6947 |
+
"step": 991
|
| 6948 |
+
},
|
| 6949 |
+
{
|
| 6950 |
+
"epoch": 0.0044098786616551715,
|
| 6951 |
+
"grad_norm": 8.9375,
|
| 6952 |
+
"learning_rate": 2.2027361940008225e-07,
|
| 6953 |
+
"loss": 3.0446,
|
| 6954 |
+
"step": 992
|
| 6955 |
+
},
|
| 6956 |
+
{
|
| 6957 |
+
"epoch": 0.004414324103854421,
|
| 6958 |
+
"grad_norm": 8.75,
|
| 6959 |
+
"learning_rate": 2.204958934862579e-07,
|
| 6960 |
+
"loss": 3.1951,
|
| 6961 |
+
"step": 993
|
| 6962 |
+
},
|
| 6963 |
+
{
|
| 6964 |
+
"epoch": 0.00441876954605367,
|
| 6965 |
+
"grad_norm": 10.8125,
|
| 6966 |
+
"learning_rate": 2.2071816757243358e-07,
|
| 6967 |
+
"loss": 2.832,
|
| 6968 |
+
"step": 994
|
| 6969 |
+
},
|
| 6970 |
+
{
|
| 6971 |
+
"epoch": 0.004423214988252919,
|
| 6972 |
+
"grad_norm": 8.8125,
|
| 6973 |
+
"learning_rate": 2.2094044165860925e-07,
|
| 6974 |
+
"loss": 3.1672,
|
| 6975 |
+
"step": 995
|
| 6976 |
+
},
|
| 6977 |
+
{
|
| 6978 |
+
"epoch": 0.004427660430452168,
|
| 6979 |
+
"grad_norm": 10.6875,
|
| 6980 |
+
"learning_rate": 2.211627157447849e-07,
|
| 6981 |
+
"loss": 2.9116,
|
| 6982 |
+
"step": 996
|
| 6983 |
+
},
|
| 6984 |
+
{
|
| 6985 |
+
"epoch": 0.004432105872651418,
|
| 6986 |
+
"grad_norm": 9.6875,
|
| 6987 |
+
"learning_rate": 2.2138498983096058e-07,
|
| 6988 |
+
"loss": 3.1124,
|
| 6989 |
+
"step": 997
|
| 6990 |
+
},
|
| 6991 |
+
{
|
| 6992 |
+
"epoch": 0.004436551314850667,
|
| 6993 |
+
"grad_norm": 9.4375,
|
| 6994 |
+
"learning_rate": 2.2160726391713623e-07,
|
| 6995 |
+
"loss": 3.0419,
|
| 6996 |
+
"step": 998
|
| 6997 |
+
},
|
| 6998 |
+
{
|
| 6999 |
+
"epoch": 0.004440996757049916,
|
| 7000 |
+
"grad_norm": 9.0625,
|
| 7001 |
+
"learning_rate": 2.218295380033119e-07,
|
| 7002 |
+
"loss": 3.1185,
|
| 7003 |
+
"step": 999
|
| 7004 |
+
},
|
| 7005 |
+
{
|
| 7006 |
+
"epoch": 0.004445442199249165,
|
| 7007 |
+
"grad_norm": 9.875,
|
| 7008 |
+
"learning_rate": 2.2205181208948756e-07,
|
| 7009 |
+
"loss": 2.9247,
|
| 7010 |
+
"step": 1000
|
| 7011 |
+
},
|
| 7012 |
+
{
|
| 7013 |
+
"epoch": 0.004449887641448414,
|
| 7014 |
+
"grad_norm": 9.9375,
|
| 7015 |
+
"learning_rate": 2.2227408617566324e-07,
|
| 7016 |
+
"loss": 3.1356,
|
| 7017 |
+
"step": 1001
|
| 7018 |
+
},
|
| 7019 |
+
{
|
| 7020 |
+
"epoch": 0.004454333083647663,
|
| 7021 |
+
"grad_norm": 11.5625,
|
| 7022 |
+
"learning_rate": 2.224963602618389e-07,
|
| 7023 |
+
"loss": 2.8265,
|
| 7024 |
+
"step": 1002
|
| 7025 |
+
},
|
| 7026 |
+
{
|
| 7027 |
+
"epoch": 0.004458778525846912,
|
| 7028 |
+
"grad_norm": 10.875,
|
| 7029 |
+
"learning_rate": 2.2271863434801457e-07,
|
| 7030 |
+
"loss": 2.8969,
|
| 7031 |
+
"step": 1003
|
| 7032 |
+
},
|
| 7033 |
+
{
|
| 7034 |
+
"epoch": 0.004463223968046161,
|
| 7035 |
+
"grad_norm": 11.5,
|
| 7036 |
+
"learning_rate": 2.2294090843419022e-07,
|
| 7037 |
+
"loss": 2.7605,
|
| 7038 |
+
"step": 1004
|
| 7039 |
+
},
|
| 7040 |
+
{
|
| 7041 |
+
"epoch": 0.004467669410245411,
|
| 7042 |
+
"grad_norm": 10.0,
|
| 7043 |
+
"learning_rate": 2.231631825203659e-07,
|
| 7044 |
+
"loss": 2.9733,
|
| 7045 |
+
"step": 1005
|
| 7046 |
+
},
|
| 7047 |
+
{
|
| 7048 |
+
"epoch": 0.00447211485244466,
|
| 7049 |
+
"grad_norm": 9.4375,
|
| 7050 |
+
"learning_rate": 2.2338545660654155e-07,
|
| 7051 |
+
"loss": 2.9336,
|
| 7052 |
+
"step": 1006
|
| 7053 |
+
},
|
| 7054 |
+
{
|
| 7055 |
+
"epoch": 0.004476560294643909,
|
| 7056 |
+
"grad_norm": 11.5,
|
| 7057 |
+
"learning_rate": 2.236077306927172e-07,
|
| 7058 |
+
"loss": 2.7993,
|
| 7059 |
+
"step": 1007
|
| 7060 |
+
},
|
| 7061 |
+
{
|
| 7062 |
+
"epoch": 0.004481005736843158,
|
| 7063 |
+
"grad_norm": 11.25,
|
| 7064 |
+
"learning_rate": 2.2383000477889285e-07,
|
| 7065 |
+
"loss": 2.8964,
|
| 7066 |
+
"step": 1008
|
| 7067 |
+
},
|
| 7068 |
+
{
|
| 7069 |
+
"epoch": 0.004485451179042407,
|
| 7070 |
+
"grad_norm": 9.75,
|
| 7071 |
+
"learning_rate": 2.2405227886506853e-07,
|
| 7072 |
+
"loss": 2.9988,
|
| 7073 |
+
"step": 1009
|
| 7074 |
+
},
|
| 7075 |
+
{
|
| 7076 |
+
"epoch": 0.004489896621241656,
|
| 7077 |
+
"grad_norm": 9.625,
|
| 7078 |
+
"learning_rate": 2.2427455295124418e-07,
|
| 7079 |
+
"loss": 3.01,
|
| 7080 |
+
"step": 1010
|
| 7081 |
+
},
|
| 7082 |
+
{
|
| 7083 |
+
"epoch": 0.0044943420634409054,
|
| 7084 |
+
"grad_norm": 7.9375,
|
| 7085 |
+
"learning_rate": 2.2449682703741986e-07,
|
| 7086 |
+
"loss": 3.2394,
|
| 7087 |
+
"step": 1011
|
| 7088 |
+
},
|
| 7089 |
+
{
|
| 7090 |
+
"epoch": 0.0044987875056401545,
|
| 7091 |
+
"grad_norm": 9.125,
|
| 7092 |
+
"learning_rate": 2.247191011235955e-07,
|
| 7093 |
+
"loss": 3.0721,
|
| 7094 |
+
"step": 1012
|
| 7095 |
+
},
|
| 7096 |
+
{
|
| 7097 |
+
"epoch": 0.0045032329478394036,
|
| 7098 |
+
"grad_norm": 10.75,
|
| 7099 |
+
"learning_rate": 2.249413752097712e-07,
|
| 7100 |
+
"loss": 2.9398,
|
| 7101 |
+
"step": 1013
|
| 7102 |
+
},
|
| 7103 |
+
{
|
| 7104 |
+
"epoch": 0.0045076783900386535,
|
| 7105 |
+
"grad_norm": 9.8125,
|
| 7106 |
+
"learning_rate": 2.2516364929594684e-07,
|
| 7107 |
+
"loss": 3.0119,
|
| 7108 |
+
"step": 1014
|
| 7109 |
+
},
|
| 7110 |
+
{
|
| 7111 |
+
"epoch": 0.0045121238322379026,
|
| 7112 |
+
"grad_norm": 10.25,
|
| 7113 |
+
"learning_rate": 2.2538592338212252e-07,
|
| 7114 |
+
"loss": 3.0231,
|
| 7115 |
+
"step": 1015
|
| 7116 |
+
},
|
| 7117 |
+
{
|
| 7118 |
+
"epoch": 0.004516569274437152,
|
| 7119 |
+
"grad_norm": 10.3125,
|
| 7120 |
+
"learning_rate": 2.2560819746829817e-07,
|
| 7121 |
+
"loss": 2.9173,
|
| 7122 |
+
"step": 1016
|
| 7123 |
+
},
|
| 7124 |
+
{
|
| 7125 |
+
"epoch": 0.004521014716636401,
|
| 7126 |
+
"grad_norm": 11.4375,
|
| 7127 |
+
"learning_rate": 2.2583047155447385e-07,
|
| 7128 |
+
"loss": 2.9078,
|
| 7129 |
+
"step": 1017
|
| 7130 |
+
},
|
| 7131 |
+
{
|
| 7132 |
+
"epoch": 0.00452546015883565,
|
| 7133 |
+
"grad_norm": 10.875,
|
| 7134 |
+
"learning_rate": 2.260527456406495e-07,
|
| 7135 |
+
"loss": 2.9214,
|
| 7136 |
+
"step": 1018
|
| 7137 |
+
},
|
| 7138 |
+
{
|
| 7139 |
+
"epoch": 0.004529905601034899,
|
| 7140 |
+
"grad_norm": 11.0625,
|
| 7141 |
+
"learning_rate": 2.2627501972682518e-07,
|
| 7142 |
+
"loss": 2.8618,
|
| 7143 |
+
"step": 1019
|
| 7144 |
+
},
|
| 7145 |
+
{
|
| 7146 |
+
"epoch": 0.004534351043234148,
|
| 7147 |
+
"grad_norm": 9.125,
|
| 7148 |
+
"learning_rate": 2.2649729381300083e-07,
|
| 7149 |
+
"loss": 3.159,
|
| 7150 |
+
"step": 1020
|
| 7151 |
+
},
|
| 7152 |
+
{
|
| 7153 |
+
"epoch": 0.004538796485433397,
|
| 7154 |
+
"grad_norm": 7.6875,
|
| 7155 |
+
"learning_rate": 2.267195678991765e-07,
|
| 7156 |
+
"loss": 3.1911,
|
| 7157 |
+
"step": 1021
|
| 7158 |
+
},
|
| 7159 |
+
{
|
| 7160 |
+
"epoch": 0.004543241927632647,
|
| 7161 |
+
"grad_norm": 9.0,
|
| 7162 |
+
"learning_rate": 2.2694184198535218e-07,
|
| 7163 |
+
"loss": 3.1503,
|
| 7164 |
+
"step": 1022
|
| 7165 |
+
},
|
| 7166 |
+
{
|
| 7167 |
+
"epoch": 0.004547687369831896,
|
| 7168 |
+
"grad_norm": 8.1875,
|
| 7169 |
+
"learning_rate": 2.2716411607152784e-07,
|
| 7170 |
+
"loss": 3.0955,
|
| 7171 |
+
"step": 1023
|
| 7172 |
+
},
|
| 7173 |
+
{
|
| 7174 |
+
"epoch": 0.004552132812031145,
|
| 7175 |
+
"grad_norm": 10.25,
|
| 7176 |
+
"learning_rate": 2.2738639015770346e-07,
|
| 7177 |
+
"loss": 2.8884,
|
| 7178 |
+
"step": 1024
|
| 7179 |
+
},
|
| 7180 |
+
{
|
| 7181 |
+
"epoch": 0.004556578254230394,
|
| 7182 |
+
"grad_norm": 10.25,
|
| 7183 |
+
"learning_rate": 2.2760866424387914e-07,
|
| 7184 |
+
"loss": 3.0228,
|
| 7185 |
+
"step": 1025
|
| 7186 |
+
},
|
| 7187 |
+
{
|
| 7188 |
+
"epoch": 0.004561023696429643,
|
| 7189 |
+
"grad_norm": 9.6875,
|
| 7190 |
+
"learning_rate": 2.278309383300548e-07,
|
| 7191 |
+
"loss": 3.0518,
|
| 7192 |
+
"step": 1026
|
| 7193 |
+
},
|
| 7194 |
+
{
|
| 7195 |
+
"epoch": 0.004565469138628892,
|
| 7196 |
+
"grad_norm": 11.1875,
|
| 7197 |
+
"learning_rate": 2.2805321241623047e-07,
|
| 7198 |
+
"loss": 2.8747,
|
| 7199 |
+
"step": 1027
|
| 7200 |
+
},
|
| 7201 |
+
{
|
| 7202 |
+
"epoch": 0.004569914580828141,
|
| 7203 |
+
"grad_norm": 9.1875,
|
| 7204 |
+
"learning_rate": 2.2827548650240612e-07,
|
| 7205 |
+
"loss": 3.0202,
|
| 7206 |
+
"step": 1028
|
| 7207 |
+
},
|
| 7208 |
+
{
|
| 7209 |
+
"epoch": 0.00457436002302739,
|
| 7210 |
+
"grad_norm": 9.125,
|
| 7211 |
+
"learning_rate": 2.284977605885818e-07,
|
| 7212 |
+
"loss": 3.0268,
|
| 7213 |
+
"step": 1029
|
| 7214 |
+
},
|
| 7215 |
+
{
|
| 7216 |
+
"epoch": 0.004578805465226639,
|
| 7217 |
+
"grad_norm": 8.5,
|
| 7218 |
+
"learning_rate": 2.2872003467475745e-07,
|
| 7219 |
+
"loss": 3.1001,
|
| 7220 |
+
"step": 1030
|
| 7221 |
+
},
|
| 7222 |
+
{
|
| 7223 |
+
"epoch": 0.004583250907425889,
|
| 7224 |
+
"grad_norm": 11.4375,
|
| 7225 |
+
"learning_rate": 2.2894230876093313e-07,
|
| 7226 |
+
"loss": 2.7481,
|
| 7227 |
+
"step": 1031
|
| 7228 |
+
},
|
| 7229 |
+
{
|
| 7230 |
+
"epoch": 0.004587696349625138,
|
| 7231 |
+
"grad_norm": 11.0,
|
| 7232 |
+
"learning_rate": 2.2916458284710878e-07,
|
| 7233 |
+
"loss": 2.8899,
|
| 7234 |
+
"step": 1032
|
| 7235 |
+
},
|
| 7236 |
+
{
|
| 7237 |
+
"epoch": 0.004592141791824387,
|
| 7238 |
+
"grad_norm": 9.375,
|
| 7239 |
+
"learning_rate": 2.2938685693328445e-07,
|
| 7240 |
+
"loss": 3.1019,
|
| 7241 |
+
"step": 1033
|
| 7242 |
+
},
|
| 7243 |
+
{
|
| 7244 |
+
"epoch": 0.0045965872340236365,
|
| 7245 |
+
"grad_norm": 11.875,
|
| 7246 |
+
"learning_rate": 2.2960913101946013e-07,
|
| 7247 |
+
"loss": 3.0228,
|
| 7248 |
+
"step": 1034
|
| 7249 |
+
},
|
| 7250 |
+
{
|
| 7251 |
+
"epoch": 0.0046010326762228855,
|
| 7252 |
+
"grad_norm": 9.1875,
|
| 7253 |
+
"learning_rate": 2.2983140510563578e-07,
|
| 7254 |
+
"loss": 3.0717,
|
| 7255 |
+
"step": 1035
|
| 7256 |
+
},
|
| 7257 |
+
{
|
| 7258 |
+
"epoch": 0.004605478118422135,
|
| 7259 |
+
"grad_norm": 12.75,
|
| 7260 |
+
"learning_rate": 2.3005367919181146e-07,
|
| 7261 |
+
"loss": 2.6542,
|
| 7262 |
+
"step": 1036
|
| 7263 |
+
},
|
| 7264 |
+
{
|
| 7265 |
+
"epoch": 0.004609923560621384,
|
| 7266 |
+
"grad_norm": 11.125,
|
| 7267 |
+
"learning_rate": 2.302759532779871e-07,
|
| 7268 |
+
"loss": 2.9248,
|
| 7269 |
+
"step": 1037
|
| 7270 |
+
},
|
| 7271 |
+
{
|
| 7272 |
+
"epoch": 0.004614369002820633,
|
| 7273 |
+
"grad_norm": 9.1875,
|
| 7274 |
+
"learning_rate": 2.304982273641628e-07,
|
| 7275 |
+
"loss": 3.1013,
|
| 7276 |
+
"step": 1038
|
| 7277 |
+
},
|
| 7278 |
+
{
|
| 7279 |
+
"epoch": 0.004618814445019883,
|
| 7280 |
+
"grad_norm": 10.4375,
|
| 7281 |
+
"learning_rate": 2.3072050145033844e-07,
|
| 7282 |
+
"loss": 2.9626,
|
| 7283 |
+
"step": 1039
|
| 7284 |
+
},
|
| 7285 |
+
{
|
| 7286 |
+
"epoch": 0.004623259887219132,
|
| 7287 |
+
"grad_norm": 10.0625,
|
| 7288 |
+
"learning_rate": 2.3094277553651407e-07,
|
| 7289 |
+
"loss": 2.9294,
|
| 7290 |
+
"step": 1040
|
| 7291 |
+
},
|
| 7292 |
+
{
|
| 7293 |
+
"epoch": 0.004627705329418381,
|
| 7294 |
+
"grad_norm": 10.3125,
|
| 7295 |
+
"learning_rate": 2.3116504962268974e-07,
|
| 7296 |
+
"loss": 2.9481,
|
| 7297 |
+
"step": 1041
|
| 7298 |
+
},
|
| 7299 |
+
{
|
| 7300 |
+
"epoch": 0.00463215077161763,
|
| 7301 |
+
"grad_norm": 11.0,
|
| 7302 |
+
"learning_rate": 2.313873237088654e-07,
|
| 7303 |
+
"loss": 2.9254,
|
| 7304 |
+
"step": 1042
|
| 7305 |
+
},
|
| 7306 |
+
{
|
| 7307 |
+
"epoch": 0.004636596213816879,
|
| 7308 |
+
"grad_norm": 8.3125,
|
| 7309 |
+
"learning_rate": 2.3160959779504107e-07,
|
| 7310 |
+
"loss": 3.2911,
|
| 7311 |
+
"step": 1043
|
| 7312 |
+
},
|
| 7313 |
+
{
|
| 7314 |
+
"epoch": 0.004641041656016128,
|
| 7315 |
+
"grad_norm": 10.5,
|
| 7316 |
+
"learning_rate": 2.3183187188121675e-07,
|
| 7317 |
+
"loss": 3.117,
|
| 7318 |
+
"step": 1044
|
| 7319 |
+
},
|
| 7320 |
+
{
|
| 7321 |
+
"epoch": 0.004645487098215377,
|
| 7322 |
+
"grad_norm": 9.0,
|
| 7323 |
+
"learning_rate": 2.320541459673924e-07,
|
| 7324 |
+
"loss": 3.0878,
|
| 7325 |
+
"step": 1045
|
| 7326 |
+
},
|
| 7327 |
+
{
|
| 7328 |
+
"epoch": 0.004649932540414626,
|
| 7329 |
+
"grad_norm": 9.0625,
|
| 7330 |
+
"learning_rate": 2.3227642005356808e-07,
|
| 7331 |
+
"loss": 2.9498,
|
| 7332 |
+
"step": 1046
|
| 7333 |
+
},
|
| 7334 |
+
{
|
| 7335 |
+
"epoch": 0.004654377982613875,
|
| 7336 |
+
"grad_norm": 9.125,
|
| 7337 |
+
"learning_rate": 2.3249869413974373e-07,
|
| 7338 |
+
"loss": 3.2355,
|
| 7339 |
+
"step": 1047
|
| 7340 |
+
},
|
| 7341 |
+
{
|
| 7342 |
+
"epoch": 0.004658823424813125,
|
| 7343 |
+
"grad_norm": 9.3125,
|
| 7344 |
+
"learning_rate": 2.327209682259194e-07,
|
| 7345 |
+
"loss": 3.093,
|
| 7346 |
+
"step": 1048
|
| 7347 |
+
},
|
| 7348 |
+
{
|
| 7349 |
+
"epoch": 0.004663268867012374,
|
| 7350 |
+
"grad_norm": 9.5625,
|
| 7351 |
+
"learning_rate": 2.3294324231209506e-07,
|
| 7352 |
+
"loss": 3.1023,
|
| 7353 |
+
"step": 1049
|
| 7354 |
+
},
|
| 7355 |
+
{
|
| 7356 |
+
"epoch": 0.004667714309211623,
|
| 7357 |
+
"grad_norm": 8.25,
|
| 7358 |
+
"learning_rate": 2.3316551639827074e-07,
|
| 7359 |
+
"loss": 3.1607,
|
| 7360 |
+
"step": 1050
|
| 7361 |
+
},
|
| 7362 |
+
{
|
| 7363 |
+
"epoch": 0.004672159751410872,
|
| 7364 |
+
"grad_norm": 10.0,
|
| 7365 |
+
"learning_rate": 2.333877904844464e-07,
|
| 7366 |
+
"loss": 2.9663,
|
| 7367 |
+
"step": 1051
|
| 7368 |
+
},
|
| 7369 |
+
{
|
| 7370 |
+
"epoch": 0.004676605193610121,
|
| 7371 |
+
"grad_norm": 11.5625,
|
| 7372 |
+
"learning_rate": 2.3361006457062207e-07,
|
| 7373 |
+
"loss": 2.9628,
|
| 7374 |
+
"step": 1052
|
| 7375 |
+
},
|
| 7376 |
+
{
|
| 7377 |
+
"epoch": 0.00468105063580937,
|
| 7378 |
+
"grad_norm": 8.9375,
|
| 7379 |
+
"learning_rate": 2.3383233865679772e-07,
|
| 7380 |
+
"loss": 3.1418,
|
| 7381 |
+
"step": 1053
|
| 7382 |
+
},
|
| 7383 |
+
{
|
| 7384 |
+
"epoch": 0.0046854960780086195,
|
| 7385 |
+
"grad_norm": 10.75,
|
| 7386 |
+
"learning_rate": 2.340546127429734e-07,
|
| 7387 |
+
"loss": 2.8077,
|
| 7388 |
+
"step": 1054
|
| 7389 |
+
},
|
| 7390 |
+
{
|
| 7391 |
+
"epoch": 0.0046899415202078685,
|
| 7392 |
+
"grad_norm": 9.75,
|
| 7393 |
+
"learning_rate": 2.3427688682914905e-07,
|
| 7394 |
+
"loss": 3.0485,
|
| 7395 |
+
"step": 1055
|
| 7396 |
+
},
|
| 7397 |
+
{
|
| 7398 |
+
"epoch": 0.0046943869624071185,
|
| 7399 |
+
"grad_norm": 10.25,
|
| 7400 |
+
"learning_rate": 2.3449916091532473e-07,
|
| 7401 |
+
"loss": 3.0086,
|
| 7402 |
+
"step": 1056
|
| 7403 |
+
},
|
| 7404 |
+
{
|
| 7405 |
+
"epoch": 0.0046988324046063675,
|
| 7406 |
+
"grad_norm": 9.75,
|
| 7407 |
+
"learning_rate": 2.3472143500150035e-07,
|
| 7408 |
+
"loss": 3.0199,
|
| 7409 |
+
"step": 1057
|
| 7410 |
+
},
|
| 7411 |
+
{
|
| 7412 |
+
"epoch": 0.004703277846805617,
|
| 7413 |
+
"grad_norm": 9.0625,
|
| 7414 |
+
"learning_rate": 2.3494370908767603e-07,
|
| 7415 |
+
"loss": 2.9958,
|
| 7416 |
+
"step": 1058
|
| 7417 |
+
},
|
| 7418 |
+
{
|
| 7419 |
+
"epoch": 0.004707723289004866,
|
| 7420 |
+
"grad_norm": 9.0625,
|
| 7421 |
+
"learning_rate": 2.3516598317385168e-07,
|
| 7422 |
+
"loss": 3.1202,
|
| 7423 |
+
"step": 1059
|
| 7424 |
+
},
|
| 7425 |
+
{
|
| 7426 |
+
"epoch": 0.004712168731204115,
|
| 7427 |
+
"grad_norm": 10.0625,
|
| 7428 |
+
"learning_rate": 2.3538825726002736e-07,
|
| 7429 |
+
"loss": 2.9328,
|
| 7430 |
+
"step": 1060
|
| 7431 |
+
},
|
| 7432 |
+
{
|
| 7433 |
+
"epoch": 0.004716614173403364,
|
| 7434 |
+
"grad_norm": 11.9375,
|
| 7435 |
+
"learning_rate": 2.35610531346203e-07,
|
| 7436 |
+
"loss": 2.8503,
|
| 7437 |
+
"step": 1061
|
| 7438 |
+
},
|
| 7439 |
+
{
|
| 7440 |
+
"epoch": 0.004721059615602613,
|
| 7441 |
+
"grad_norm": 10.6875,
|
| 7442 |
+
"learning_rate": 2.358328054323787e-07,
|
| 7443 |
+
"loss": 2.8671,
|
| 7444 |
+
"step": 1062
|
| 7445 |
+
},
|
| 7446 |
+
{
|
| 7447 |
+
"epoch": 0.004725505057801862,
|
| 7448 |
+
"grad_norm": 9.4375,
|
| 7449 |
+
"learning_rate": 2.3605507951855434e-07,
|
| 7450 |
+
"loss": 3.127,
|
| 7451 |
+
"step": 1063
|
| 7452 |
+
},
|
| 7453 |
+
{
|
| 7454 |
+
"epoch": 0.004729950500001111,
|
| 7455 |
+
"grad_norm": 8.4375,
|
| 7456 |
+
"learning_rate": 2.3627735360473002e-07,
|
| 7457 |
+
"loss": 3.1994,
|
| 7458 |
+
"step": 1064
|
| 7459 |
+
},
|
| 7460 |
+
{
|
| 7461 |
+
"epoch": 0.004734395942200361,
|
| 7462 |
+
"grad_norm": 11.375,
|
| 7463 |
+
"learning_rate": 2.3649962769090567e-07,
|
| 7464 |
+
"loss": 2.8112,
|
| 7465 |
+
"step": 1065
|
| 7466 |
+
},
|
| 7467 |
+
{
|
| 7468 |
+
"epoch": 0.00473884138439961,
|
| 7469 |
+
"grad_norm": 9.75,
|
| 7470 |
+
"learning_rate": 2.3672190177708135e-07,
|
| 7471 |
+
"loss": 3.0305,
|
| 7472 |
+
"step": 1066
|
| 7473 |
+
},
|
| 7474 |
+
{
|
| 7475 |
+
"epoch": 0.004743286826598859,
|
| 7476 |
+
"grad_norm": 10.375,
|
| 7477 |
+
"learning_rate": 2.36944175863257e-07,
|
| 7478 |
+
"loss": 2.8869,
|
| 7479 |
+
"step": 1067
|
| 7480 |
+
},
|
| 7481 |
+
{
|
| 7482 |
+
"epoch": 0.004747732268798108,
|
| 7483 |
+
"grad_norm": 11.0625,
|
| 7484 |
+
"learning_rate": 2.3716644994943267e-07,
|
| 7485 |
+
"loss": 2.9174,
|
| 7486 |
+
"step": 1068
|
| 7487 |
+
},
|
| 7488 |
+
{
|
| 7489 |
+
"epoch": 0.004752177710997357,
|
| 7490 |
+
"grad_norm": 9.875,
|
| 7491 |
+
"learning_rate": 2.3738872403560833e-07,
|
| 7492 |
+
"loss": 3.0506,
|
| 7493 |
+
"step": 1069
|
| 7494 |
+
},
|
| 7495 |
+
{
|
| 7496 |
+
"epoch": 0.004756623153196606,
|
| 7497 |
+
"grad_norm": 10.0625,
|
| 7498 |
+
"learning_rate": 2.37610998121784e-07,
|
| 7499 |
+
"loss": 2.9781,
|
| 7500 |
+
"step": 1070
|
| 7501 |
+
},
|
| 7502 |
+
{
|
| 7503 |
+
"epoch": 0.004761068595395855,
|
| 7504 |
+
"grad_norm": 10.6875,
|
| 7505 |
+
"learning_rate": 2.3783327220795965e-07,
|
| 7506 |
+
"loss": 2.9292,
|
| 7507 |
+
"step": 1071
|
| 7508 |
+
},
|
| 7509 |
+
{
|
| 7510 |
+
"epoch": 0.004765514037595104,
|
| 7511 |
+
"grad_norm": 9.75,
|
| 7512 |
+
"learning_rate": 2.3805554629413533e-07,
|
| 7513 |
+
"loss": 2.9742,
|
| 7514 |
+
"step": 1072
|
| 7515 |
+
},
|
| 7516 |
+
{
|
| 7517 |
+
"epoch": 0.004769959479794354,
|
| 7518 |
+
"grad_norm": 11.5625,
|
| 7519 |
+
"learning_rate": 2.38277820380311e-07,
|
| 7520 |
+
"loss": 2.9182,
|
| 7521 |
+
"step": 1073
|
| 7522 |
+
},
|
| 7523 |
+
{
|
| 7524 |
+
"epoch": 0.004774404921993603,
|
| 7525 |
+
"grad_norm": 9.75,
|
| 7526 |
+
"learning_rate": 2.385000944664866e-07,
|
| 7527 |
+
"loss": 2.9846,
|
| 7528 |
+
"step": 1074
|
| 7529 |
+
},
|
| 7530 |
+
{
|
| 7531 |
+
"epoch": 0.004778850364192852,
|
| 7532 |
+
"grad_norm": 11.3125,
|
| 7533 |
+
"learning_rate": 2.387223685526623e-07,
|
| 7534 |
+
"loss": 2.8255,
|
| 7535 |
+
"step": 1075
|
| 7536 |
+
},
|
| 7537 |
+
{
|
| 7538 |
+
"epoch": 0.0047832958063921014,
|
| 7539 |
+
"grad_norm": 10.75,
|
| 7540 |
+
"learning_rate": 2.3894464263883796e-07,
|
| 7541 |
+
"loss": 3.0087,
|
| 7542 |
+
"step": 1076
|
| 7543 |
+
},
|
| 7544 |
+
{
|
| 7545 |
+
"epoch": 0.0047877412485913505,
|
| 7546 |
+
"grad_norm": 8.75,
|
| 7547 |
+
"learning_rate": 2.3916691672501364e-07,
|
| 7548 |
+
"loss": 3.033,
|
| 7549 |
+
"step": 1077
|
| 7550 |
+
},
|
| 7551 |
+
{
|
| 7552 |
+
"epoch": 0.0047921866907905996,
|
| 7553 |
+
"grad_norm": 10.625,
|
| 7554 |
+
"learning_rate": 2.3938919081118927e-07,
|
| 7555 |
+
"loss": 3.021,
|
| 7556 |
+
"step": 1078
|
| 7557 |
+
},
|
| 7558 |
+
{
|
| 7559 |
+
"epoch": 0.004796632132989849,
|
| 7560 |
+
"grad_norm": 10.625,
|
| 7561 |
+
"learning_rate": 2.3961146489736495e-07,
|
| 7562 |
+
"loss": 3.0047,
|
| 7563 |
+
"step": 1079
|
| 7564 |
+
},
|
| 7565 |
+
{
|
| 7566 |
+
"epoch": 0.004801077575189098,
|
| 7567 |
+
"grad_norm": 9.9375,
|
| 7568 |
+
"learning_rate": 2.398337389835406e-07,
|
| 7569 |
+
"loss": 3.1063,
|
| 7570 |
+
"step": 1080
|
| 7571 |
+
},
|
| 7572 |
+
{
|
| 7573 |
+
"epoch": 0.004805523017388347,
|
| 7574 |
+
"grad_norm": 12.375,
|
| 7575 |
+
"learning_rate": 2.400560130697163e-07,
|
| 7576 |
+
"loss": 2.7354,
|
| 7577 |
+
"step": 1081
|
| 7578 |
+
},
|
| 7579 |
+
{
|
| 7580 |
+
"epoch": 0.004809968459587597,
|
| 7581 |
+
"grad_norm": 9.8125,
|
| 7582 |
+
"learning_rate": 2.402782871558919e-07,
|
| 7583 |
+
"loss": 3.0045,
|
| 7584 |
+
"step": 1082
|
| 7585 |
+
},
|
| 7586 |
+
{
|
| 7587 |
+
"epoch": 0.004814413901786846,
|
| 7588 |
+
"grad_norm": 9.125,
|
| 7589 |
+
"learning_rate": 2.405005612420676e-07,
|
| 7590 |
+
"loss": 3.1526,
|
| 7591 |
+
"step": 1083
|
| 7592 |
+
},
|
| 7593 |
+
{
|
| 7594 |
+
"epoch": 0.004818859343986095,
|
| 7595 |
+
"grad_norm": 9.5,
|
| 7596 |
+
"learning_rate": 2.407228353282433e-07,
|
| 7597 |
+
"loss": 3.0959,
|
| 7598 |
+
"step": 1084
|
| 7599 |
+
},
|
| 7600 |
+
{
|
| 7601 |
+
"epoch": 0.004823304786185344,
|
| 7602 |
+
"grad_norm": 11.0,
|
| 7603 |
+
"learning_rate": 2.4094510941441896e-07,
|
| 7604 |
+
"loss": 2.8924,
|
| 7605 |
+
"step": 1085
|
| 7606 |
+
},
|
| 7607 |
+
{
|
| 7608 |
+
"epoch": 0.004827750228384593,
|
| 7609 |
+
"grad_norm": 9.5,
|
| 7610 |
+
"learning_rate": 2.4116738350059464e-07,
|
| 7611 |
+
"loss": 3.0039,
|
| 7612 |
+
"step": 1086
|
| 7613 |
+
},
|
| 7614 |
+
{
|
| 7615 |
+
"epoch": 0.004832195670583842,
|
| 7616 |
+
"grad_norm": 11.375,
|
| 7617 |
+
"learning_rate": 2.4138965758677026e-07,
|
| 7618 |
+
"loss": 2.7872,
|
| 7619 |
+
"step": 1087
|
| 7620 |
+
},
|
| 7621 |
+
{
|
| 7622 |
+
"epoch": 0.004836641112783091,
|
| 7623 |
+
"grad_norm": 9.25,
|
| 7624 |
+
"learning_rate": 2.4161193167294594e-07,
|
| 7625 |
+
"loss": 3.0193,
|
| 7626 |
+
"step": 1088
|
| 7627 |
+
},
|
| 7628 |
+
{
|
| 7629 |
+
"epoch": 0.00484108655498234,
|
| 7630 |
+
"grad_norm": 10.8125,
|
| 7631 |
+
"learning_rate": 2.418342057591216e-07,
|
| 7632 |
+
"loss": 2.8507,
|
| 7633 |
+
"step": 1089
|
| 7634 |
+
},
|
| 7635 |
+
{
|
| 7636 |
+
"epoch": 0.00484553199718159,
|
| 7637 |
+
"grad_norm": 9.4375,
|
| 7638 |
+
"learning_rate": 2.4205647984529724e-07,
|
| 7639 |
+
"loss": 2.9794,
|
| 7640 |
+
"step": 1090
|
| 7641 |
+
},
|
| 7642 |
+
{
|
| 7643 |
+
"epoch": 0.004849977439380839,
|
| 7644 |
+
"grad_norm": 7.46875,
|
| 7645 |
+
"learning_rate": 2.422787539314729e-07,
|
| 7646 |
+
"loss": 3.2601,
|
| 7647 |
+
"step": 1091
|
| 7648 |
+
},
|
| 7649 |
+
{
|
| 7650 |
+
"epoch": 0.004854422881580088,
|
| 7651 |
+
"grad_norm": 10.75,
|
| 7652 |
+
"learning_rate": 2.4250102801764854e-07,
|
| 7653 |
+
"loss": 2.9433,
|
| 7654 |
+
"step": 1092
|
| 7655 |
+
},
|
| 7656 |
+
{
|
| 7657 |
+
"epoch": 0.004858868323779337,
|
| 7658 |
+
"grad_norm": 9.625,
|
| 7659 |
+
"learning_rate": 2.427233021038242e-07,
|
| 7660 |
+
"loss": 3.0437,
|
| 7661 |
+
"step": 1093
|
| 7662 |
+
},
|
| 7663 |
+
{
|
| 7664 |
+
"epoch": 0.004863313765978586,
|
| 7665 |
+
"grad_norm": 10.9375,
|
| 7666 |
+
"learning_rate": 2.429455761899999e-07,
|
| 7667 |
+
"loss": 2.9793,
|
| 7668 |
+
"step": 1094
|
| 7669 |
+
},
|
| 7670 |
+
{
|
| 7671 |
+
"epoch": 0.004867759208177835,
|
| 7672 |
+
"grad_norm": 9.875,
|
| 7673 |
+
"learning_rate": 2.431678502761756e-07,
|
| 7674 |
+
"loss": 2.9553,
|
| 7675 |
+
"step": 1095
|
| 7676 |
+
},
|
| 7677 |
+
{
|
| 7678 |
+
"epoch": 0.004872204650377084,
|
| 7679 |
+
"grad_norm": 9.4375,
|
| 7680 |
+
"learning_rate": 2.4339012436235126e-07,
|
| 7681 |
+
"loss": 3.2021,
|
| 7682 |
+
"step": 1096
|
| 7683 |
+
},
|
| 7684 |
+
{
|
| 7685 |
+
"epoch": 0.0048766500925763335,
|
| 7686 |
+
"grad_norm": 9.6875,
|
| 7687 |
+
"learning_rate": 2.436123984485269e-07,
|
| 7688 |
+
"loss": 3.0393,
|
| 7689 |
+
"step": 1097
|
| 7690 |
+
},
|
| 7691 |
+
{
|
| 7692 |
+
"epoch": 0.0048810955347755825,
|
| 7693 |
+
"grad_norm": 10.0,
|
| 7694 |
+
"learning_rate": 2.4383467253470256e-07,
|
| 7695 |
+
"loss": 3.0392,
|
| 7696 |
+
"step": 1098
|
| 7697 |
+
},
|
| 7698 |
+
{
|
| 7699 |
+
"epoch": 0.0048855409769748325,
|
| 7700 |
+
"grad_norm": 10.25,
|
| 7701 |
+
"learning_rate": 2.4405694662087824e-07,
|
| 7702 |
+
"loss": 3.0669,
|
| 7703 |
+
"step": 1099
|
| 7704 |
+
},
|
| 7705 |
+
{
|
| 7706 |
+
"epoch": 0.0048899864191740815,
|
| 7707 |
+
"grad_norm": 9.0,
|
| 7708 |
+
"learning_rate": 2.442792207070539e-07,
|
| 7709 |
+
"loss": 3.1861,
|
| 7710 |
+
"step": 1100
|
| 7711 |
+
},
|
| 7712 |
+
{
|
| 7713 |
+
"epoch": 0.004894431861373331,
|
| 7714 |
+
"grad_norm": 12.75,
|
| 7715 |
+
"learning_rate": 2.4450149479322954e-07,
|
| 7716 |
+
"loss": 2.8885,
|
| 7717 |
+
"step": 1101
|
| 7718 |
+
},
|
| 7719 |
+
{
|
| 7720 |
+
"epoch": 0.00489887730357258,
|
| 7721 |
+
"grad_norm": 10.0625,
|
| 7722 |
+
"learning_rate": 2.447237688794052e-07,
|
| 7723 |
+
"loss": 3.0389,
|
| 7724 |
+
"step": 1102
|
| 7725 |
+
},
|
| 7726 |
+
{
|
| 7727 |
+
"epoch": 0.004903322745771829,
|
| 7728 |
+
"grad_norm": 8.5,
|
| 7729 |
+
"learning_rate": 2.449460429655809e-07,
|
| 7730 |
+
"loss": 3.0667,
|
| 7731 |
+
"step": 1103
|
| 7732 |
+
},
|
| 7733 |
+
{
|
| 7734 |
+
"epoch": 0.004907768187971078,
|
| 7735 |
+
"grad_norm": 8.875,
|
| 7736 |
+
"learning_rate": 2.4516831705175657e-07,
|
| 7737 |
+
"loss": 3.2318,
|
| 7738 |
+
"step": 1104
|
| 7739 |
+
},
|
| 7740 |
+
{
|
| 7741 |
+
"epoch": 0.004912213630170327,
|
| 7742 |
+
"grad_norm": 8.4375,
|
| 7743 |
+
"learning_rate": 2.453905911379322e-07,
|
| 7744 |
+
"loss": 3.1945,
|
| 7745 |
+
"step": 1105
|
| 7746 |
+
},
|
| 7747 |
+
{
|
| 7748 |
+
"epoch": 0.004916659072369576,
|
| 7749 |
+
"grad_norm": 8.875,
|
| 7750 |
+
"learning_rate": 2.456128652241079e-07,
|
| 7751 |
+
"loss": 3.0713,
|
| 7752 |
+
"step": 1106
|
| 7753 |
+
},
|
| 7754 |
+
{
|
| 7755 |
+
"epoch": 0.004921104514568826,
|
| 7756 |
+
"grad_norm": 8.3125,
|
| 7757 |
+
"learning_rate": 2.458351393102835e-07,
|
| 7758 |
+
"loss": 3.2106,
|
| 7759 |
+
"step": 1107
|
| 7760 |
+
},
|
| 7761 |
+
{
|
| 7762 |
+
"epoch": 0.004925549956768075,
|
| 7763 |
+
"grad_norm": 9.4375,
|
| 7764 |
+
"learning_rate": 2.460574133964592e-07,
|
| 7765 |
+
"loss": 3.0079,
|
| 7766 |
+
"step": 1108
|
| 7767 |
+
},
|
| 7768 |
+
{
|
| 7769 |
+
"epoch": 0.004929995398967324,
|
| 7770 |
+
"grad_norm": 11.1875,
|
| 7771 |
+
"learning_rate": 2.4627968748263486e-07,
|
| 7772 |
+
"loss": 2.8788,
|
| 7773 |
+
"step": 1109
|
| 7774 |
+
},
|
| 7775 |
+
{
|
| 7776 |
+
"epoch": 0.004934440841166573,
|
| 7777 |
+
"grad_norm": 10.1875,
|
| 7778 |
+
"learning_rate": 2.4650196156881053e-07,
|
| 7779 |
+
"loss": 2.9214,
|
| 7780 |
+
"step": 1110
|
| 7781 |
+
},
|
| 7782 |
+
{
|
| 7783 |
+
"epoch": 0.004938886283365822,
|
| 7784 |
+
"grad_norm": 9.5,
|
| 7785 |
+
"learning_rate": 2.4672423565498616e-07,
|
| 7786 |
+
"loss": 3.0695,
|
| 7787 |
+
"step": 1111
|
| 7788 |
+
},
|
| 7789 |
+
{
|
| 7790 |
+
"epoch": 0.004943331725565071,
|
| 7791 |
+
"grad_norm": 8.8125,
|
| 7792 |
+
"learning_rate": 2.4694650974116184e-07,
|
| 7793 |
+
"loss": 3.0699,
|
| 7794 |
+
"step": 1112
|
| 7795 |
+
},
|
| 7796 |
+
{
|
| 7797 |
+
"epoch": 0.00494777716776432,
|
| 7798 |
+
"grad_norm": 10.625,
|
| 7799 |
+
"learning_rate": 2.471687838273375e-07,
|
| 7800 |
+
"loss": 2.8647,
|
| 7801 |
+
"step": 1113
|
| 7802 |
+
},
|
| 7803 |
+
{
|
| 7804 |
+
"epoch": 0.004952222609963569,
|
| 7805 |
+
"grad_norm": 9.25,
|
| 7806 |
+
"learning_rate": 2.473910579135132e-07,
|
| 7807 |
+
"loss": 3.187,
|
| 7808 |
+
"step": 1114
|
| 7809 |
+
},
|
| 7810 |
+
{
|
| 7811 |
+
"epoch": 0.004956668052162818,
|
| 7812 |
+
"grad_norm": 7.65625,
|
| 7813 |
+
"learning_rate": 2.476133319996888e-07,
|
| 7814 |
+
"loss": 3.2968,
|
| 7815 |
+
"step": 1115
|
| 7816 |
+
},
|
| 7817 |
+
{
|
| 7818 |
+
"epoch": 0.004961113494362068,
|
| 7819 |
+
"grad_norm": 8.0625,
|
| 7820 |
+
"learning_rate": 2.478356060858645e-07,
|
| 7821 |
+
"loss": 3.2926,
|
| 7822 |
+
"step": 1116
|
| 7823 |
+
},
|
| 7824 |
+
{
|
| 7825 |
+
"epoch": 0.004965558936561317,
|
| 7826 |
+
"grad_norm": 9.4375,
|
| 7827 |
+
"learning_rate": 2.4805788017204017e-07,
|
| 7828 |
+
"loss": 3.079,
|
| 7829 |
+
"step": 1117
|
| 7830 |
+
},
|
| 7831 |
+
{
|
| 7832 |
+
"epoch": 0.004970004378760566,
|
| 7833 |
+
"grad_norm": 8.875,
|
| 7834 |
+
"learning_rate": 2.4828015425821585e-07,
|
| 7835 |
+
"loss": 3.1489,
|
| 7836 |
+
"step": 1118
|
| 7837 |
+
},
|
| 7838 |
+
{
|
| 7839 |
+
"epoch": 0.0049744498209598155,
|
| 7840 |
+
"grad_norm": 11.3125,
|
| 7841 |
+
"learning_rate": 2.485024283443915e-07,
|
| 7842 |
+
"loss": 2.8599,
|
| 7843 |
+
"step": 1119
|
| 7844 |
+
},
|
| 7845 |
+
{
|
| 7846 |
+
"epoch": 0.0049788952631590645,
|
| 7847 |
+
"grad_norm": 8.8125,
|
| 7848 |
+
"learning_rate": 2.4872470243056715e-07,
|
| 7849 |
+
"loss": 3.1747,
|
| 7850 |
+
"step": 1120
|
| 7851 |
+
},
|
| 7852 |
+
{
|
| 7853 |
+
"epoch": 0.004983340705358314,
|
| 7854 |
+
"grad_norm": 9.5,
|
| 7855 |
+
"learning_rate": 2.4894697651674283e-07,
|
| 7856 |
+
"loss": 3.0296,
|
| 7857 |
+
"step": 1121
|
| 7858 |
+
},
|
| 7859 |
+
{
|
| 7860 |
+
"epoch": 0.004987786147557563,
|
| 7861 |
+
"grad_norm": 9.5,
|
| 7862 |
+
"learning_rate": 2.491692506029185e-07,
|
| 7863 |
+
"loss": 3.054,
|
| 7864 |
+
"step": 1122
|
| 7865 |
+
},
|
| 7866 |
+
{
|
| 7867 |
+
"epoch": 0.004992231589756812,
|
| 7868 |
+
"grad_norm": 9.6875,
|
| 7869 |
+
"learning_rate": 2.493915246890942e-07,
|
| 7870 |
+
"loss": 3.0177,
|
| 7871 |
+
"step": 1123
|
| 7872 |
+
},
|
| 7873 |
+
{
|
| 7874 |
+
"epoch": 0.004996677031956062,
|
| 7875 |
+
"grad_norm": 9.375,
|
| 7876 |
+
"learning_rate": 2.496137987752698e-07,
|
| 7877 |
+
"loss": 3.0309,
|
| 7878 |
+
"step": 1124
|
| 7879 |
+
},
|
| 7880 |
+
{
|
| 7881 |
+
"epoch": 0.005001122474155311,
|
| 7882 |
+
"grad_norm": 9.0625,
|
| 7883 |
+
"learning_rate": 2.4983607286144544e-07,
|
| 7884 |
+
"loss": 3.1507,
|
| 7885 |
+
"step": 1125
|
| 7886 |
+
},
|
| 7887 |
+
{
|
| 7888 |
+
"epoch": 0.00500556791635456,
|
| 7889 |
+
"grad_norm": 7.09375,
|
| 7890 |
+
"learning_rate": 2.500583469476211e-07,
|
| 7891 |
+
"loss": 3.2147,
|
| 7892 |
+
"step": 1126
|
| 7893 |
+
},
|
| 7894 |
+
{
|
| 7895 |
+
"epoch": 0.005010013358553809,
|
| 7896 |
+
"grad_norm": 8.5,
|
| 7897 |
+
"learning_rate": 2.502806210337968e-07,
|
| 7898 |
+
"loss": 3.2409,
|
| 7899 |
+
"step": 1127
|
| 7900 |
+
},
|
| 7901 |
+
{
|
| 7902 |
+
"epoch": 0.005014458800753058,
|
| 7903 |
+
"grad_norm": 11.5625,
|
| 7904 |
+
"learning_rate": 2.5050289511997247e-07,
|
| 7905 |
+
"loss": 2.8354,
|
| 7906 |
+
"step": 1128
|
| 7907 |
+
},
|
| 7908 |
+
{
|
| 7909 |
+
"epoch": 0.005018904242952307,
|
| 7910 |
+
"grad_norm": 8.3125,
|
| 7911 |
+
"learning_rate": 2.507251692061481e-07,
|
| 7912 |
+
"loss": 3.1717,
|
| 7913 |
+
"step": 1129
|
| 7914 |
+
},
|
| 7915 |
+
{
|
| 7916 |
+
"epoch": 0.005023349685151556,
|
| 7917 |
+
"grad_norm": 8.6875,
|
| 7918 |
+
"learning_rate": 2.5094744329232377e-07,
|
| 7919 |
+
"loss": 3.2802,
|
| 7920 |
+
"step": 1130
|
| 7921 |
+
},
|
| 7922 |
+
{
|
| 7923 |
+
"epoch": 0.005027795127350805,
|
| 7924 |
+
"grad_norm": 7.875,
|
| 7925 |
+
"learning_rate": 2.5116971737849945e-07,
|
| 7926 |
+
"loss": 3.3287,
|
| 7927 |
+
"step": 1131
|
| 7928 |
+
},
|
| 7929 |
+
{
|
| 7930 |
+
"epoch": 0.005032240569550054,
|
| 7931 |
+
"grad_norm": 6.25,
|
| 7932 |
+
"learning_rate": 2.5139199146467513e-07,
|
| 7933 |
+
"loss": 3.3918,
|
| 7934 |
+
"step": 1132
|
| 7935 |
+
},
|
| 7936 |
+
{
|
| 7937 |
+
"epoch": 0.005036686011749304,
|
| 7938 |
+
"grad_norm": 10.3125,
|
| 7939 |
+
"learning_rate": 2.5161426555085075e-07,
|
| 7940 |
+
"loss": 3.2482,
|
| 7941 |
+
"step": 1133
|
| 7942 |
+
},
|
| 7943 |
+
{
|
| 7944 |
+
"epoch": 0.005041131453948553,
|
| 7945 |
+
"grad_norm": 7.1875,
|
| 7946 |
+
"learning_rate": 2.5183653963702643e-07,
|
| 7947 |
+
"loss": 3.2685,
|
| 7948 |
+
"step": 1134
|
| 7949 |
+
},
|
| 7950 |
+
{
|
| 7951 |
+
"epoch": 0.005045576896147802,
|
| 7952 |
+
"grad_norm": 10.5,
|
| 7953 |
+
"learning_rate": 2.520588137232021e-07,
|
| 7954 |
+
"loss": 2.9439,
|
| 7955 |
+
"step": 1135
|
| 7956 |
+
},
|
| 7957 |
+
{
|
| 7958 |
+
"epoch": 0.005050022338347051,
|
| 7959 |
+
"grad_norm": 11.5,
|
| 7960 |
+
"learning_rate": 2.522810878093778e-07,
|
| 7961 |
+
"loss": 2.8252,
|
| 7962 |
+
"step": 1136
|
| 7963 |
+
},
|
| 7964 |
+
{
|
| 7965 |
+
"epoch": 0.0050544677805463,
|
| 7966 |
+
"grad_norm": 7.6875,
|
| 7967 |
+
"learning_rate": 2.5250336189555346e-07,
|
| 7968 |
+
"loss": 3.297,
|
| 7969 |
+
"step": 1137
|
| 7970 |
+
},
|
| 7971 |
+
{
|
| 7972 |
+
"epoch": 0.005058913222745549,
|
| 7973 |
+
"grad_norm": 10.375,
|
| 7974 |
+
"learning_rate": 2.527256359817291e-07,
|
| 7975 |
+
"loss": 2.9264,
|
| 7976 |
+
"step": 1138
|
| 7977 |
+
},
|
| 7978 |
+
{
|
| 7979 |
+
"epoch": 0.0050633586649447984,
|
| 7980 |
+
"grad_norm": 6.59375,
|
| 7981 |
+
"learning_rate": 2.5294791006790477e-07,
|
| 7982 |
+
"loss": 3.3846,
|
| 7983 |
+
"step": 1139
|
| 7984 |
+
},
|
| 7985 |
+
{
|
| 7986 |
+
"epoch": 0.0050678041071440475,
|
| 7987 |
+
"grad_norm": 9.875,
|
| 7988 |
+
"learning_rate": 2.531701841540804e-07,
|
| 7989 |
+
"loss": 3.038,
|
| 7990 |
+
"step": 1140
|
| 7991 |
+
},
|
| 7992 |
+
{
|
| 7993 |
+
"epoch": 0.0050722495493432974,
|
| 7994 |
+
"grad_norm": 10.6875,
|
| 7995 |
+
"learning_rate": 2.5339245824025607e-07,
|
| 7996 |
+
"loss": 2.8303,
|
| 7997 |
+
"step": 1141
|
| 7998 |
+
},
|
| 7999 |
+
{
|
| 8000 |
+
"epoch": 0.0050766949915425465,
|
| 8001 |
+
"grad_norm": 9.9375,
|
| 8002 |
+
"learning_rate": 2.5361473232643175e-07,
|
| 8003 |
+
"loss": 2.967,
|
| 8004 |
+
"step": 1142
|
| 8005 |
+
},
|
| 8006 |
+
{
|
| 8007 |
+
"epoch": 0.005081140433741796,
|
| 8008 |
+
"grad_norm": 9.8125,
|
| 8009 |
+
"learning_rate": 2.5383700641260737e-07,
|
| 8010 |
+
"loss": 3.1252,
|
| 8011 |
+
"step": 1143
|
| 8012 |
+
},
|
| 8013 |
+
{
|
| 8014 |
+
"epoch": 0.005085585875941045,
|
| 8015 |
+
"grad_norm": 9.5,
|
| 8016 |
+
"learning_rate": 2.5405928049878305e-07,
|
| 8017 |
+
"loss": 3.0528,
|
| 8018 |
+
"step": 1144
|
| 8019 |
+
},
|
| 8020 |
+
{
|
| 8021 |
+
"epoch": 0.005090031318140294,
|
| 8022 |
+
"grad_norm": 9.9375,
|
| 8023 |
+
"learning_rate": 2.5428155458495873e-07,
|
| 8024 |
+
"loss": 3.0946,
|
| 8025 |
+
"step": 1145
|
| 8026 |
+
},
|
| 8027 |
+
{
|
| 8028 |
+
"epoch": 0.005094476760339543,
|
| 8029 |
+
"grad_norm": 10.25,
|
| 8030 |
+
"learning_rate": 2.545038286711344e-07,
|
| 8031 |
+
"loss": 2.976,
|
| 8032 |
+
"step": 1146
|
| 8033 |
+
},
|
| 8034 |
+
{
|
| 8035 |
+
"epoch": 0.005098922202538792,
|
| 8036 |
+
"grad_norm": 11.25,
|
| 8037 |
+
"learning_rate": 2.547261027573101e-07,
|
| 8038 |
+
"loss": 2.8889,
|
| 8039 |
+
"step": 1147
|
| 8040 |
+
},
|
| 8041 |
+
{
|
| 8042 |
+
"epoch": 0.005103367644738041,
|
| 8043 |
+
"grad_norm": 10.125,
|
| 8044 |
+
"learning_rate": 2.549483768434857e-07,
|
| 8045 |
+
"loss": 3.0064,
|
| 8046 |
+
"step": 1148
|
| 8047 |
+
},
|
| 8048 |
+
{
|
| 8049 |
+
"epoch": 0.00510781308693729,
|
| 8050 |
+
"grad_norm": 11.9375,
|
| 8051 |
+
"learning_rate": 2.551706509296614e-07,
|
| 8052 |
+
"loss": 2.7114,
|
| 8053 |
+
"step": 1149
|
| 8054 |
+
},
|
| 8055 |
+
{
|
| 8056 |
+
"epoch": 0.00511225852913654,
|
| 8057 |
+
"grad_norm": 8.875,
|
| 8058 |
+
"learning_rate": 2.5539292501583706e-07,
|
| 8059 |
+
"loss": 3.1548,
|
| 8060 |
+
"step": 1150
|
| 8061 |
+
},
|
| 8062 |
+
{
|
| 8063 |
+
"epoch": 0.005116703971335789,
|
| 8064 |
+
"grad_norm": 7.90625,
|
| 8065 |
+
"learning_rate": 2.5561519910201274e-07,
|
| 8066 |
+
"loss": 3.3142,
|
| 8067 |
+
"step": 1151
|
| 8068 |
+
},
|
| 8069 |
+
{
|
| 8070 |
+
"epoch": 0.005121149413535038,
|
| 8071 |
+
"grad_norm": 10.25,
|
| 8072 |
+
"learning_rate": 2.5583747318818837e-07,
|
| 8073 |
+
"loss": 2.9388,
|
| 8074 |
+
"step": 1152
|
| 8075 |
+
},
|
| 8076 |
+
{
|
| 8077 |
+
"epoch": 0.005125594855734287,
|
| 8078 |
+
"grad_norm": 10.3125,
|
| 8079 |
+
"learning_rate": 2.5605974727436404e-07,
|
| 8080 |
+
"loss": 2.9134,
|
| 8081 |
+
"step": 1153
|
| 8082 |
+
},
|
| 8083 |
+
{
|
| 8084 |
+
"epoch": 0.005130040297933536,
|
| 8085 |
+
"grad_norm": 8.625,
|
| 8086 |
+
"learning_rate": 2.562820213605397e-07,
|
| 8087 |
+
"loss": 3.1508,
|
| 8088 |
+
"step": 1154
|
| 8089 |
+
},
|
| 8090 |
+
{
|
| 8091 |
+
"epoch": 0.005134485740132785,
|
| 8092 |
+
"grad_norm": 9.8125,
|
| 8093 |
+
"learning_rate": 2.565042954467154e-07,
|
| 8094 |
+
"loss": 3.1071,
|
| 8095 |
+
"step": 1155
|
| 8096 |
+
},
|
| 8097 |
+
{
|
| 8098 |
+
"epoch": 0.005138931182332034,
|
| 8099 |
+
"grad_norm": 10.4375,
|
| 8100 |
+
"learning_rate": 2.56726569532891e-07,
|
| 8101 |
+
"loss": 2.9386,
|
| 8102 |
+
"step": 1156
|
| 8103 |
+
},
|
| 8104 |
+
{
|
| 8105 |
+
"epoch": 0.005143376624531283,
|
| 8106 |
+
"grad_norm": 8.5625,
|
| 8107 |
+
"learning_rate": 2.569488436190667e-07,
|
| 8108 |
+
"loss": 3.2167,
|
| 8109 |
+
"step": 1157
|
| 8110 |
+
},
|
| 8111 |
+
{
|
| 8112 |
+
"epoch": 0.005147822066730533,
|
| 8113 |
+
"grad_norm": 8.75,
|
| 8114 |
+
"learning_rate": 2.5717111770524233e-07,
|
| 8115 |
+
"loss": 3.0796,
|
| 8116 |
+
"step": 1158
|
| 8117 |
+
},
|
| 8118 |
+
{
|
| 8119 |
+
"epoch": 0.005152267508929782,
|
| 8120 |
+
"grad_norm": 9.5,
|
| 8121 |
+
"learning_rate": 2.57393391791418e-07,
|
| 8122 |
+
"loss": 3.084,
|
| 8123 |
+
"step": 1159
|
| 8124 |
+
},
|
| 8125 |
+
{
|
| 8126 |
+
"epoch": 0.005156712951129031,
|
| 8127 |
+
"grad_norm": 13.3125,
|
| 8128 |
+
"learning_rate": 2.576156658775937e-07,
|
| 8129 |
+
"loss": 2.6222,
|
| 8130 |
+
"step": 1160
|
| 8131 |
+
},
|
| 8132 |
+
{
|
| 8133 |
+
"epoch": 0.00516115839332828,
|
| 8134 |
+
"grad_norm": 8.25,
|
| 8135 |
+
"learning_rate": 2.5783793996376936e-07,
|
| 8136 |
+
"loss": 3.2654,
|
| 8137 |
+
"step": 1161
|
| 8138 |
+
},
|
| 8139 |
+
{
|
| 8140 |
+
"epoch": 0.0051656038355275295,
|
| 8141 |
+
"grad_norm": 12.0625,
|
| 8142 |
+
"learning_rate": 2.58060214049945e-07,
|
| 8143 |
+
"loss": 2.8711,
|
| 8144 |
+
"step": 1162
|
| 8145 |
+
},
|
| 8146 |
+
{
|
| 8147 |
+
"epoch": 0.0051700492777267786,
|
| 8148 |
+
"grad_norm": 12.0,
|
| 8149 |
+
"learning_rate": 2.5828248813612066e-07,
|
| 8150 |
+
"loss": 2.8799,
|
| 8151 |
+
"step": 1163
|
| 8152 |
+
},
|
| 8153 |
+
{
|
| 8154 |
+
"epoch": 0.005174494719926028,
|
| 8155 |
+
"grad_norm": 12.0,
|
| 8156 |
+
"learning_rate": 2.5850476222229634e-07,
|
| 8157 |
+
"loss": 2.8347,
|
| 8158 |
+
"step": 1164
|
| 8159 |
+
},
|
| 8160 |
+
{
|
| 8161 |
+
"epoch": 0.005178940162125277,
|
| 8162 |
+
"grad_norm": 11.125,
|
| 8163 |
+
"learning_rate": 2.58727036308472e-07,
|
| 8164 |
+
"loss": 2.9166,
|
| 8165 |
+
"step": 1165
|
| 8166 |
+
},
|
| 8167 |
+
{
|
| 8168 |
+
"epoch": 0.005183385604324526,
|
| 8169 |
+
"grad_norm": 9.125,
|
| 8170 |
+
"learning_rate": 2.5894931039464764e-07,
|
| 8171 |
+
"loss": 3.1374,
|
| 8172 |
+
"step": 1166
|
| 8173 |
+
},
|
| 8174 |
+
{
|
| 8175 |
+
"epoch": 0.005187831046523776,
|
| 8176 |
+
"grad_norm": 8.3125,
|
| 8177 |
+
"learning_rate": 2.591715844808233e-07,
|
| 8178 |
+
"loss": 3.1146,
|
| 8179 |
+
"step": 1167
|
| 8180 |
+
},
|
| 8181 |
+
{
|
| 8182 |
+
"epoch": 0.005192276488723025,
|
| 8183 |
+
"grad_norm": 11.0625,
|
| 8184 |
+
"learning_rate": 2.59393858566999e-07,
|
| 8185 |
+
"loss": 2.7386,
|
| 8186 |
+
"step": 1168
|
| 8187 |
+
},
|
| 8188 |
+
{
|
| 8189 |
+
"epoch": 0.005196721930922274,
|
| 8190 |
+
"grad_norm": 9.8125,
|
| 8191 |
+
"learning_rate": 2.596161326531747e-07,
|
| 8192 |
+
"loss": 3.0491,
|
| 8193 |
+
"step": 1169
|
| 8194 |
+
},
|
| 8195 |
+
{
|
| 8196 |
+
"epoch": 0.005201167373121523,
|
| 8197 |
+
"grad_norm": 10.4375,
|
| 8198 |
+
"learning_rate": 2.598384067393503e-07,
|
| 8199 |
+
"loss": 2.9397,
|
| 8200 |
+
"step": 1170
|
| 8201 |
+
},
|
| 8202 |
+
{
|
| 8203 |
+
"epoch": 0.005205612815320772,
|
| 8204 |
+
"grad_norm": 9.75,
|
| 8205 |
+
"learning_rate": 2.60060680825526e-07,
|
| 8206 |
+
"loss": 3.101,
|
| 8207 |
+
"step": 1171
|
| 8208 |
+
},
|
| 8209 |
+
{
|
| 8210 |
+
"epoch": 0.005210058257520021,
|
| 8211 |
+
"grad_norm": 9.0625,
|
| 8212 |
+
"learning_rate": 2.6028295491170166e-07,
|
| 8213 |
+
"loss": 3.0531,
|
| 8214 |
+
"step": 1172
|
| 8215 |
+
},
|
| 8216 |
+
{
|
| 8217 |
+
"epoch": 0.00521450369971927,
|
| 8218 |
+
"grad_norm": 9.5625,
|
| 8219 |
+
"learning_rate": 2.605052289978773e-07,
|
| 8220 |
+
"loss": 2.9893,
|
| 8221 |
+
"step": 1173
|
| 8222 |
+
},
|
| 8223 |
+
{
|
| 8224 |
+
"epoch": 0.005218949141918519,
|
| 8225 |
+
"grad_norm": 10.0625,
|
| 8226 |
+
"learning_rate": 2.6072750308405296e-07,
|
| 8227 |
+
"loss": 2.9335,
|
| 8228 |
+
"step": 1174
|
| 8229 |
+
},
|
| 8230 |
+
{
|
| 8231 |
+
"epoch": 0.005223394584117769,
|
| 8232 |
+
"grad_norm": 10.125,
|
| 8233 |
+
"learning_rate": 2.6094977717022864e-07,
|
| 8234 |
+
"loss": 3.0214,
|
| 8235 |
+
"step": 1175
|
| 8236 |
+
},
|
| 8237 |
+
{
|
| 8238 |
+
"epoch": 0.005227840026317018,
|
| 8239 |
+
"grad_norm": 9.4375,
|
| 8240 |
+
"learning_rate": 2.6117205125640426e-07,
|
| 8241 |
+
"loss": 3.0275,
|
| 8242 |
+
"step": 1176
|
| 8243 |
+
},
|
| 8244 |
+
{
|
| 8245 |
+
"epoch": 0.005232285468516267,
|
| 8246 |
+
"grad_norm": 9.3125,
|
| 8247 |
+
"learning_rate": 2.6139432534257994e-07,
|
| 8248 |
+
"loss": 3.0003,
|
| 8249 |
+
"step": 1177
|
| 8250 |
+
},
|
| 8251 |
+
{
|
| 8252 |
+
"epoch": 0.005236730910715516,
|
| 8253 |
+
"grad_norm": 9.75,
|
| 8254 |
+
"learning_rate": 2.616165994287556e-07,
|
| 8255 |
+
"loss": 2.9723,
|
| 8256 |
+
"step": 1178
|
| 8257 |
+
},
|
| 8258 |
+
{
|
| 8259 |
+
"epoch": 0.005241176352914765,
|
| 8260 |
+
"grad_norm": 9.4375,
|
| 8261 |
+
"learning_rate": 2.618388735149313e-07,
|
| 8262 |
+
"loss": 3.0696,
|
| 8263 |
+
"step": 1179
|
| 8264 |
+
},
|
| 8265 |
+
{
|
| 8266 |
+
"epoch": 0.005245621795114014,
|
| 8267 |
+
"grad_norm": 8.5625,
|
| 8268 |
+
"learning_rate": 2.620611476011069e-07,
|
| 8269 |
+
"loss": 3.104,
|
| 8270 |
+
"step": 1180
|
| 8271 |
+
},
|
| 8272 |
+
{
|
| 8273 |
+
"epoch": 0.005250067237313263,
|
| 8274 |
+
"grad_norm": 12.1875,
|
| 8275 |
+
"learning_rate": 2.622834216872826e-07,
|
| 8276 |
+
"loss": 2.8341,
|
| 8277 |
+
"step": 1181
|
| 8278 |
+
},
|
| 8279 |
+
{
|
| 8280 |
+
"epoch": 0.0052545126795125125,
|
| 8281 |
+
"grad_norm": 11.1875,
|
| 8282 |
+
"learning_rate": 2.625056957734583e-07,
|
| 8283 |
+
"loss": 2.935,
|
| 8284 |
+
"step": 1182
|
| 8285 |
+
},
|
| 8286 |
+
{
|
| 8287 |
+
"epoch": 0.005258958121711762,
|
| 8288 |
+
"grad_norm": 9.5,
|
| 8289 |
+
"learning_rate": 2.6272796985963395e-07,
|
| 8290 |
+
"loss": 2.9294,
|
| 8291 |
+
"step": 1183
|
| 8292 |
+
},
|
| 8293 |
+
{
|
| 8294 |
+
"epoch": 0.0052634035639110115,
|
| 8295 |
+
"grad_norm": 9.5625,
|
| 8296 |
+
"learning_rate": 2.629502439458096e-07,
|
| 8297 |
+
"loss": 3.052,
|
| 8298 |
+
"step": 1184
|
| 8299 |
+
},
|
| 8300 |
+
{
|
| 8301 |
+
"epoch": 0.0052678490061102605,
|
| 8302 |
+
"grad_norm": 10.375,
|
| 8303 |
+
"learning_rate": 2.6317251803198526e-07,
|
| 8304 |
+
"loss": 3.006,
|
| 8305 |
+
"step": 1185
|
| 8306 |
+
},
|
| 8307 |
+
{
|
| 8308 |
+
"epoch": 0.00527229444830951,
|
| 8309 |
+
"grad_norm": 9.8125,
|
| 8310 |
+
"learning_rate": 2.6339479211816093e-07,
|
| 8311 |
+
"loss": 3.0649,
|
| 8312 |
+
"step": 1186
|
| 8313 |
+
},
|
| 8314 |
+
{
|
| 8315 |
+
"epoch": 0.005276739890508759,
|
| 8316 |
+
"grad_norm": 7.0,
|
| 8317 |
+
"learning_rate": 2.636170662043366e-07,
|
| 8318 |
+
"loss": 3.393,
|
| 8319 |
+
"step": 1187
|
| 8320 |
+
},
|
| 8321 |
+
{
|
| 8322 |
+
"epoch": 0.005281185332708008,
|
| 8323 |
+
"grad_norm": 10.6875,
|
| 8324 |
+
"learning_rate": 2.638393402905123e-07,
|
| 8325 |
+
"loss": 3.0712,
|
| 8326 |
+
"step": 1188
|
| 8327 |
+
},
|
| 8328 |
+
{
|
| 8329 |
+
"epoch": 0.005285630774907257,
|
| 8330 |
+
"grad_norm": 6.8125,
|
| 8331 |
+
"learning_rate": 2.640616143766879e-07,
|
| 8332 |
+
"loss": 3.2026,
|
| 8333 |
+
"step": 1189
|
| 8334 |
+
},
|
| 8335 |
+
{
|
| 8336 |
+
"epoch": 0.005290076217106506,
|
| 8337 |
+
"grad_norm": 10.25,
|
| 8338 |
+
"learning_rate": 2.6428388846286354e-07,
|
| 8339 |
+
"loss": 3.0145,
|
| 8340 |
+
"step": 1190
|
| 8341 |
+
},
|
| 8342 |
+
{
|
| 8343 |
+
"epoch": 0.005294521659305755,
|
| 8344 |
+
"grad_norm": 8.8125,
|
| 8345 |
+
"learning_rate": 2.645061625490392e-07,
|
| 8346 |
+
"loss": 2.9562,
|
| 8347 |
+
"step": 1191
|
| 8348 |
+
},
|
| 8349 |
+
{
|
| 8350 |
+
"epoch": 0.005298967101505005,
|
| 8351 |
+
"grad_norm": 9.25,
|
| 8352 |
+
"learning_rate": 2.647284366352149e-07,
|
| 8353 |
+
"loss": 3.1811,
|
| 8354 |
+
"step": 1192
|
| 8355 |
+
},
|
| 8356 |
+
{
|
| 8357 |
+
"epoch": 0.005303412543704254,
|
| 8358 |
+
"grad_norm": 7.84375,
|
| 8359 |
+
"learning_rate": 2.6495071072139057e-07,
|
| 8360 |
+
"loss": 3.1817,
|
| 8361 |
+
"step": 1193
|
| 8362 |
+
},
|
| 8363 |
+
{
|
| 8364 |
+
"epoch": 0.005307857985903503,
|
| 8365 |
+
"grad_norm": 13.125,
|
| 8366 |
+
"learning_rate": 2.651729848075662e-07,
|
| 8367 |
+
"loss": 2.6848,
|
| 8368 |
+
"step": 1194
|
| 8369 |
+
},
|
| 8370 |
+
{
|
| 8371 |
+
"epoch": 0.005312303428102752,
|
| 8372 |
+
"grad_norm": 8.8125,
|
| 8373 |
+
"learning_rate": 2.653952588937419e-07,
|
| 8374 |
+
"loss": 3.0787,
|
| 8375 |
+
"step": 1195
|
| 8376 |
+
},
|
| 8377 |
+
{
|
| 8378 |
+
"epoch": 0.005316748870302001,
|
| 8379 |
+
"grad_norm": 10.5,
|
| 8380 |
+
"learning_rate": 2.6561753297991755e-07,
|
| 8381 |
+
"loss": 2.9864,
|
| 8382 |
+
"step": 1196
|
| 8383 |
+
},
|
| 8384 |
+
{
|
| 8385 |
+
"epoch": 0.00532119431250125,
|
| 8386 |
+
"grad_norm": 8.5625,
|
| 8387 |
+
"learning_rate": 2.6583980706609323e-07,
|
| 8388 |
+
"loss": 3.0954,
|
| 8389 |
+
"step": 1197
|
| 8390 |
+
},
|
| 8391 |
+
{
|
| 8392 |
+
"epoch": 0.005325639754700499,
|
| 8393 |
+
"grad_norm": 10.8125,
|
| 8394 |
+
"learning_rate": 2.660620811522689e-07,
|
| 8395 |
+
"loss": 2.9631,
|
| 8396 |
+
"step": 1198
|
| 8397 |
+
},
|
| 8398 |
+
{
|
| 8399 |
+
"epoch": 0.005330085196899748,
|
| 8400 |
+
"grad_norm": 10.5,
|
| 8401 |
+
"learning_rate": 2.6628435523844453e-07,
|
| 8402 |
+
"loss": 2.9371,
|
| 8403 |
+
"step": 1199
|
| 8404 |
+
},
|
| 8405 |
+
{
|
| 8406 |
+
"epoch": 0.005334530639098998,
|
| 8407 |
+
"grad_norm": 10.6875,
|
| 8408 |
+
"learning_rate": 2.665066293246202e-07,
|
| 8409 |
+
"loss": 3.0277,
|
| 8410 |
+
"step": 1200
|
| 8411 |
+
},
|
| 8412 |
+
{
|
| 8413 |
+
"epoch": 0.005338976081298247,
|
| 8414 |
+
"grad_norm": 12.25,
|
| 8415 |
+
"learning_rate": 2.667289034107959e-07,
|
| 8416 |
+
"loss": 2.7114,
|
| 8417 |
+
"step": 1201
|
| 8418 |
+
},
|
| 8419 |
+
{
|
| 8420 |
+
"epoch": 0.005343421523497496,
|
| 8421 |
+
"grad_norm": 8.5,
|
| 8422 |
+
"learning_rate": 2.6695117749697157e-07,
|
| 8423 |
+
"loss": 3.0918,
|
| 8424 |
+
"step": 1202
|
| 8425 |
+
},
|
| 8426 |
+
{
|
| 8427 |
+
"epoch": 0.005347866965696745,
|
| 8428 |
+
"grad_norm": 8.9375,
|
| 8429 |
+
"learning_rate": 2.671734515831472e-07,
|
| 8430 |
+
"loss": 3.2499,
|
| 8431 |
+
"step": 1203
|
| 8432 |
+
},
|
| 8433 |
+
{
|
| 8434 |
+
"epoch": 0.0053523124078959945,
|
| 8435 |
+
"grad_norm": 9.5625,
|
| 8436 |
+
"learning_rate": 2.6739572566932287e-07,
|
| 8437 |
+
"loss": 3.0669,
|
| 8438 |
+
"step": 1204
|
| 8439 |
+
},
|
| 8440 |
+
{
|
| 8441 |
+
"epoch": 0.0053567578500952435,
|
| 8442 |
+
"grad_norm": 10.1875,
|
| 8443 |
+
"learning_rate": 2.6761799975549855e-07,
|
| 8444 |
+
"loss": 3.0489,
|
| 8445 |
+
"step": 1205
|
| 8446 |
+
},
|
| 8447 |
+
{
|
| 8448 |
+
"epoch": 0.005361203292294493,
|
| 8449 |
+
"grad_norm": 10.1875,
|
| 8450 |
+
"learning_rate": 2.678402738416742e-07,
|
| 8451 |
+
"loss": 2.9393,
|
| 8452 |
+
"step": 1206
|
| 8453 |
+
},
|
| 8454 |
+
{
|
| 8455 |
+
"epoch": 0.005365648734493742,
|
| 8456 |
+
"grad_norm": 9.375,
|
| 8457 |
+
"learning_rate": 2.6806254792784985e-07,
|
| 8458 |
+
"loss": 3.0972,
|
| 8459 |
+
"step": 1207
|
| 8460 |
+
},
|
| 8461 |
+
{
|
| 8462 |
+
"epoch": 0.005370094176692991,
|
| 8463 |
+
"grad_norm": 10.5,
|
| 8464 |
+
"learning_rate": 2.6828482201402553e-07,
|
| 8465 |
+
"loss": 2.9102,
|
| 8466 |
+
"step": 1208
|
| 8467 |
+
},
|
| 8468 |
+
{
|
| 8469 |
+
"epoch": 0.005374539618892241,
|
| 8470 |
+
"grad_norm": 10.6875,
|
| 8471 |
+
"learning_rate": 2.6850709610020115e-07,
|
| 8472 |
+
"loss": 3.0099,
|
| 8473 |
+
"step": 1209
|
| 8474 |
+
},
|
| 8475 |
+
{
|
| 8476 |
+
"epoch": 0.00537898506109149,
|
| 8477 |
+
"grad_norm": 10.5,
|
| 8478 |
+
"learning_rate": 2.6872937018637683e-07,
|
| 8479 |
+
"loss": 2.9322,
|
| 8480 |
+
"step": 1210
|
| 8481 |
+
},
|
| 8482 |
+
{
|
| 8483 |
+
"epoch": 0.005383430503290739,
|
| 8484 |
+
"grad_norm": 10.375,
|
| 8485 |
+
"learning_rate": 2.689516442725525e-07,
|
| 8486 |
+
"loss": 2.9966,
|
| 8487 |
+
"step": 1211
|
| 8488 |
+
},
|
| 8489 |
+
{
|
| 8490 |
+
"epoch": 0.005387875945489988,
|
| 8491 |
+
"grad_norm": 10.6875,
|
| 8492 |
+
"learning_rate": 2.691739183587282e-07,
|
| 8493 |
+
"loss": 2.9666,
|
| 8494 |
+
"step": 1212
|
| 8495 |
+
},
|
| 8496 |
+
{
|
| 8497 |
+
"epoch": 0.005392321387689237,
|
| 8498 |
+
"grad_norm": 11.5,
|
| 8499 |
+
"learning_rate": 2.693961924449038e-07,
|
| 8500 |
+
"loss": 2.8478,
|
| 8501 |
+
"step": 1213
|
| 8502 |
+
},
|
| 8503 |
+
{
|
| 8504 |
+
"epoch": 0.005396766829888486,
|
| 8505 |
+
"grad_norm": 10.375,
|
| 8506 |
+
"learning_rate": 2.696184665310795e-07,
|
| 8507 |
+
"loss": 2.9671,
|
| 8508 |
+
"step": 1214
|
| 8509 |
+
},
|
| 8510 |
+
{
|
| 8511 |
+
"epoch": 0.005401212272087735,
|
| 8512 |
+
"grad_norm": 9.9375,
|
| 8513 |
+
"learning_rate": 2.6984074061725517e-07,
|
| 8514 |
+
"loss": 2.9147,
|
| 8515 |
+
"step": 1215
|
| 8516 |
+
},
|
| 8517 |
+
{
|
| 8518 |
+
"epoch": 0.005405657714286984,
|
| 8519 |
+
"grad_norm": 9.875,
|
| 8520 |
+
"learning_rate": 2.7006301470343084e-07,
|
| 8521 |
+
"loss": 3.0142,
|
| 8522 |
+
"step": 1216
|
| 8523 |
+
},
|
| 8524 |
+
{
|
| 8525 |
+
"epoch": 0.005410103156486234,
|
| 8526 |
+
"grad_norm": 9.875,
|
| 8527 |
+
"learning_rate": 2.7028528878960647e-07,
|
| 8528 |
+
"loss": 3.0616,
|
| 8529 |
+
"step": 1217
|
| 8530 |
+
},
|
| 8531 |
+
{
|
| 8532 |
+
"epoch": 0.005414548598685483,
|
| 8533 |
+
"grad_norm": 8.0,
|
| 8534 |
+
"learning_rate": 2.7050756287578215e-07,
|
| 8535 |
+
"loss": 3.2485,
|
| 8536 |
+
"step": 1218
|
| 8537 |
+
},
|
| 8538 |
+
{
|
| 8539 |
+
"epoch": 0.005418994040884732,
|
| 8540 |
+
"grad_norm": 5.375,
|
| 8541 |
+
"learning_rate": 2.707298369619578e-07,
|
| 8542 |
+
"loss": 3.5312,
|
| 8543 |
+
"step": 1219
|
| 8544 |
+
},
|
| 8545 |
+
{
|
| 8546 |
+
"epoch": 0.005423439483083981,
|
| 8547 |
+
"grad_norm": 9.375,
|
| 8548 |
+
"learning_rate": 2.709521110481335e-07,
|
| 8549 |
+
"loss": 3.0494,
|
| 8550 |
+
"step": 1220
|
| 8551 |
+
},
|
| 8552 |
+
{
|
| 8553 |
+
"epoch": 0.00542788492528323,
|
| 8554 |
+
"grad_norm": 9.375,
|
| 8555 |
+
"learning_rate": 2.7117438513430913e-07,
|
| 8556 |
+
"loss": 3.1738,
|
| 8557 |
+
"step": 1221
|
| 8558 |
+
},
|
| 8559 |
+
{
|
| 8560 |
+
"epoch": 0.005432330367482479,
|
| 8561 |
+
"grad_norm": 11.75,
|
| 8562 |
+
"learning_rate": 2.713966592204848e-07,
|
| 8563 |
+
"loss": 2.8339,
|
| 8564 |
+
"step": 1222
|
| 8565 |
+
},
|
| 8566 |
+
{
|
| 8567 |
+
"epoch": 0.005436775809681728,
|
| 8568 |
+
"grad_norm": 10.75,
|
| 8569 |
+
"learning_rate": 2.7161893330666043e-07,
|
| 8570 |
+
"loss": 2.9493,
|
| 8571 |
+
"step": 1223
|
| 8572 |
+
},
|
| 8573 |
+
{
|
| 8574 |
+
"epoch": 0.0054412212518809774,
|
| 8575 |
+
"grad_norm": 8.5,
|
| 8576 |
+
"learning_rate": 2.718412073928361e-07,
|
| 8577 |
+
"loss": 3.2086,
|
| 8578 |
+
"step": 1224
|
| 8579 |
+
},
|
| 8580 |
+
{
|
| 8581 |
+
"epoch": 0.0054456666940802265,
|
| 8582 |
+
"grad_norm": 10.6875,
|
| 8583 |
+
"learning_rate": 2.720634814790118e-07,
|
| 8584 |
+
"loss": 2.8396,
|
| 8585 |
+
"step": 1225
|
| 8586 |
+
},
|
| 8587 |
+
{
|
| 8588 |
+
"epoch": 0.005450112136279476,
|
| 8589 |
+
"grad_norm": 11.375,
|
| 8590 |
+
"learning_rate": 2.7228575556518746e-07,
|
| 8591 |
+
"loss": 2.7995,
|
| 8592 |
+
"step": 1226
|
| 8593 |
+
},
|
| 8594 |
+
{
|
| 8595 |
+
"epoch": 0.0054545575784787255,
|
| 8596 |
+
"grad_norm": 10.125,
|
| 8597 |
+
"learning_rate": 2.725080296513631e-07,
|
| 8598 |
+
"loss": 2.9795,
|
| 8599 |
+
"step": 1227
|
| 8600 |
+
},
|
| 8601 |
+
{
|
| 8602 |
+
"epoch": 0.0054590030206779746,
|
| 8603 |
+
"grad_norm": 9.125,
|
| 8604 |
+
"learning_rate": 2.7273030373753877e-07,
|
| 8605 |
+
"loss": 3.1222,
|
| 8606 |
+
"step": 1228
|
| 8607 |
+
},
|
| 8608 |
+
{
|
| 8609 |
+
"epoch": 0.005463448462877224,
|
| 8610 |
+
"grad_norm": 9.8125,
|
| 8611 |
+
"learning_rate": 2.7295257782371444e-07,
|
| 8612 |
+
"loss": 3.0153,
|
| 8613 |
+
"step": 1229
|
| 8614 |
+
},
|
| 8615 |
+
{
|
| 8616 |
+
"epoch": 0.005467893905076473,
|
| 8617 |
+
"grad_norm": 9.75,
|
| 8618 |
+
"learning_rate": 2.731748519098901e-07,
|
| 8619 |
+
"loss": 3.1154,
|
| 8620 |
+
"step": 1230
|
| 8621 |
+
},
|
| 8622 |
+
{
|
| 8623 |
+
"epoch": 0.005472339347275722,
|
| 8624 |
+
"grad_norm": 10.0625,
|
| 8625 |
+
"learning_rate": 2.7339712599606575e-07,
|
| 8626 |
+
"loss": 2.9918,
|
| 8627 |
+
"step": 1231
|
| 8628 |
+
},
|
| 8629 |
+
{
|
| 8630 |
+
"epoch": 0.005476784789474971,
|
| 8631 |
+
"grad_norm": 8.875,
|
| 8632 |
+
"learning_rate": 2.736194000822414e-07,
|
| 8633 |
+
"loss": 3.2199,
|
| 8634 |
+
"step": 1232
|
| 8635 |
+
},
|
| 8636 |
+
{
|
| 8637 |
+
"epoch": 0.00548123023167422,
|
| 8638 |
+
"grad_norm": 10.375,
|
| 8639 |
+
"learning_rate": 2.738416741684171e-07,
|
| 8640 |
+
"loss": 2.9295,
|
| 8641 |
+
"step": 1233
|
| 8642 |
+
},
|
| 8643 |
+
{
|
| 8644 |
+
"epoch": 0.00548567567387347,
|
| 8645 |
+
"grad_norm": 11.25,
|
| 8646 |
+
"learning_rate": 2.740639482545928e-07,
|
| 8647 |
+
"loss": 2.8678,
|
| 8648 |
+
"step": 1234
|
| 8649 |
+
},
|
| 8650 |
+
{
|
| 8651 |
+
"epoch": 0.005490121116072719,
|
| 8652 |
+
"grad_norm": 10.125,
|
| 8653 |
+
"learning_rate": 2.7428622234076846e-07,
|
| 8654 |
+
"loss": 3.1695,
|
| 8655 |
+
"step": 1235
|
| 8656 |
+
},
|
| 8657 |
+
{
|
| 8658 |
+
"epoch": 0.005494566558271968,
|
| 8659 |
+
"grad_norm": 10.0625,
|
| 8660 |
+
"learning_rate": 2.745084964269441e-07,
|
| 8661 |
+
"loss": 3.0594,
|
| 8662 |
+
"step": 1236
|
| 8663 |
+
},
|
| 8664 |
+
{
|
| 8665 |
+
"epoch": 0.005499012000471217,
|
| 8666 |
+
"grad_norm": 9.0,
|
| 8667 |
+
"learning_rate": 2.7473077051311976e-07,
|
| 8668 |
+
"loss": 2.9845,
|
| 8669 |
+
"step": 1237
|
| 8670 |
+
},
|
| 8671 |
+
{
|
| 8672 |
+
"epoch": 0.005503457442670466,
|
| 8673 |
+
"grad_norm": 8.5625,
|
| 8674 |
+
"learning_rate": 2.7495304459929544e-07,
|
| 8675 |
+
"loss": 3.0625,
|
| 8676 |
+
"step": 1238
|
| 8677 |
+
},
|
| 8678 |
+
{
|
| 8679 |
+
"epoch": 0.005507902884869715,
|
| 8680 |
+
"grad_norm": 9.5625,
|
| 8681 |
+
"learning_rate": 2.751753186854711e-07,
|
| 8682 |
+
"loss": 3.0609,
|
| 8683 |
+
"step": 1239
|
| 8684 |
+
},
|
| 8685 |
+
{
|
| 8686 |
+
"epoch": 0.005512348327068964,
|
| 8687 |
+
"grad_norm": 9.25,
|
| 8688 |
+
"learning_rate": 2.7539759277164674e-07,
|
| 8689 |
+
"loss": 3.0943,
|
| 8690 |
+
"step": 1240
|
| 8691 |
+
},
|
| 8692 |
+
{
|
| 8693 |
+
"epoch": 0.005516793769268213,
|
| 8694 |
+
"grad_norm": 12.375,
|
| 8695 |
+
"learning_rate": 2.7561986685782237e-07,
|
| 8696 |
+
"loss": 2.8101,
|
| 8697 |
+
"step": 1241
|
| 8698 |
+
},
|
| 8699 |
+
{
|
| 8700 |
+
"epoch": 0.005521239211467462,
|
| 8701 |
+
"grad_norm": 10.0625,
|
| 8702 |
+
"learning_rate": 2.7584214094399804e-07,
|
| 8703 |
+
"loss": 3.0231,
|
| 8704 |
+
"step": 1242
|
| 8705 |
+
},
|
| 8706 |
+
{
|
| 8707 |
+
"epoch": 0.005525684653666712,
|
| 8708 |
+
"grad_norm": 8.8125,
|
| 8709 |
+
"learning_rate": 2.760644150301737e-07,
|
| 8710 |
+
"loss": 3.1517,
|
| 8711 |
+
"step": 1243
|
| 8712 |
+
},
|
| 8713 |
+
{
|
| 8714 |
+
"epoch": 0.005530130095865961,
|
| 8715 |
+
"grad_norm": 8.0625,
|
| 8716 |
+
"learning_rate": 2.762866891163494e-07,
|
| 8717 |
+
"loss": 3.2189,
|
| 8718 |
+
"step": 1244
|
| 8719 |
+
},
|
| 8720 |
+
{
|
| 8721 |
+
"epoch": 0.00553457553806521,
|
| 8722 |
+
"grad_norm": 9.4375,
|
| 8723 |
+
"learning_rate": 2.76508963202525e-07,
|
| 8724 |
+
"loss": 3.0806,
|
| 8725 |
+
"step": 1245
|
| 8726 |
+
},
|
| 8727 |
+
{
|
| 8728 |
+
"epoch": 0.005539020980264459,
|
| 8729 |
+
"grad_norm": 9.125,
|
| 8730 |
+
"learning_rate": 2.767312372887007e-07,
|
| 8731 |
+
"loss": 3.1077,
|
| 8732 |
+
"step": 1246
|
| 8733 |
+
},
|
| 8734 |
+
{
|
| 8735 |
+
"epoch": 0.0055434664224637085,
|
| 8736 |
+
"grad_norm": 10.75,
|
| 8737 |
+
"learning_rate": 2.769535113748764e-07,
|
| 8738 |
+
"loss": 2.8885,
|
| 8739 |
+
"step": 1247
|
| 8740 |
+
},
|
| 8741 |
+
{
|
| 8742 |
+
"epoch": 0.0055479118646629575,
|
| 8743 |
+
"grad_norm": 10.125,
|
| 8744 |
+
"learning_rate": 2.7717578546105206e-07,
|
| 8745 |
+
"loss": 2.9466,
|
| 8746 |
+
"step": 1248
|
| 8747 |
+
},
|
| 8748 |
+
{
|
| 8749 |
+
"epoch": 0.005552357306862207,
|
| 8750 |
+
"grad_norm": 9.6875,
|
| 8751 |
+
"learning_rate": 2.7739805954722774e-07,
|
| 8752 |
+
"loss": 3.0036,
|
| 8753 |
+
"step": 1249
|
| 8754 |
+
},
|
| 8755 |
+
{
|
| 8756 |
+
"epoch": 0.005556802749061456,
|
| 8757 |
+
"grad_norm": 9.6875,
|
| 8758 |
+
"learning_rate": 2.7762033363340336e-07,
|
| 8759 |
+
"loss": 3.0176,
|
| 8760 |
+
"step": 1250
|
| 8761 |
+
},
|
| 8762 |
+
{
|
| 8763 |
+
"epoch": 0.005561248191260706,
|
| 8764 |
+
"grad_norm": 10.875,
|
| 8765 |
+
"learning_rate": 2.7784260771957904e-07,
|
| 8766 |
+
"loss": 2.9511,
|
| 8767 |
+
"step": 1251
|
| 8768 |
+
},
|
| 8769 |
+
{
|
| 8770 |
+
"epoch": 0.005565693633459955,
|
| 8771 |
+
"grad_norm": 9.4375,
|
| 8772 |
+
"learning_rate": 2.780648818057547e-07,
|
| 8773 |
+
"loss": 3.2522,
|
| 8774 |
+
"step": 1252
|
| 8775 |
+
},
|
| 8776 |
+
{
|
| 8777 |
+
"epoch": 0.005570139075659204,
|
| 8778 |
+
"grad_norm": 7.1875,
|
| 8779 |
+
"learning_rate": 2.782871558919304e-07,
|
| 8780 |
+
"loss": 3.3091,
|
| 8781 |
+
"step": 1253
|
| 8782 |
+
},
|
| 8783 |
+
{
|
| 8784 |
+
"epoch": 0.005574584517858453,
|
| 8785 |
+
"grad_norm": 7.90625,
|
| 8786 |
+
"learning_rate": 2.78509429978106e-07,
|
| 8787 |
+
"loss": 3.1846,
|
| 8788 |
+
"step": 1254
|
| 8789 |
+
},
|
| 8790 |
+
{
|
| 8791 |
+
"epoch": 0.005579029960057702,
|
| 8792 |
+
"grad_norm": 8.625,
|
| 8793 |
+
"learning_rate": 2.787317040642817e-07,
|
| 8794 |
+
"loss": 3.1589,
|
| 8795 |
+
"step": 1255
|
| 8796 |
+
},
|
| 8797 |
+
{
|
| 8798 |
+
"epoch": 0.005583475402256951,
|
| 8799 |
+
"grad_norm": 8.0,
|
| 8800 |
+
"learning_rate": 2.789539781504574e-07,
|
| 8801 |
+
"loss": 3.2919,
|
| 8802 |
+
"step": 1256
|
| 8803 |
+
},
|
| 8804 |
+
{
|
| 8805 |
+
"epoch": 0.0055879208444562,
|
| 8806 |
+
"grad_norm": 10.125,
|
| 8807 |
+
"learning_rate": 2.79176252236633e-07,
|
| 8808 |
+
"loss": 2.9361,
|
| 8809 |
+
"step": 1257
|
| 8810 |
+
},
|
| 8811 |
+
{
|
| 8812 |
+
"epoch": 0.005592366286655449,
|
| 8813 |
+
"grad_norm": 8.375,
|
| 8814 |
+
"learning_rate": 2.793985263228087e-07,
|
| 8815 |
+
"loss": 3.2461,
|
| 8816 |
+
"step": 1258
|
| 8817 |
+
},
|
| 8818 |
+
{
|
| 8819 |
+
"epoch": 0.005596811728854698,
|
| 8820 |
+
"grad_norm": 8.125,
|
| 8821 |
+
"learning_rate": 2.7962080040898435e-07,
|
| 8822 |
+
"loss": 3.2751,
|
| 8823 |
+
"step": 1259
|
| 8824 |
+
},
|
| 8825 |
+
{
|
| 8826 |
+
"epoch": 0.005601257171053948,
|
| 8827 |
+
"grad_norm": 9.9375,
|
| 8828 |
+
"learning_rate": 2.7984307449516e-07,
|
| 8829 |
+
"loss": 2.9777,
|
| 8830 |
+
"step": 1260
|
| 8831 |
+
},
|
| 8832 |
+
{
|
| 8833 |
+
"epoch": 0.005605702613253197,
|
| 8834 |
+
"grad_norm": 9.625,
|
| 8835 |
+
"learning_rate": 2.8006534858133566e-07,
|
| 8836 |
+
"loss": 3.2067,
|
| 8837 |
+
"step": 1261
|
| 8838 |
+
},
|
| 8839 |
+
{
|
| 8840 |
+
"epoch": 0.005610148055452446,
|
| 8841 |
+
"grad_norm": 8.625,
|
| 8842 |
+
"learning_rate": 2.8028762266751134e-07,
|
| 8843 |
+
"loss": 3.229,
|
| 8844 |
+
"step": 1262
|
| 8845 |
+
},
|
| 8846 |
+
{
|
| 8847 |
+
"epoch": 0.005614593497651695,
|
| 8848 |
+
"grad_norm": 9.0,
|
| 8849 |
+
"learning_rate": 2.80509896753687e-07,
|
| 8850 |
+
"loss": 3.2134,
|
| 8851 |
+
"step": 1263
|
| 8852 |
+
},
|
| 8853 |
+
{
|
| 8854 |
+
"epoch": 0.005619038939850944,
|
| 8855 |
+
"grad_norm": 9.8125,
|
| 8856 |
+
"learning_rate": 2.8073217083986264e-07,
|
| 8857 |
+
"loss": 2.9657,
|
| 8858 |
+
"step": 1264
|
| 8859 |
+
},
|
| 8860 |
+
{
|
| 8861 |
+
"epoch": 0.005623484382050193,
|
| 8862 |
+
"grad_norm": 9.1875,
|
| 8863 |
+
"learning_rate": 2.809544449260383e-07,
|
| 8864 |
+
"loss": 3.1097,
|
| 8865 |
+
"step": 1265
|
| 8866 |
+
},
|
| 8867 |
+
{
|
| 8868 |
+
"epoch": 0.005627929824249442,
|
| 8869 |
+
"grad_norm": 9.25,
|
| 8870 |
+
"learning_rate": 2.81176719012214e-07,
|
| 8871 |
+
"loss": 3.0363,
|
| 8872 |
+
"step": 1266
|
| 8873 |
+
},
|
| 8874 |
+
{
|
| 8875 |
+
"epoch": 0.0056323752664486915,
|
| 8876 |
+
"grad_norm": 9.9375,
|
| 8877 |
+
"learning_rate": 2.8139899309838967e-07,
|
| 8878 |
+
"loss": 2.9348,
|
| 8879 |
+
"step": 1267
|
| 8880 |
+
},
|
| 8881 |
+
{
|
| 8882 |
+
"epoch": 0.005636820708647941,
|
| 8883 |
+
"grad_norm": 12.4375,
|
| 8884 |
+
"learning_rate": 2.816212671845653e-07,
|
| 8885 |
+
"loss": 2.815,
|
| 8886 |
+
"step": 1268
|
| 8887 |
+
},
|
| 8888 |
+
{
|
| 8889 |
+
"epoch": 0.0056412661508471905,
|
| 8890 |
+
"grad_norm": 8.4375,
|
| 8891 |
+
"learning_rate": 2.81843541270741e-07,
|
| 8892 |
+
"loss": 3.2688,
|
| 8893 |
+
"step": 1269
|
| 8894 |
+
},
|
| 8895 |
+
{
|
| 8896 |
+
"epoch": 0.0056457115930464395,
|
| 8897 |
+
"grad_norm": 10.8125,
|
| 8898 |
+
"learning_rate": 2.8206581535691665e-07,
|
| 8899 |
+
"loss": 3.0391,
|
| 8900 |
+
"step": 1270
|
| 8901 |
}
|
| 8902 |
],
|
| 8903 |
"logging_steps": 1,
|
|
|
|
| 8917 |
"attributes": {}
|
| 8918 |
}
|
| 8919 |
},
|
| 8920 |
+
"total_flos": 5.02759358988288e+16,
|
| 8921 |
"train_batch_size": 1,
|
| 8922 |
"trial_name": null,
|
| 8923 |
"trial_params": null
|