Training in progress, step 614, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2503003904
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70bd2f1fafcc7b47703939366d5a875aff52142b49bf6285307a5af2ad10a8c7
|
| 3 |
size 2503003904
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5006244836
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8c344e2c84cb015817f049934ef360b2461e56bde3cfedcd901f935b22064df
|
| 3 |
size 5006244836
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09c47ca366ccbbd06cb763290143357c3f8edb356ab78715c93165d5577f5bd0
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5b27a5164f4a92724958d35729275d1de6caf155726547d148f0c77c37fa93ec
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3353,6 +3353,958 @@
|
|
| 3353 |
"learning_rate": 1.6301935118745826e-05,
|
| 3354 |
"loss": 0.1043,
|
| 3355 |
"step": 478
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3356 |
}
|
| 3357 |
],
|
| 3358 |
"logging_steps": 1,
|
|
@@ -3367,12 +4319,12 @@
|
|
| 3367 |
"should_evaluate": false,
|
| 3368 |
"should_log": false,
|
| 3369 |
"should_save": true,
|
| 3370 |
-
"should_training_stop":
|
| 3371 |
},
|
| 3372 |
"attributes": {}
|
| 3373 |
}
|
| 3374 |
},
|
| 3375 |
-
"total_flos":
|
| 3376 |
"train_batch_size": 4,
|
| 3377 |
"trial_name": null,
|
| 3378 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9995929995929996,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 614,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3353 |
"learning_rate": 1.6301935118745826e-05,
|
| 3354 |
"loss": 0.1043,
|
| 3355 |
"step": 478
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.7798127798127799,
|
| 3359 |
+
"grad_norm": 0.7857057452201843,
|
| 3360 |
+
"learning_rate": 1.6076797092734573e-05,
|
| 3361 |
+
"loss": 0.118,
|
| 3362 |
+
"step": 479
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.7814407814407814,
|
| 3366 |
+
"grad_norm": 0.7554141879081726,
|
| 3367 |
+
"learning_rate": 1.5852926336289925e-05,
|
| 3368 |
+
"loss": 0.0752,
|
| 3369 |
+
"step": 480
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.783068783068783,
|
| 3373 |
+
"grad_norm": 1.1241683959960938,
|
| 3374 |
+
"learning_rate": 1.5630331212554905e-05,
|
| 3375 |
+
"loss": 0.1308,
|
| 3376 |
+
"step": 481
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.7846967846967847,
|
| 3380 |
+
"grad_norm": 1.1640019416809082,
|
| 3381 |
+
"learning_rate": 1.5409020037018652e-05,
|
| 3382 |
+
"loss": 0.115,
|
| 3383 |
+
"step": 482
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.7863247863247863,
|
| 3387 |
+
"grad_norm": 1.619645357131958,
|
| 3388 |
+
"learning_rate": 1.5189001077205833e-05,
|
| 3389 |
+
"loss": 0.2098,
|
| 3390 |
+
"step": 483
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.787952787952788,
|
| 3394 |
+
"grad_norm": 1.3909748792648315,
|
| 3395 |
+
"learning_rate": 1.4970282552367854e-05,
|
| 3396 |
+
"loss": 0.1389,
|
| 3397 |
+
"step": 484
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.7895807895807896,
|
| 3401 |
+
"grad_norm": 0.983561635017395,
|
| 3402 |
+
"learning_rate": 1.475287263317569e-05,
|
| 3403 |
+
"loss": 0.0782,
|
| 3404 |
+
"step": 485
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.7912087912087912,
|
| 3408 |
+
"grad_norm": 1.0795811414718628,
|
| 3409 |
+
"learning_rate": 1.453677944141474e-05,
|
| 3410 |
+
"loss": 0.1023,
|
| 3411 |
+
"step": 486
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.7928367928367929,
|
| 3415 |
+
"grad_norm": 1.4191927909851074,
|
| 3416 |
+
"learning_rate": 1.4322011049681411e-05,
|
| 3417 |
+
"loss": 0.1648,
|
| 3418 |
+
"step": 487
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.7944647944647945,
|
| 3422 |
+
"grad_norm": 1.2225549221038818,
|
| 3423 |
+
"learning_rate": 1.4108575481081521e-05,
|
| 3424 |
+
"loss": 0.1021,
|
| 3425 |
+
"step": 488
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.796092796092796,
|
| 3429 |
+
"grad_norm": 1.0409220457077026,
|
| 3430 |
+
"learning_rate": 1.3896480708930576e-05,
|
| 3431 |
+
"loss": 0.0923,
|
| 3432 |
+
"step": 489
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.7977207977207977,
|
| 3436 |
+
"grad_norm": 1.3591772317886353,
|
| 3437 |
+
"learning_rate": 1.3685734656455989e-05,
|
| 3438 |
+
"loss": 0.1194,
|
| 3439 |
+
"step": 490
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.7993487993487993,
|
| 3443 |
+
"grad_norm": 1.3432077169418335,
|
| 3444 |
+
"learning_rate": 1.3476345196500978e-05,
|
| 3445 |
+
"loss": 0.0894,
|
| 3446 |
+
"step": 491
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.800976800976801,
|
| 3450 |
+
"grad_norm": 1.5234460830688477,
|
| 3451 |
+
"learning_rate": 1.3268320151230517e-05,
|
| 3452 |
+
"loss": 0.1638,
|
| 3453 |
+
"step": 492
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.8026048026048026,
|
| 3457 |
+
"grad_norm": 0.6838420033454895,
|
| 3458 |
+
"learning_rate": 1.306166729183918e-05,
|
| 3459 |
+
"loss": 0.0699,
|
| 3460 |
+
"step": 493
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.8042328042328042,
|
| 3464 |
+
"grad_norm": 1.3897405862808228,
|
| 3465 |
+
"learning_rate": 1.2856394338260691e-05,
|
| 3466 |
+
"loss": 0.0923,
|
| 3467 |
+
"step": 494
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.8058608058608059,
|
| 3471 |
+
"grad_norm": 1.5889415740966797,
|
| 3472 |
+
"learning_rate": 1.2652508958879672e-05,
|
| 3473 |
+
"loss": 0.1909,
|
| 3474 |
+
"step": 495
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.8074888074888075,
|
| 3478 |
+
"grad_norm": 1.4787379503250122,
|
| 3479 |
+
"learning_rate": 1.245001877024512e-05,
|
| 3480 |
+
"loss": 0.1033,
|
| 3481 |
+
"step": 496
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.8091168091168092,
|
| 3485 |
+
"grad_norm": 0.7427048087120056,
|
| 3486 |
+
"learning_rate": 1.224893133678583e-05,
|
| 3487 |
+
"loss": 0.0342,
|
| 3488 |
+
"step": 497
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.8107448107448108,
|
| 3492 |
+
"grad_norm": 0.8333874344825745,
|
| 3493 |
+
"learning_rate": 1.2049254170527857e-05,
|
| 3494 |
+
"loss": 0.0373,
|
| 3495 |
+
"step": 498
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.8123728123728123,
|
| 3499 |
+
"grad_norm": 1.2029796838760376,
|
| 3500 |
+
"learning_rate": 1.1850994730813936e-05,
|
| 3501 |
+
"loss": 0.0835,
|
| 3502 |
+
"step": 499
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.814000814000814,
|
| 3506 |
+
"grad_norm": 2.5518958568573,
|
| 3507 |
+
"learning_rate": 1.1654160424024719e-05,
|
| 3508 |
+
"loss": 0.2705,
|
| 3509 |
+
"step": 500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.8156288156288156,
|
| 3513 |
+
"grad_norm": 0.028806041926145554,
|
| 3514 |
+
"learning_rate": 1.1458758603302144e-05,
|
| 3515 |
+
"loss": 0.0011,
|
| 3516 |
+
"step": 501
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.8172568172568172,
|
| 3520 |
+
"grad_norm": 0.03748834505677223,
|
| 3521 |
+
"learning_rate": 1.1264796568274811e-05,
|
| 3522 |
+
"loss": 0.0014,
|
| 3523 |
+
"step": 502
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.8188848188848189,
|
| 3527 |
+
"grad_norm": 0.2620696425437927,
|
| 3528 |
+
"learning_rate": 1.1072281564785147e-05,
|
| 3529 |
+
"loss": 0.0363,
|
| 3530 |
+
"step": 503
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.8205128205128205,
|
| 3534 |
+
"grad_norm": 0.5484391450881958,
|
| 3535 |
+
"learning_rate": 1.088122078461884e-05,
|
| 3536 |
+
"loss": 0.1093,
|
| 3537 |
+
"step": 504
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.8221408221408222,
|
| 3541 |
+
"grad_norm": 0.5863115191459656,
|
| 3542 |
+
"learning_rate": 1.0691621365236153e-05,
|
| 3543 |
+
"loss": 0.0954,
|
| 3544 |
+
"step": 505
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.8237688237688238,
|
| 3548 |
+
"grad_norm": 0.39074957370758057,
|
| 3549 |
+
"learning_rate": 1.0503490389505244e-05,
|
| 3550 |
+
"loss": 0.0387,
|
| 3551 |
+
"step": 506
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.8253968253968254,
|
| 3555 |
+
"grad_norm": 0.5790150165557861,
|
| 3556 |
+
"learning_rate": 1.0316834885437593e-05,
|
| 3557 |
+
"loss": 0.0768,
|
| 3558 |
+
"step": 507
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.827024827024827,
|
| 3562 |
+
"grad_norm": 0.44617122411727905,
|
| 3563 |
+
"learning_rate": 1.013166182592551e-05,
|
| 3564 |
+
"loss": 0.0545,
|
| 3565 |
+
"step": 508
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.8286528286528286,
|
| 3569 |
+
"grad_norm": 0.18992449343204498,
|
| 3570 |
+
"learning_rate": 9.947978128481556e-06,
|
| 3571 |
+
"loss": 0.0099,
|
| 3572 |
+
"step": 509
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.8302808302808303,
|
| 3576 |
+
"grad_norm": 0.3735041916370392,
|
| 3577 |
+
"learning_rate": 9.765790654980195e-06,
|
| 3578 |
+
"loss": 0.026,
|
| 3579 |
+
"step": 510
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.8319088319088319,
|
| 3583 |
+
"grad_norm": 0.7374064326286316,
|
| 3584 |
+
"learning_rate": 9.585106211401452e-06,
|
| 3585 |
+
"loss": 0.0849,
|
| 3586 |
+
"step": 511
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.8335368335368335,
|
| 3590 |
+
"grad_norm": 0.46673384308815,
|
| 3591 |
+
"learning_rate": 9.40593154757659e-06,
|
| 3592 |
+
"loss": 0.037,
|
| 3593 |
+
"step": 512
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.8351648351648352,
|
| 3597 |
+
"grad_norm": 0.2250240594148636,
|
| 3598 |
+
"learning_rate": 9.228273356936046e-06,
|
| 3599 |
+
"loss": 0.0151,
|
| 3600 |
+
"step": 513
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.8367928367928368,
|
| 3604 |
+
"grad_norm": 0.30334368348121643,
|
| 3605 |
+
"learning_rate": 9.052138276259348e-06,
|
| 3606 |
+
"loss": 0.0115,
|
| 3607 |
+
"step": 514
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.8384208384208384,
|
| 3611 |
+
"grad_norm": 0.448330819606781,
|
| 3612 |
+
"learning_rate": 8.87753288542717e-06,
|
| 3613 |
+
"loss": 0.0438,
|
| 3614 |
+
"step": 515
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.8400488400488401,
|
| 3618 |
+
"grad_norm": 1.012182593345642,
|
| 3619 |
+
"learning_rate": 8.704463707175526e-06,
|
| 3620 |
+
"loss": 0.1588,
|
| 3621 |
+
"step": 516
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.8416768416768416,
|
| 3625 |
+
"grad_norm": 0.4254077970981598,
|
| 3626 |
+
"learning_rate": 8.532937206852165e-06,
|
| 3627 |
+
"loss": 0.0321,
|
| 3628 |
+
"step": 517
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.8433048433048433,
|
| 3632 |
+
"grad_norm": 0.5330860614776611,
|
| 3633 |
+
"learning_rate": 8.36295979217494e-06,
|
| 3634 |
+
"loss": 0.0212,
|
| 3635 |
+
"step": 518
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.8449328449328449,
|
| 3639 |
+
"grad_norm": 0.38128530979156494,
|
| 3640 |
+
"learning_rate": 8.194537812992532e-06,
|
| 3641 |
+
"loss": 0.0188,
|
| 3642 |
+
"step": 519
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.8465608465608465,
|
| 3646 |
+
"grad_norm": 0.26952776312828064,
|
| 3647 |
+
"learning_rate": 8.027677561047175e-06,
|
| 3648 |
+
"loss": 0.0135,
|
| 3649 |
+
"step": 520
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.8481888481888482,
|
| 3653 |
+
"grad_norm": 0.6845637559890747,
|
| 3654 |
+
"learning_rate": 7.862385269739624e-06,
|
| 3655 |
+
"loss": 0.0881,
|
| 3656 |
+
"step": 521
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.8498168498168498,
|
| 3660 |
+
"grad_norm": 0.5055005550384521,
|
| 3661 |
+
"learning_rate": 7.698667113896346e-06,
|
| 3662 |
+
"loss": 0.0429,
|
| 3663 |
+
"step": 522
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.8514448514448515,
|
| 3667 |
+
"grad_norm": 0.6151329874992371,
|
| 3668 |
+
"learning_rate": 7.536529209538773e-06,
|
| 3669 |
+
"loss": 0.0654,
|
| 3670 |
+
"step": 523
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.8530728530728531,
|
| 3674 |
+
"grad_norm": 0.880127489566803,
|
| 3675 |
+
"learning_rate": 7.375977613654861e-06,
|
| 3676 |
+
"loss": 0.0937,
|
| 3677 |
+
"step": 524
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.8547008547008547,
|
| 3681 |
+
"grad_norm": 0.48843473196029663,
|
| 3682 |
+
"learning_rate": 7.217018323972851e-06,
|
| 3683 |
+
"loss": 0.063,
|
| 3684 |
+
"step": 525
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.8563288563288564,
|
| 3688 |
+
"grad_norm": 0.6441347599029541,
|
| 3689 |
+
"learning_rate": 7.059657278737136e-06,
|
| 3690 |
+
"loss": 0.0974,
|
| 3691 |
+
"step": 526
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.8579568579568579,
|
| 3695 |
+
"grad_norm": 0.4741074740886688,
|
| 3696 |
+
"learning_rate": 6.903900356486504e-06,
|
| 3697 |
+
"loss": 0.0428,
|
| 3698 |
+
"step": 527
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.8595848595848596,
|
| 3702 |
+
"grad_norm": 0.7434227466583252,
|
| 3703 |
+
"learning_rate": 6.7497533758344665e-06,
|
| 3704 |
+
"loss": 0.0964,
|
| 3705 |
+
"step": 528
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.8612128612128612,
|
| 3709 |
+
"grad_norm": 0.7430245876312256,
|
| 3710 |
+
"learning_rate": 6.597222095251965e-06,
|
| 3711 |
+
"loss": 0.1001,
|
| 3712 |
+
"step": 529
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.8628408628408628,
|
| 3716 |
+
"grad_norm": 0.8883737325668335,
|
| 3717 |
+
"learning_rate": 6.446312212852163e-06,
|
| 3718 |
+
"loss": 0.1359,
|
| 3719 |
+
"step": 530
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.8644688644688645,
|
| 3723 |
+
"grad_norm": 0.9965616464614868,
|
| 3724 |
+
"learning_rate": 6.29702936617767e-06,
|
| 3725 |
+
"loss": 0.1649,
|
| 3726 |
+
"step": 531
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.8660968660968661,
|
| 3730 |
+
"grad_norm": 1.306498646736145,
|
| 3731 |
+
"learning_rate": 6.14937913198988e-06,
|
| 3732 |
+
"loss": 0.075,
|
| 3733 |
+
"step": 532
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.8677248677248677,
|
| 3737 |
+
"grad_norm": 1.1316754817962646,
|
| 3738 |
+
"learning_rate": 6.003367026060646e-06,
|
| 3739 |
+
"loss": 0.1295,
|
| 3740 |
+
"step": 533
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.8693528693528694,
|
| 3744 |
+
"grad_norm": 1.1253420114517212,
|
| 3745 |
+
"learning_rate": 5.858998502966273e-06,
|
| 3746 |
+
"loss": 0.1071,
|
| 3747 |
+
"step": 534
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.870980870980871,
|
| 3751 |
+
"grad_norm": 1.2386155128479004,
|
| 3752 |
+
"learning_rate": 5.7162789558837025e-06,
|
| 3753 |
+
"loss": 0.1373,
|
| 3754 |
+
"step": 535
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.8726088726088727,
|
| 3758 |
+
"grad_norm": 1.7070964574813843,
|
| 3759 |
+
"learning_rate": 5.575213716389039e-06,
|
| 3760 |
+
"loss": 0.2068,
|
| 3761 |
+
"step": 536
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.8742368742368742,
|
| 3765 |
+
"grad_norm": 1.152567982673645,
|
| 3766 |
+
"learning_rate": 5.435808054258429e-06,
|
| 3767 |
+
"loss": 0.1045,
|
| 3768 |
+
"step": 537
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.8758648758648758,
|
| 3772 |
+
"grad_norm": 0.7139473557472229,
|
| 3773 |
+
"learning_rate": 5.298067177271143e-06,
|
| 3774 |
+
"loss": 0.0588,
|
| 3775 |
+
"step": 538
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.8774928774928775,
|
| 3779 |
+
"grad_norm": 1.0029757022857666,
|
| 3780 |
+
"learning_rate": 5.161996231015049e-06,
|
| 3781 |
+
"loss": 0.0709,
|
| 3782 |
+
"step": 539
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.8791208791208791,
|
| 3786 |
+
"grad_norm": 0.9710745215415955,
|
| 3787 |
+
"learning_rate": 5.027600298694396e-06,
|
| 3788 |
+
"loss": 0.0748,
|
| 3789 |
+
"step": 540
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.8807488807488808,
|
| 3793 |
+
"grad_norm": 1.085533618927002,
|
| 3794 |
+
"learning_rate": 4.89488440093992e-06,
|
| 3795 |
+
"loss": 0.0728,
|
| 3796 |
+
"step": 541
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.8823768823768824,
|
| 3800 |
+
"grad_norm": 1.0848852396011353,
|
| 3801 |
+
"learning_rate": 4.763853495621251e-06,
|
| 3802 |
+
"loss": 0.0837,
|
| 3803 |
+
"step": 542
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.884004884004884,
|
| 3807 |
+
"grad_norm": 1.214345932006836,
|
| 3808 |
+
"learning_rate": 4.634512477661784e-06,
|
| 3809 |
+
"loss": 0.0998,
|
| 3810 |
+
"step": 543
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.8856328856328857,
|
| 3814 |
+
"grad_norm": 1.3322100639343262,
|
| 3815 |
+
"learning_rate": 4.5068661788557344e-06,
|
| 3816 |
+
"loss": 0.1066,
|
| 3817 |
+
"step": 544
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.8872608872608873,
|
| 3821 |
+
"grad_norm": 1.40041983127594,
|
| 3822 |
+
"learning_rate": 4.380919367687658e-06,
|
| 3823 |
+
"loss": 0.1525,
|
| 3824 |
+
"step": 545
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.8888888888888888,
|
| 3828 |
+
"grad_norm": 1.1170860528945923,
|
| 3829 |
+
"learning_rate": 4.2566767491543706e-06,
|
| 3830 |
+
"loss": 0.0883,
|
| 3831 |
+
"step": 546
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.8905168905168905,
|
| 3835 |
+
"grad_norm": 0.7227723002433777,
|
| 3836 |
+
"learning_rate": 4.134142964589105e-06,
|
| 3837 |
+
"loss": 0.058,
|
| 3838 |
+
"step": 547
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.8921448921448921,
|
| 3842 |
+
"grad_norm": 0.6305646896362305,
|
| 3843 |
+
"learning_rate": 4.01332259148815e-06,
|
| 3844 |
+
"loss": 0.0107,
|
| 3845 |
+
"step": 548
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.8937728937728938,
|
| 3849 |
+
"grad_norm": 1.3228040933609009,
|
| 3850 |
+
"learning_rate": 3.894220143339905e-06,
|
| 3851 |
+
"loss": 0.087,
|
| 3852 |
+
"step": 549
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.8954008954008954,
|
| 3856 |
+
"grad_norm": 2.63421893119812,
|
| 3857 |
+
"learning_rate": 3.776840069456189e-06,
|
| 3858 |
+
"loss": 0.3158,
|
| 3859 |
+
"step": 550
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.897028897028897,
|
| 3863 |
+
"grad_norm": 0.2332572638988495,
|
| 3864 |
+
"learning_rate": 3.6611867548060507e-06,
|
| 3865 |
+
"loss": 0.0344,
|
| 3866 |
+
"step": 551
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.8986568986568987,
|
| 3870 |
+
"grad_norm": 0.09918678551912308,
|
| 3871 |
+
"learning_rate": 3.547264519852006e-06,
|
| 3872 |
+
"loss": 0.0022,
|
| 3873 |
+
"step": 552
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.9002849002849003,
|
| 3877 |
+
"grad_norm": 0.173350527882576,
|
| 3878 |
+
"learning_rate": 3.435077620388577e-06,
|
| 3879 |
+
"loss": 0.0141,
|
| 3880 |
+
"step": 553
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.901912901912902,
|
| 3884 |
+
"grad_norm": 0.8093030452728271,
|
| 3885 |
+
"learning_rate": 3.3246302473833367e-06,
|
| 3886 |
+
"loss": 0.2129,
|
| 3887 |
+
"step": 554
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 0.9035409035409036,
|
| 3891 |
+
"grad_norm": 0.7507438063621521,
|
| 3892 |
+
"learning_rate": 3.215926526820351e-06,
|
| 3893 |
+
"loss": 0.0989,
|
| 3894 |
+
"step": 555
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 0.9051689051689051,
|
| 3898 |
+
"grad_norm": 0.7121350765228271,
|
| 3899 |
+
"learning_rate": 3.108970519546034e-06,
|
| 3900 |
+
"loss": 0.0491,
|
| 3901 |
+
"step": 556
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 0.9067969067969068,
|
| 3905 |
+
"grad_norm": 0.49473196268081665,
|
| 3906 |
+
"learning_rate": 3.0037662211174434e-06,
|
| 3907 |
+
"loss": 0.0481,
|
| 3908 |
+
"step": 557
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 0.9084249084249084,
|
| 3912 |
+
"grad_norm": 0.6143136620521545,
|
| 3913 |
+
"learning_rate": 2.9003175616530265e-06,
|
| 3914 |
+
"loss": 0.0706,
|
| 3915 |
+
"step": 558
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 0.91005291005291,
|
| 3919 |
+
"grad_norm": 0.4149825870990753,
|
| 3920 |
+
"learning_rate": 2.798628405685827e-06,
|
| 3921 |
+
"loss": 0.0424,
|
| 3922 |
+
"step": 559
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 0.9116809116809117,
|
| 3926 |
+
"grad_norm": 0.659127414226532,
|
| 3927 |
+
"learning_rate": 2.6987025520190447e-06,
|
| 3928 |
+
"loss": 0.0724,
|
| 3929 |
+
"step": 560
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.9133089133089133,
|
| 3933 |
+
"grad_norm": 0.48161572217941284,
|
| 3934 |
+
"learning_rate": 2.600543733584215e-06,
|
| 3935 |
+
"loss": 0.0414,
|
| 3936 |
+
"step": 561
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 0.914936914936915,
|
| 3940 |
+
"grad_norm": 0.3644472062587738,
|
| 3941 |
+
"learning_rate": 2.50415561730169e-06,
|
| 3942 |
+
"loss": 0.0234,
|
| 3943 |
+
"step": 562
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.9165649165649166,
|
| 3947 |
+
"grad_norm": 0.34914886951446533,
|
| 3948 |
+
"learning_rate": 2.4095418039436655e-06,
|
| 3949 |
+
"loss": 0.0272,
|
| 3950 |
+
"step": 563
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.9181929181929182,
|
| 3954 |
+
"grad_norm": 0.4708094298839569,
|
| 3955 |
+
"learning_rate": 2.3167058279997156e-06,
|
| 3956 |
+
"loss": 0.0213,
|
| 3957 |
+
"step": 564
|
| 3958 |
+
},
|
| 3959 |
+
{
|
| 3960 |
+
"epoch": 0.9198209198209198,
|
| 3961 |
+
"grad_norm": 0.49146759510040283,
|
| 3962 |
+
"learning_rate": 2.2256511575446836e-06,
|
| 3963 |
+
"loss": 0.0422,
|
| 3964 |
+
"step": 565
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 0.9214489214489214,
|
| 3968 |
+
"grad_norm": 0.3840540051460266,
|
| 3969 |
+
"learning_rate": 2.136381194109166e-06,
|
| 3970 |
+
"loss": 0.024,
|
| 3971 |
+
"step": 566
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 0.9230769230769231,
|
| 3975 |
+
"grad_norm": 0.5289592742919922,
|
| 3976 |
+
"learning_rate": 2.048899272552457e-06,
|
| 3977 |
+
"loss": 0.0557,
|
| 3978 |
+
"step": 567
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 0.9247049247049247,
|
| 3982 |
+
"grad_norm": 0.3621179461479187,
|
| 3983 |
+
"learning_rate": 1.963208660937904e-06,
|
| 3984 |
+
"loss": 0.0231,
|
| 3985 |
+
"step": 568
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.9263329263329263,
|
| 3989 |
+
"grad_norm": 0.39359891414642334,
|
| 3990 |
+
"learning_rate": 1.8793125604109007e-06,
|
| 3991 |
+
"loss": 0.0304,
|
| 3992 |
+
"step": 569
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 0.927960927960928,
|
| 3996 |
+
"grad_norm": 0.10071241110563278,
|
| 3997 |
+
"learning_rate": 1.797214105079248e-06,
|
| 3998 |
+
"loss": 0.0043,
|
| 3999 |
+
"step": 570
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 0.9295889295889296,
|
| 4003 |
+
"grad_norm": 0.4732748866081238,
|
| 4004 |
+
"learning_rate": 1.7169163618960814e-06,
|
| 4005 |
+
"loss": 0.0399,
|
| 4006 |
+
"step": 571
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 0.9312169312169312,
|
| 4010 |
+
"grad_norm": 0.572767436504364,
|
| 4011 |
+
"learning_rate": 1.6384223305453416e-06,
|
| 4012 |
+
"loss": 0.052,
|
| 4013 |
+
"step": 572
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.9328449328449329,
|
| 4017 |
+
"grad_norm": 1.0503902435302734,
|
| 4018 |
+
"learning_rate": 1.5617349433296403e-06,
|
| 4019 |
+
"loss": 0.0512,
|
| 4020 |
+
"step": 573
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 0.9344729344729344,
|
| 4024 |
+
"grad_norm": 0.798310399055481,
|
| 4025 |
+
"learning_rate": 1.4868570650607815e-06,
|
| 4026 |
+
"loss": 0.0763,
|
| 4027 |
+
"step": 574
|
| 4028 |
+
},
|
| 4029 |
+
{
|
| 4030 |
+
"epoch": 0.9361009361009361,
|
| 4031 |
+
"grad_norm": 0.3898273706436157,
|
| 4032 |
+
"learning_rate": 1.4137914929527096e-06,
|
| 4033 |
+
"loss": 0.0358,
|
| 4034 |
+
"step": 575
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 0.9377289377289377,
|
| 4038 |
+
"grad_norm": 0.45904216170310974,
|
| 4039 |
+
"learning_rate": 1.342540956517041e-06,
|
| 4040 |
+
"loss": 0.0437,
|
| 4041 |
+
"step": 576
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 0.9393569393569393,
|
| 4045 |
+
"grad_norm": 1.0378749370574951,
|
| 4046 |
+
"learning_rate": 1.2731081174610526e-06,
|
| 4047 |
+
"loss": 0.1158,
|
| 4048 |
+
"step": 577
|
| 4049 |
+
},
|
| 4050 |
+
{
|
| 4051 |
+
"epoch": 0.940984940984941,
|
| 4052 |
+
"grad_norm": 0.831689715385437,
|
| 4053 |
+
"learning_rate": 1.205495569588283e-06,
|
| 4054 |
+
"loss": 0.1591,
|
| 4055 |
+
"step": 578
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 0.9426129426129426,
|
| 4059 |
+
"grad_norm": 0.9298439621925354,
|
| 4060 |
+
"learning_rate": 1.139705838701638e-06,
|
| 4061 |
+
"loss": 0.1746,
|
| 4062 |
+
"step": 579
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 0.9442409442409443,
|
| 4066 |
+
"grad_norm": 1.2845919132232666,
|
| 4067 |
+
"learning_rate": 1.0757413825090212e-06,
|
| 4068 |
+
"loss": 0.1089,
|
| 4069 |
+
"step": 580
|
| 4070 |
+
},
|
| 4071 |
+
{
|
| 4072 |
+
"epoch": 0.9458689458689459,
|
| 4073 |
+
"grad_norm": 1.1161198616027832,
|
| 4074 |
+
"learning_rate": 1.0136045905315028e-06,
|
| 4075 |
+
"loss": 0.1823,
|
| 4076 |
+
"step": 581
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 0.9474969474969475,
|
| 4080 |
+
"grad_norm": 0.9601607918739319,
|
| 4081 |
+
"learning_rate": 9.532977840141122e-07,
|
| 4082 |
+
"loss": 0.0872,
|
| 4083 |
+
"step": 582
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 0.9491249491249492,
|
| 4087 |
+
"grad_norm": 1.4678707122802734,
|
| 4088 |
+
"learning_rate": 8.948232158390468e-07,
|
| 4089 |
+
"loss": 0.2021,
|
| 4090 |
+
"step": 583
|
| 4091 |
+
},
|
| 4092 |
+
{
|
| 4093 |
+
"epoch": 0.9507529507529507,
|
| 4094 |
+
"grad_norm": 1.3035527467727661,
|
| 4095 |
+
"learning_rate": 8.381830704415839e-07,
|
| 4096 |
+
"loss": 0.1604,
|
| 4097 |
+
"step": 584
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 0.9523809523809523,
|
| 4101 |
+
"grad_norm": 0.9295395612716675,
|
| 4102 |
+
"learning_rate": 7.833794637284231e-07,
|
| 4103 |
+
"loss": 0.0591,
|
| 4104 |
+
"step": 585
|
| 4105 |
+
},
|
| 4106 |
+
{
|
| 4107 |
+
"epoch": 0.954008954008954,
|
| 4108 |
+
"grad_norm": 1.5823544263839722,
|
| 4109 |
+
"learning_rate": 7.30414442998667e-07,
|
| 4110 |
+
"loss": 0.1536,
|
| 4111 |
+
"step": 586
|
| 4112 |
+
},
|
| 4113 |
+
{
|
| 4114 |
+
"epoch": 0.9556369556369556,
|
| 4115 |
+
"grad_norm": 1.4963399171829224,
|
| 4116 |
+
"learning_rate": 6.792899868673486e-07,
|
| 4117 |
+
"loss": 0.1601,
|
| 4118 |
+
"step": 587
|
| 4119 |
+
},
|
| 4120 |
+
{
|
| 4121 |
+
"epoch": 0.9572649572649573,
|
| 4122 |
+
"grad_norm": 1.1686265468597412,
|
| 4123 |
+
"learning_rate": 6.300080051914791e-07,
|
| 4124 |
+
"loss": 0.0715,
|
| 4125 |
+
"step": 588
|
| 4126 |
+
},
|
| 4127 |
+
{
|
| 4128 |
+
"epoch": 0.9588929588929589,
|
| 4129 |
+
"grad_norm": 0.9449349641799927,
|
| 4130 |
+
"learning_rate": 5.825703389987391e-07,
|
| 4131 |
+
"loss": 0.071,
|
| 4132 |
+
"step": 589
|
| 4133 |
+
},
|
| 4134 |
+
{
|
| 4135 |
+
"epoch": 0.9605209605209605,
|
| 4136 |
+
"grad_norm": 1.001676321029663,
|
| 4137 |
+
"learning_rate": 5.369787604186993e-07,
|
| 4138 |
+
"loss": 0.0702,
|
| 4139 |
+
"step": 590
|
| 4140 |
+
},
|
| 4141 |
+
{
|
| 4142 |
+
"epoch": 0.9621489621489622,
|
| 4143 |
+
"grad_norm": 1.1436750888824463,
|
| 4144 |
+
"learning_rate": 4.932349726165964e-07,
|
| 4145 |
+
"loss": 0.0824,
|
| 4146 |
+
"step": 591
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 0.9637769637769638,
|
| 4150 |
+
"grad_norm": 1.280293583869934,
|
| 4151 |
+
"learning_rate": 4.513406097297224e-07,
|
| 4152 |
+
"loss": 0.0754,
|
| 4153 |
+
"step": 592
|
| 4154 |
+
},
|
| 4155 |
+
{
|
| 4156 |
+
"epoch": 0.9654049654049655,
|
| 4157 |
+
"grad_norm": 1.658920168876648,
|
| 4158 |
+
"learning_rate": 4.1129723680637946e-07,
|
| 4159 |
+
"loss": 0.1284,
|
| 4160 |
+
"step": 593
|
| 4161 |
+
},
|
| 4162 |
+
{
|
| 4163 |
+
"epoch": 0.967032967032967,
|
| 4164 |
+
"grad_norm": 0.8406931757926941,
|
| 4165 |
+
"learning_rate": 3.7310634974741523e-07,
|
| 4166 |
+
"loss": 0.0539,
|
| 4167 |
+
"step": 594
|
| 4168 |
+
},
|
| 4169 |
+
{
|
| 4170 |
+
"epoch": 0.9686609686609686,
|
| 4171 |
+
"grad_norm": 0.9294666647911072,
|
| 4172 |
+
"learning_rate": 3.3676937525032314e-07,
|
| 4173 |
+
"loss": 0.0635,
|
| 4174 |
+
"step": 595
|
| 4175 |
+
},
|
| 4176 |
+
{
|
| 4177 |
+
"epoch": 0.9702889702889703,
|
| 4178 |
+
"grad_norm": 1.2156299352645874,
|
| 4179 |
+
"learning_rate": 3.022876707559796e-07,
|
| 4180 |
+
"loss": 0.074,
|
| 4181 |
+
"step": 596
|
| 4182 |
+
},
|
| 4183 |
+
{
|
| 4184 |
+
"epoch": 0.9719169719169719,
|
| 4185 |
+
"grad_norm": 1.1088374853134155,
|
| 4186 |
+
"learning_rate": 2.696625243979012e-07,
|
| 4187 |
+
"loss": 0.0819,
|
| 4188 |
+
"step": 597
|
| 4189 |
+
},
|
| 4190 |
+
{
|
| 4191 |
+
"epoch": 0.9735449735449735,
|
| 4192 |
+
"grad_norm": 1.7963135242462158,
|
| 4193 |
+
"learning_rate": 2.3889515495413296e-07,
|
| 4194 |
+
"loss": 0.2016,
|
| 4195 |
+
"step": 598
|
| 4196 |
+
},
|
| 4197 |
+
{
|
| 4198 |
+
"epoch": 0.9751729751729752,
|
| 4199 |
+
"grad_norm": 2.0805530548095703,
|
| 4200 |
+
"learning_rate": 2.0998671180172957e-07,
|
| 4201 |
+
"loss": 0.2091,
|
| 4202 |
+
"step": 599
|
| 4203 |
+
},
|
| 4204 |
+
{
|
| 4205 |
+
"epoch": 0.9768009768009768,
|
| 4206 |
+
"grad_norm": 2.48085618019104,
|
| 4207 |
+
"learning_rate": 1.829382748738062e-07,
|
| 4208 |
+
"loss": 0.3043,
|
| 4209 |
+
"step": 600
|
| 4210 |
+
},
|
| 4211 |
+
{
|
| 4212 |
+
"epoch": 0.9784289784289785,
|
| 4213 |
+
"grad_norm": 0.025913316756486893,
|
| 4214 |
+
"learning_rate": 1.5775085461919858e-07,
|
| 4215 |
+
"loss": 0.0012,
|
| 4216 |
+
"step": 601
|
| 4217 |
+
},
|
| 4218 |
+
{
|
| 4219 |
+
"epoch": 0.98005698005698,
|
| 4220 |
+
"grad_norm": 0.6956274509429932,
|
| 4221 |
+
"learning_rate": 1.3442539196472648e-07,
|
| 4222 |
+
"loss": 0.1641,
|
| 4223 |
+
"step": 602
|
| 4224 |
+
},
|
| 4225 |
+
{
|
| 4226 |
+
"epoch": 0.9816849816849816,
|
| 4227 |
+
"grad_norm": 0.24362115561962128,
|
| 4228 |
+
"learning_rate": 1.1296275828001635e-07,
|
| 4229 |
+
"loss": 0.0106,
|
| 4230 |
+
"step": 603
|
| 4231 |
+
},
|
| 4232 |
+
{
|
| 4233 |
+
"epoch": 0.9833129833129833,
|
| 4234 |
+
"grad_norm": 0.3809750974178314,
|
| 4235 |
+
"learning_rate": 9.336375534497733e-08,
|
| 4236 |
+
"loss": 0.0239,
|
| 4237 |
+
"step": 604
|
| 4238 |
+
},
|
| 4239 |
+
{
|
| 4240 |
+
"epoch": 0.9849409849409849,
|
| 4241 |
+
"grad_norm": 0.530231773853302,
|
| 4242 |
+
"learning_rate": 7.56291153198363e-08,
|
| 4243 |
+
"loss": 0.042,
|
| 4244 |
+
"step": 605
|
| 4245 |
+
},
|
| 4246 |
+
{
|
| 4247 |
+
"epoch": 0.9865689865689866,
|
| 4248 |
+
"grad_norm": 0.7303449511528015,
|
| 4249 |
+
"learning_rate": 5.975950071779312e-08,
|
| 4250 |
+
"loss": 0.0821,
|
| 4251 |
+
"step": 606
|
| 4252 |
+
},
|
| 4253 |
+
{
|
| 4254 |
+
"epoch": 0.9881969881969882,
|
| 4255 |
+
"grad_norm": 0.4484477937221527,
|
| 4256 |
+
"learning_rate": 4.575550438026266e-08,
|
| 4257 |
+
"loss": 0.0574,
|
| 4258 |
+
"step": 607
|
| 4259 |
+
},
|
| 4260 |
+
{
|
| 4261 |
+
"epoch": 0.9898249898249898,
|
| 4262 |
+
"grad_norm": 0.6158083081245422,
|
| 4263 |
+
"learning_rate": 3.361764945473134e-08,
|
| 4264 |
+
"loss": 0.0628,
|
| 4265 |
+
"step": 608
|
| 4266 |
+
},
|
| 4267 |
+
{
|
| 4268 |
+
"epoch": 0.9914529914529915,
|
| 4269 |
+
"grad_norm": 1.289014220237732,
|
| 4270 |
+
"learning_rate": 2.3346389375211718e-08,
|
| 4271 |
+
"loss": 0.2014,
|
| 4272 |
+
"step": 609
|
| 4273 |
+
},
|
| 4274 |
+
{
|
| 4275 |
+
"epoch": 0.9930809930809931,
|
| 4276 |
+
"grad_norm": 1.2264184951782227,
|
| 4277 |
+
"learning_rate": 1.4942107845317133e-08,
|
| 4278 |
+
"loss": 0.1331,
|
| 4279 |
+
"step": 610
|
| 4280 |
+
},
|
| 4281 |
+
{
|
| 4282 |
+
"epoch": 0.9947089947089947,
|
| 4283 |
+
"grad_norm": 0.9105086922645569,
|
| 4284 |
+
"learning_rate": 8.405118823906489e-09,
|
| 4285 |
+
"loss": 0.0738,
|
| 4286 |
+
"step": 611
|
| 4287 |
+
},
|
| 4288 |
+
{
|
| 4289 |
+
"epoch": 0.9963369963369964,
|
| 4290 |
+
"grad_norm": 1.066785454750061,
|
| 4291 |
+
"learning_rate": 3.7356665133714274e-09,
|
| 4292 |
+
"loss": 0.0894,
|
| 4293 |
+
"step": 612
|
| 4294 |
+
},
|
| 4295 |
+
{
|
| 4296 |
+
"epoch": 0.9979649979649979,
|
| 4297 |
+
"grad_norm": 1.2152076959609985,
|
| 4298 |
+
"learning_rate": 9.339253505102896e-10,
|
| 4299 |
+
"loss": 0.1075,
|
| 4300 |
+
"step": 613
|
| 4301 |
+
},
|
| 4302 |
+
{
|
| 4303 |
+
"epoch": 0.9995929995929996,
|
| 4304 |
+
"grad_norm": 1.6802031993865967,
|
| 4305 |
+
"learning_rate": 0.0,
|
| 4306 |
+
"loss": 0.1659,
|
| 4307 |
+
"step": 614
|
| 4308 |
}
|
| 4309 |
],
|
| 4310 |
"logging_steps": 1,
|
|
|
|
| 4319 |
"should_evaluate": false,
|
| 4320 |
"should_log": false,
|
| 4321 |
"should_save": true,
|
| 4322 |
+
"should_training_stop": true
|
| 4323 |
},
|
| 4324 |
"attributes": {}
|
| 4325 |
}
|
| 4326 |
},
|
| 4327 |
+
"total_flos": 2.4815911417203917e+17,
|
| 4328 |
"train_batch_size": 4,
|
| 4329 |
"trial_name": null,
|
| 4330 |
"trial_params": null
|