Training in progress, step 6000, 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 685354800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3fab20f32621c18f7e567e201b068fd4e502966a6cf5803e5436f2004a34e1fe
|
| 3 |
size 685354800
|
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": 4.
|
| 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,
|
|
@@ -41308,6 +41308,706 @@
|
|
| 41308 |
"learning_rate": 0.0005,
|
| 41309 |
"loss": 1.3641,
|
| 41310 |
"step": 5900
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41311 |
}
|
| 41312 |
],
|
| 41313 |
"logging_steps": 1,
|
|
@@ -41322,12 +42022,12 @@
|
|
| 41322 |
"should_evaluate": false,
|
| 41323 |
"should_log": false,
|
| 41324 |
"should_save": true,
|
| 41325 |
-
"should_training_stop":
|
| 41326 |
},
|
| 41327 |
"attributes": {}
|
| 41328 |
}
|
| 41329 |
},
|
| 41330 |
-
"total_flos": 1.
|
| 41331 |
"train_batch_size": 8,
|
| 41332 |
"trial_name": null,
|
| 41333 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 4.930156121610517,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 6000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 41308 |
"learning_rate": 0.0005,
|
| 41309 |
"loss": 1.3641,
|
| 41310 |
"step": 5900
|
| 41311 |
+
},
|
| 41312 |
+
{
|
| 41313 |
+
"epoch": 4.848808545603944,
|
| 41314 |
+
"grad_norm": 0.5559499859809875,
|
| 41315 |
+
"learning_rate": 0.0005,
|
| 41316 |
+
"loss": 1.3687,
|
| 41317 |
+
"step": 5901
|
| 41318 |
+
},
|
| 41319 |
+
{
|
| 41320 |
+
"epoch": 4.849630238290879,
|
| 41321 |
+
"grad_norm": 0.535542905330658,
|
| 41322 |
+
"learning_rate": 0.0005,
|
| 41323 |
+
"loss": 1.2836,
|
| 41324 |
+
"step": 5902
|
| 41325 |
+
},
|
| 41326 |
+
{
|
| 41327 |
+
"epoch": 4.8504519309778145,
|
| 41328 |
+
"grad_norm": 0.5417159199714661,
|
| 41329 |
+
"learning_rate": 0.0005,
|
| 41330 |
+
"loss": 1.3408,
|
| 41331 |
+
"step": 5903
|
| 41332 |
+
},
|
| 41333 |
+
{
|
| 41334 |
+
"epoch": 4.851273623664749,
|
| 41335 |
+
"grad_norm": 0.543630838394165,
|
| 41336 |
+
"learning_rate": 0.0005,
|
| 41337 |
+
"loss": 1.3996,
|
| 41338 |
+
"step": 5904
|
| 41339 |
+
},
|
| 41340 |
+
{
|
| 41341 |
+
"epoch": 4.852095316351685,
|
| 41342 |
+
"grad_norm": 0.5195862054824829,
|
| 41343 |
+
"learning_rate": 0.0005,
|
| 41344 |
+
"loss": 1.3169,
|
| 41345 |
+
"step": 5905
|
| 41346 |
+
},
|
| 41347 |
+
{
|
| 41348 |
+
"epoch": 4.852917009038619,
|
| 41349 |
+
"grad_norm": 0.5713696479797363,
|
| 41350 |
+
"learning_rate": 0.0005,
|
| 41351 |
+
"loss": 1.4069,
|
| 41352 |
+
"step": 5906
|
| 41353 |
+
},
|
| 41354 |
+
{
|
| 41355 |
+
"epoch": 4.853738701725555,
|
| 41356 |
+
"grad_norm": 0.5813581943511963,
|
| 41357 |
+
"learning_rate": 0.0005,
|
| 41358 |
+
"loss": 1.3037,
|
| 41359 |
+
"step": 5907
|
| 41360 |
+
},
|
| 41361 |
+
{
|
| 41362 |
+
"epoch": 4.8545603944124895,
|
| 41363 |
+
"grad_norm": 0.5637884140014648,
|
| 41364 |
+
"learning_rate": 0.0005,
|
| 41365 |
+
"loss": 1.3991,
|
| 41366 |
+
"step": 5908
|
| 41367 |
+
},
|
| 41368 |
+
{
|
| 41369 |
+
"epoch": 4.855382087099425,
|
| 41370 |
+
"grad_norm": 0.5755516290664673,
|
| 41371 |
+
"learning_rate": 0.0005,
|
| 41372 |
+
"loss": 1.4464,
|
| 41373 |
+
"step": 5909
|
| 41374 |
+
},
|
| 41375 |
+
{
|
| 41376 |
+
"epoch": 4.85620377978636,
|
| 41377 |
+
"grad_norm": 0.5279106497764587,
|
| 41378 |
+
"learning_rate": 0.0005,
|
| 41379 |
+
"loss": 1.3729,
|
| 41380 |
+
"step": 5910
|
| 41381 |
+
},
|
| 41382 |
+
{
|
| 41383 |
+
"epoch": 4.857025472473295,
|
| 41384 |
+
"grad_norm": 0.5568866729736328,
|
| 41385 |
+
"learning_rate": 0.0005,
|
| 41386 |
+
"loss": 1.3846,
|
| 41387 |
+
"step": 5911
|
| 41388 |
+
},
|
| 41389 |
+
{
|
| 41390 |
+
"epoch": 4.85784716516023,
|
| 41391 |
+
"grad_norm": 0.5258705615997314,
|
| 41392 |
+
"learning_rate": 0.0005,
|
| 41393 |
+
"loss": 1.3301,
|
| 41394 |
+
"step": 5912
|
| 41395 |
+
},
|
| 41396 |
+
{
|
| 41397 |
+
"epoch": 4.8586688578471655,
|
| 41398 |
+
"grad_norm": 0.5256040692329407,
|
| 41399 |
+
"learning_rate": 0.0005,
|
| 41400 |
+
"loss": 1.3543,
|
| 41401 |
+
"step": 5913
|
| 41402 |
+
},
|
| 41403 |
+
{
|
| 41404 |
+
"epoch": 4.8594905505341,
|
| 41405 |
+
"grad_norm": 0.5508636236190796,
|
| 41406 |
+
"learning_rate": 0.0005,
|
| 41407 |
+
"loss": 1.3168,
|
| 41408 |
+
"step": 5914
|
| 41409 |
+
},
|
| 41410 |
+
{
|
| 41411 |
+
"epoch": 4.860312243221035,
|
| 41412 |
+
"grad_norm": 0.5376938581466675,
|
| 41413 |
+
"learning_rate": 0.0005,
|
| 41414 |
+
"loss": 1.3618,
|
| 41415 |
+
"step": 5915
|
| 41416 |
+
},
|
| 41417 |
+
{
|
| 41418 |
+
"epoch": 4.86113393590797,
|
| 41419 |
+
"grad_norm": 0.513633668422699,
|
| 41420 |
+
"learning_rate": 0.0005,
|
| 41421 |
+
"loss": 1.279,
|
| 41422 |
+
"step": 5916
|
| 41423 |
+
},
|
| 41424 |
+
{
|
| 41425 |
+
"epoch": 4.861955628594906,
|
| 41426 |
+
"grad_norm": 0.5526177287101746,
|
| 41427 |
+
"learning_rate": 0.0005,
|
| 41428 |
+
"loss": 1.2778,
|
| 41429 |
+
"step": 5917
|
| 41430 |
+
},
|
| 41431 |
+
{
|
| 41432 |
+
"epoch": 4.862777321281841,
|
| 41433 |
+
"grad_norm": 0.5580543279647827,
|
| 41434 |
+
"learning_rate": 0.0005,
|
| 41435 |
+
"loss": 1.3931,
|
| 41436 |
+
"step": 5918
|
| 41437 |
+
},
|
| 41438 |
+
{
|
| 41439 |
+
"epoch": 4.863599013968775,
|
| 41440 |
+
"grad_norm": 0.5640444159507751,
|
| 41441 |
+
"learning_rate": 0.0005,
|
| 41442 |
+
"loss": 1.4172,
|
| 41443 |
+
"step": 5919
|
| 41444 |
+
},
|
| 41445 |
+
{
|
| 41446 |
+
"epoch": 4.864420706655711,
|
| 41447 |
+
"grad_norm": 0.5366266369819641,
|
| 41448 |
+
"learning_rate": 0.0005,
|
| 41449 |
+
"loss": 1.2472,
|
| 41450 |
+
"step": 5920
|
| 41451 |
+
},
|
| 41452 |
+
{
|
| 41453 |
+
"epoch": 4.8652423993426455,
|
| 41454 |
+
"grad_norm": 0.5291760563850403,
|
| 41455 |
+
"learning_rate": 0.0005,
|
| 41456 |
+
"loss": 1.2402,
|
| 41457 |
+
"step": 5921
|
| 41458 |
+
},
|
| 41459 |
+
{
|
| 41460 |
+
"epoch": 4.866064092029581,
|
| 41461 |
+
"grad_norm": 0.5365450978279114,
|
| 41462 |
+
"learning_rate": 0.0005,
|
| 41463 |
+
"loss": 1.2807,
|
| 41464 |
+
"step": 5922
|
| 41465 |
+
},
|
| 41466 |
+
{
|
| 41467 |
+
"epoch": 4.866885784716516,
|
| 41468 |
+
"grad_norm": 0.5284454822540283,
|
| 41469 |
+
"learning_rate": 0.0005,
|
| 41470 |
+
"loss": 1.4549,
|
| 41471 |
+
"step": 5923
|
| 41472 |
+
},
|
| 41473 |
+
{
|
| 41474 |
+
"epoch": 4.867707477403451,
|
| 41475 |
+
"grad_norm": 0.541029155254364,
|
| 41476 |
+
"learning_rate": 0.0005,
|
| 41477 |
+
"loss": 1.3154,
|
| 41478 |
+
"step": 5924
|
| 41479 |
+
},
|
| 41480 |
+
{
|
| 41481 |
+
"epoch": 4.868529170090386,
|
| 41482 |
+
"grad_norm": 0.5600323677062988,
|
| 41483 |
+
"learning_rate": 0.0005,
|
| 41484 |
+
"loss": 1.3268,
|
| 41485 |
+
"step": 5925
|
| 41486 |
+
},
|
| 41487 |
+
{
|
| 41488 |
+
"epoch": 4.869350862777321,
|
| 41489 |
+
"grad_norm": 0.5529478788375854,
|
| 41490 |
+
"learning_rate": 0.0005,
|
| 41491 |
+
"loss": 1.3629,
|
| 41492 |
+
"step": 5926
|
| 41493 |
+
},
|
| 41494 |
+
{
|
| 41495 |
+
"epoch": 4.870172555464256,
|
| 41496 |
+
"grad_norm": 0.5471394658088684,
|
| 41497 |
+
"learning_rate": 0.0005,
|
| 41498 |
+
"loss": 1.2708,
|
| 41499 |
+
"step": 5927
|
| 41500 |
+
},
|
| 41501 |
+
{
|
| 41502 |
+
"epoch": 4.870994248151192,
|
| 41503 |
+
"grad_norm": 0.5359401702880859,
|
| 41504 |
+
"learning_rate": 0.0005,
|
| 41505 |
+
"loss": 1.3363,
|
| 41506 |
+
"step": 5928
|
| 41507 |
+
},
|
| 41508 |
+
{
|
| 41509 |
+
"epoch": 4.871815940838126,
|
| 41510 |
+
"grad_norm": 0.5148870348930359,
|
| 41511 |
+
"learning_rate": 0.0005,
|
| 41512 |
+
"loss": 1.2534,
|
| 41513 |
+
"step": 5929
|
| 41514 |
+
},
|
| 41515 |
+
{
|
| 41516 |
+
"epoch": 4.872637633525062,
|
| 41517 |
+
"grad_norm": 0.5293238759040833,
|
| 41518 |
+
"learning_rate": 0.0005,
|
| 41519 |
+
"loss": 1.3152,
|
| 41520 |
+
"step": 5930
|
| 41521 |
+
},
|
| 41522 |
+
{
|
| 41523 |
+
"epoch": 4.8734593262119965,
|
| 41524 |
+
"grad_norm": 0.5174310207366943,
|
| 41525 |
+
"learning_rate": 0.0005,
|
| 41526 |
+
"loss": 1.3534,
|
| 41527 |
+
"step": 5931
|
| 41528 |
+
},
|
| 41529 |
+
{
|
| 41530 |
+
"epoch": 4.874281018898932,
|
| 41531 |
+
"grad_norm": 0.5341978073120117,
|
| 41532 |
+
"learning_rate": 0.0005,
|
| 41533 |
+
"loss": 1.3888,
|
| 41534 |
+
"step": 5932
|
| 41535 |
+
},
|
| 41536 |
+
{
|
| 41537 |
+
"epoch": 4.875102711585867,
|
| 41538 |
+
"grad_norm": 0.5729701519012451,
|
| 41539 |
+
"learning_rate": 0.0005,
|
| 41540 |
+
"loss": 1.4173,
|
| 41541 |
+
"step": 5933
|
| 41542 |
+
},
|
| 41543 |
+
{
|
| 41544 |
+
"epoch": 4.875924404272802,
|
| 41545 |
+
"grad_norm": 0.6612470746040344,
|
| 41546 |
+
"learning_rate": 0.0005,
|
| 41547 |
+
"loss": 1.3927,
|
| 41548 |
+
"step": 5934
|
| 41549 |
+
},
|
| 41550 |
+
{
|
| 41551 |
+
"epoch": 4.876746096959737,
|
| 41552 |
+
"grad_norm": 0.5212023258209229,
|
| 41553 |
+
"learning_rate": 0.0005,
|
| 41554 |
+
"loss": 1.3259,
|
| 41555 |
+
"step": 5935
|
| 41556 |
+
},
|
| 41557 |
+
{
|
| 41558 |
+
"epoch": 4.8775677896466725,
|
| 41559 |
+
"grad_norm": 0.5570096969604492,
|
| 41560 |
+
"learning_rate": 0.0005,
|
| 41561 |
+
"loss": 1.2884,
|
| 41562 |
+
"step": 5936
|
| 41563 |
+
},
|
| 41564 |
+
{
|
| 41565 |
+
"epoch": 4.878389482333607,
|
| 41566 |
+
"grad_norm": 0.5613424181938171,
|
| 41567 |
+
"learning_rate": 0.0005,
|
| 41568 |
+
"loss": 1.2463,
|
| 41569 |
+
"step": 5937
|
| 41570 |
+
},
|
| 41571 |
+
{
|
| 41572 |
+
"epoch": 4.879211175020543,
|
| 41573 |
+
"grad_norm": 0.5613168478012085,
|
| 41574 |
+
"learning_rate": 0.0005,
|
| 41575 |
+
"loss": 1.3115,
|
| 41576 |
+
"step": 5938
|
| 41577 |
+
},
|
| 41578 |
+
{
|
| 41579 |
+
"epoch": 4.880032867707477,
|
| 41580 |
+
"grad_norm": 0.535275936126709,
|
| 41581 |
+
"learning_rate": 0.0005,
|
| 41582 |
+
"loss": 1.3002,
|
| 41583 |
+
"step": 5939
|
| 41584 |
+
},
|
| 41585 |
+
{
|
| 41586 |
+
"epoch": 4.880854560394413,
|
| 41587 |
+
"grad_norm": 0.5672900676727295,
|
| 41588 |
+
"learning_rate": 0.0005,
|
| 41589 |
+
"loss": 1.3669,
|
| 41590 |
+
"step": 5940
|
| 41591 |
+
},
|
| 41592 |
+
{
|
| 41593 |
+
"epoch": 4.881676253081348,
|
| 41594 |
+
"grad_norm": 0.5469388961791992,
|
| 41595 |
+
"learning_rate": 0.0005,
|
| 41596 |
+
"loss": 1.3652,
|
| 41597 |
+
"step": 5941
|
| 41598 |
+
},
|
| 41599 |
+
{
|
| 41600 |
+
"epoch": 4.882497945768282,
|
| 41601 |
+
"grad_norm": 0.529625415802002,
|
| 41602 |
+
"learning_rate": 0.0005,
|
| 41603 |
+
"loss": 1.4422,
|
| 41604 |
+
"step": 5942
|
| 41605 |
+
},
|
| 41606 |
+
{
|
| 41607 |
+
"epoch": 4.883319638455218,
|
| 41608 |
+
"grad_norm": 0.5889802575111389,
|
| 41609 |
+
"learning_rate": 0.0005,
|
| 41610 |
+
"loss": 1.3788,
|
| 41611 |
+
"step": 5943
|
| 41612 |
+
},
|
| 41613 |
+
{
|
| 41614 |
+
"epoch": 4.884141331142153,
|
| 41615 |
+
"grad_norm": 0.5382450819015503,
|
| 41616 |
+
"learning_rate": 0.0005,
|
| 41617 |
+
"loss": 1.3318,
|
| 41618 |
+
"step": 5944
|
| 41619 |
+
},
|
| 41620 |
+
{
|
| 41621 |
+
"epoch": 4.884963023829088,
|
| 41622 |
+
"grad_norm": 0.5242584347724915,
|
| 41623 |
+
"learning_rate": 0.0005,
|
| 41624 |
+
"loss": 1.268,
|
| 41625 |
+
"step": 5945
|
| 41626 |
+
},
|
| 41627 |
+
{
|
| 41628 |
+
"epoch": 4.885784716516023,
|
| 41629 |
+
"grad_norm": 0.5483070611953735,
|
| 41630 |
+
"learning_rate": 0.0005,
|
| 41631 |
+
"loss": 1.4016,
|
| 41632 |
+
"step": 5946
|
| 41633 |
+
},
|
| 41634 |
+
{
|
| 41635 |
+
"epoch": 4.886606409202958,
|
| 41636 |
+
"grad_norm": 0.5779204368591309,
|
| 41637 |
+
"learning_rate": 0.0005,
|
| 41638 |
+
"loss": 1.327,
|
| 41639 |
+
"step": 5947
|
| 41640 |
+
},
|
| 41641 |
+
{
|
| 41642 |
+
"epoch": 4.887428101889893,
|
| 41643 |
+
"grad_norm": 0.5551186800003052,
|
| 41644 |
+
"learning_rate": 0.0005,
|
| 41645 |
+
"loss": 1.4228,
|
| 41646 |
+
"step": 5948
|
| 41647 |
+
},
|
| 41648 |
+
{
|
| 41649 |
+
"epoch": 4.888249794576828,
|
| 41650 |
+
"grad_norm": 0.5995270609855652,
|
| 41651 |
+
"learning_rate": 0.0005,
|
| 41652 |
+
"loss": 1.3831,
|
| 41653 |
+
"step": 5949
|
| 41654 |
+
},
|
| 41655 |
+
{
|
| 41656 |
+
"epoch": 4.889071487263763,
|
| 41657 |
+
"grad_norm": 0.5424328446388245,
|
| 41658 |
+
"learning_rate": 0.0005,
|
| 41659 |
+
"loss": 1.3815,
|
| 41660 |
+
"step": 5950
|
| 41661 |
+
},
|
| 41662 |
+
{
|
| 41663 |
+
"epoch": 4.889893179950699,
|
| 41664 |
+
"grad_norm": 0.5349864959716797,
|
| 41665 |
+
"learning_rate": 0.0005,
|
| 41666 |
+
"loss": 1.2749,
|
| 41667 |
+
"step": 5951
|
| 41668 |
+
},
|
| 41669 |
+
{
|
| 41670 |
+
"epoch": 4.890714872637633,
|
| 41671 |
+
"grad_norm": 0.5398481488227844,
|
| 41672 |
+
"learning_rate": 0.0005,
|
| 41673 |
+
"loss": 1.3635,
|
| 41674 |
+
"step": 5952
|
| 41675 |
+
},
|
| 41676 |
+
{
|
| 41677 |
+
"epoch": 4.891536565324569,
|
| 41678 |
+
"grad_norm": 0.5872131586074829,
|
| 41679 |
+
"learning_rate": 0.0005,
|
| 41680 |
+
"loss": 1.3778,
|
| 41681 |
+
"step": 5953
|
| 41682 |
+
},
|
| 41683 |
+
{
|
| 41684 |
+
"epoch": 4.8923582580115035,
|
| 41685 |
+
"grad_norm": 0.5435046553611755,
|
| 41686 |
+
"learning_rate": 0.0005,
|
| 41687 |
+
"loss": 1.3834,
|
| 41688 |
+
"step": 5954
|
| 41689 |
+
},
|
| 41690 |
+
{
|
| 41691 |
+
"epoch": 4.893179950698439,
|
| 41692 |
+
"grad_norm": 0.551789402961731,
|
| 41693 |
+
"learning_rate": 0.0005,
|
| 41694 |
+
"loss": 1.3315,
|
| 41695 |
+
"step": 5955
|
| 41696 |
+
},
|
| 41697 |
+
{
|
| 41698 |
+
"epoch": 4.894001643385374,
|
| 41699 |
+
"grad_norm": 0.5277841687202454,
|
| 41700 |
+
"learning_rate": 0.0005,
|
| 41701 |
+
"loss": 1.4072,
|
| 41702 |
+
"step": 5956
|
| 41703 |
+
},
|
| 41704 |
+
{
|
| 41705 |
+
"epoch": 4.894823336072309,
|
| 41706 |
+
"grad_norm": 0.5183998346328735,
|
| 41707 |
+
"learning_rate": 0.0005,
|
| 41708 |
+
"loss": 1.2465,
|
| 41709 |
+
"step": 5957
|
| 41710 |
+
},
|
| 41711 |
+
{
|
| 41712 |
+
"epoch": 4.895645028759244,
|
| 41713 |
+
"grad_norm": 0.561173677444458,
|
| 41714 |
+
"learning_rate": 0.0005,
|
| 41715 |
+
"loss": 1.4095,
|
| 41716 |
+
"step": 5958
|
| 41717 |
+
},
|
| 41718 |
+
{
|
| 41719 |
+
"epoch": 4.8964667214461794,
|
| 41720 |
+
"grad_norm": 0.5499217510223389,
|
| 41721 |
+
"learning_rate": 0.0005,
|
| 41722 |
+
"loss": 1.312,
|
| 41723 |
+
"step": 5959
|
| 41724 |
+
},
|
| 41725 |
+
{
|
| 41726 |
+
"epoch": 4.897288414133114,
|
| 41727 |
+
"grad_norm": 0.534344494342804,
|
| 41728 |
+
"learning_rate": 0.0005,
|
| 41729 |
+
"loss": 1.326,
|
| 41730 |
+
"step": 5960
|
| 41731 |
+
},
|
| 41732 |
+
{
|
| 41733 |
+
"epoch": 4.89811010682005,
|
| 41734 |
+
"grad_norm": 0.5524152517318726,
|
| 41735 |
+
"learning_rate": 0.0005,
|
| 41736 |
+
"loss": 1.3709,
|
| 41737 |
+
"step": 5961
|
| 41738 |
+
},
|
| 41739 |
+
{
|
| 41740 |
+
"epoch": 4.898931799506984,
|
| 41741 |
+
"grad_norm": 0.5235154032707214,
|
| 41742 |
+
"learning_rate": 0.0005,
|
| 41743 |
+
"loss": 1.212,
|
| 41744 |
+
"step": 5962
|
| 41745 |
+
},
|
| 41746 |
+
{
|
| 41747 |
+
"epoch": 4.89975349219392,
|
| 41748 |
+
"grad_norm": 0.5188632607460022,
|
| 41749 |
+
"learning_rate": 0.0005,
|
| 41750 |
+
"loss": 1.2143,
|
| 41751 |
+
"step": 5963
|
| 41752 |
+
},
|
| 41753 |
+
{
|
| 41754 |
+
"epoch": 4.9005751848808545,
|
| 41755 |
+
"grad_norm": 0.5360555648803711,
|
| 41756 |
+
"learning_rate": 0.0005,
|
| 41757 |
+
"loss": 1.4402,
|
| 41758 |
+
"step": 5964
|
| 41759 |
+
},
|
| 41760 |
+
{
|
| 41761 |
+
"epoch": 4.901396877567789,
|
| 41762 |
+
"grad_norm": 0.5215834379196167,
|
| 41763 |
+
"learning_rate": 0.0005,
|
| 41764 |
+
"loss": 1.2509,
|
| 41765 |
+
"step": 5965
|
| 41766 |
+
},
|
| 41767 |
+
{
|
| 41768 |
+
"epoch": 4.902218570254725,
|
| 41769 |
+
"grad_norm": 0.526918888092041,
|
| 41770 |
+
"learning_rate": 0.0005,
|
| 41771 |
+
"loss": 1.2802,
|
| 41772 |
+
"step": 5966
|
| 41773 |
+
},
|
| 41774 |
+
{
|
| 41775 |
+
"epoch": 4.90304026294166,
|
| 41776 |
+
"grad_norm": 0.5387428402900696,
|
| 41777 |
+
"learning_rate": 0.0005,
|
| 41778 |
+
"loss": 1.294,
|
| 41779 |
+
"step": 5967
|
| 41780 |
+
},
|
| 41781 |
+
{
|
| 41782 |
+
"epoch": 4.903861955628595,
|
| 41783 |
+
"grad_norm": 0.5555245280265808,
|
| 41784 |
+
"learning_rate": 0.0005,
|
| 41785 |
+
"loss": 1.4213,
|
| 41786 |
+
"step": 5968
|
| 41787 |
+
},
|
| 41788 |
+
{
|
| 41789 |
+
"epoch": 4.90468364831553,
|
| 41790 |
+
"grad_norm": 0.5404963493347168,
|
| 41791 |
+
"learning_rate": 0.0005,
|
| 41792 |
+
"loss": 1.227,
|
| 41793 |
+
"step": 5969
|
| 41794 |
+
},
|
| 41795 |
+
{
|
| 41796 |
+
"epoch": 4.905505341002465,
|
| 41797 |
+
"grad_norm": 0.5623646974563599,
|
| 41798 |
+
"learning_rate": 0.0005,
|
| 41799 |
+
"loss": 1.4319,
|
| 41800 |
+
"step": 5970
|
| 41801 |
+
},
|
| 41802 |
+
{
|
| 41803 |
+
"epoch": 4.9063270336894,
|
| 41804 |
+
"grad_norm": 0.5389553904533386,
|
| 41805 |
+
"learning_rate": 0.0005,
|
| 41806 |
+
"loss": 1.297,
|
| 41807 |
+
"step": 5971
|
| 41808 |
+
},
|
| 41809 |
+
{
|
| 41810 |
+
"epoch": 4.907148726376335,
|
| 41811 |
+
"grad_norm": 0.5477744936943054,
|
| 41812 |
+
"learning_rate": 0.0005,
|
| 41813 |
+
"loss": 1.3659,
|
| 41814 |
+
"step": 5972
|
| 41815 |
+
},
|
| 41816 |
+
{
|
| 41817 |
+
"epoch": 4.90797041906327,
|
| 41818 |
+
"grad_norm": 0.5301917791366577,
|
| 41819 |
+
"learning_rate": 0.0005,
|
| 41820 |
+
"loss": 1.2658,
|
| 41821 |
+
"step": 5973
|
| 41822 |
+
},
|
| 41823 |
+
{
|
| 41824 |
+
"epoch": 4.908792111750206,
|
| 41825 |
+
"grad_norm": 0.5291617512702942,
|
| 41826 |
+
"learning_rate": 0.0005,
|
| 41827 |
+
"loss": 1.3592,
|
| 41828 |
+
"step": 5974
|
| 41829 |
+
},
|
| 41830 |
+
{
|
| 41831 |
+
"epoch": 4.90961380443714,
|
| 41832 |
+
"grad_norm": 0.5419930815696716,
|
| 41833 |
+
"learning_rate": 0.0005,
|
| 41834 |
+
"loss": 1.3125,
|
| 41835 |
+
"step": 5975
|
| 41836 |
+
},
|
| 41837 |
+
{
|
| 41838 |
+
"epoch": 4.910435497124076,
|
| 41839 |
+
"grad_norm": 0.5221468210220337,
|
| 41840 |
+
"learning_rate": 0.0005,
|
| 41841 |
+
"loss": 1.3643,
|
| 41842 |
+
"step": 5976
|
| 41843 |
+
},
|
| 41844 |
+
{
|
| 41845 |
+
"epoch": 4.9112571898110104,
|
| 41846 |
+
"grad_norm": 0.5470991134643555,
|
| 41847 |
+
"learning_rate": 0.0005,
|
| 41848 |
+
"loss": 1.3461,
|
| 41849 |
+
"step": 5977
|
| 41850 |
+
},
|
| 41851 |
+
{
|
| 41852 |
+
"epoch": 4.912078882497946,
|
| 41853 |
+
"grad_norm": 0.5354620814323425,
|
| 41854 |
+
"learning_rate": 0.0005,
|
| 41855 |
+
"loss": 1.2681,
|
| 41856 |
+
"step": 5978
|
| 41857 |
+
},
|
| 41858 |
+
{
|
| 41859 |
+
"epoch": 4.912900575184881,
|
| 41860 |
+
"grad_norm": 0.5851417779922485,
|
| 41861 |
+
"learning_rate": 0.0005,
|
| 41862 |
+
"loss": 1.3237,
|
| 41863 |
+
"step": 5979
|
| 41864 |
+
},
|
| 41865 |
+
{
|
| 41866 |
+
"epoch": 4.913722267871816,
|
| 41867 |
+
"grad_norm": 0.5492926239967346,
|
| 41868 |
+
"learning_rate": 0.0005,
|
| 41869 |
+
"loss": 1.417,
|
| 41870 |
+
"step": 5980
|
| 41871 |
+
},
|
| 41872 |
+
{
|
| 41873 |
+
"epoch": 4.914543960558751,
|
| 41874 |
+
"grad_norm": 0.5446394681930542,
|
| 41875 |
+
"learning_rate": 0.0005,
|
| 41876 |
+
"loss": 1.275,
|
| 41877 |
+
"step": 5981
|
| 41878 |
+
},
|
| 41879 |
+
{
|
| 41880 |
+
"epoch": 4.915365653245686,
|
| 41881 |
+
"grad_norm": 0.5484851002693176,
|
| 41882 |
+
"learning_rate": 0.0005,
|
| 41883 |
+
"loss": 1.3842,
|
| 41884 |
+
"step": 5982
|
| 41885 |
+
},
|
| 41886 |
+
{
|
| 41887 |
+
"epoch": 4.916187345932621,
|
| 41888 |
+
"grad_norm": 0.5432127714157104,
|
| 41889 |
+
"learning_rate": 0.0005,
|
| 41890 |
+
"loss": 1.4247,
|
| 41891 |
+
"step": 5983
|
| 41892 |
+
},
|
| 41893 |
+
{
|
| 41894 |
+
"epoch": 4.917009038619557,
|
| 41895 |
+
"grad_norm": 0.5324352979660034,
|
| 41896 |
+
"learning_rate": 0.0005,
|
| 41897 |
+
"loss": 1.3269,
|
| 41898 |
+
"step": 5984
|
| 41899 |
+
},
|
| 41900 |
+
{
|
| 41901 |
+
"epoch": 4.917830731306491,
|
| 41902 |
+
"grad_norm": 0.5508584976196289,
|
| 41903 |
+
"learning_rate": 0.0005,
|
| 41904 |
+
"loss": 1.1933,
|
| 41905 |
+
"step": 5985
|
| 41906 |
+
},
|
| 41907 |
+
{
|
| 41908 |
+
"epoch": 4.918652423993427,
|
| 41909 |
+
"grad_norm": 0.5484975576400757,
|
| 41910 |
+
"learning_rate": 0.0005,
|
| 41911 |
+
"loss": 1.2294,
|
| 41912 |
+
"step": 5986
|
| 41913 |
+
},
|
| 41914 |
+
{
|
| 41915 |
+
"epoch": 4.9194741166803615,
|
| 41916 |
+
"grad_norm": 0.5371730923652649,
|
| 41917 |
+
"learning_rate": 0.0005,
|
| 41918 |
+
"loss": 1.293,
|
| 41919 |
+
"step": 5987
|
| 41920 |
+
},
|
| 41921 |
+
{
|
| 41922 |
+
"epoch": 4.920295809367296,
|
| 41923 |
+
"grad_norm": 0.5321599841117859,
|
| 41924 |
+
"learning_rate": 0.0005,
|
| 41925 |
+
"loss": 1.3553,
|
| 41926 |
+
"step": 5988
|
| 41927 |
+
},
|
| 41928 |
+
{
|
| 41929 |
+
"epoch": 4.921117502054232,
|
| 41930 |
+
"grad_norm": 0.5347539186477661,
|
| 41931 |
+
"learning_rate": 0.0005,
|
| 41932 |
+
"loss": 1.3904,
|
| 41933 |
+
"step": 5989
|
| 41934 |
+
},
|
| 41935 |
+
{
|
| 41936 |
+
"epoch": 4.921939194741167,
|
| 41937 |
+
"grad_norm": 0.5540315508842468,
|
| 41938 |
+
"learning_rate": 0.0005,
|
| 41939 |
+
"loss": 1.3358,
|
| 41940 |
+
"step": 5990
|
| 41941 |
+
},
|
| 41942 |
+
{
|
| 41943 |
+
"epoch": 4.922760887428102,
|
| 41944 |
+
"grad_norm": 0.543171226978302,
|
| 41945 |
+
"learning_rate": 0.0005,
|
| 41946 |
+
"loss": 1.3914,
|
| 41947 |
+
"step": 5991
|
| 41948 |
+
},
|
| 41949 |
+
{
|
| 41950 |
+
"epoch": 4.923582580115037,
|
| 41951 |
+
"grad_norm": 0.5221793055534363,
|
| 41952 |
+
"learning_rate": 0.0005,
|
| 41953 |
+
"loss": 1.3852,
|
| 41954 |
+
"step": 5992
|
| 41955 |
+
},
|
| 41956 |
+
{
|
| 41957 |
+
"epoch": 4.924404272801972,
|
| 41958 |
+
"grad_norm": 0.5648449659347534,
|
| 41959 |
+
"learning_rate": 0.0005,
|
| 41960 |
+
"loss": 1.3695,
|
| 41961 |
+
"step": 5993
|
| 41962 |
+
},
|
| 41963 |
+
{
|
| 41964 |
+
"epoch": 4.925225965488907,
|
| 41965 |
+
"grad_norm": 0.5360020995140076,
|
| 41966 |
+
"learning_rate": 0.0005,
|
| 41967 |
+
"loss": 1.296,
|
| 41968 |
+
"step": 5994
|
| 41969 |
+
},
|
| 41970 |
+
{
|
| 41971 |
+
"epoch": 4.926047658175842,
|
| 41972 |
+
"grad_norm": 0.5223022699356079,
|
| 41973 |
+
"learning_rate": 0.0005,
|
| 41974 |
+
"loss": 1.2654,
|
| 41975 |
+
"step": 5995
|
| 41976 |
+
},
|
| 41977 |
+
{
|
| 41978 |
+
"epoch": 4.926869350862777,
|
| 41979 |
+
"grad_norm": 0.5244916677474976,
|
| 41980 |
+
"learning_rate": 0.0005,
|
| 41981 |
+
"loss": 1.2287,
|
| 41982 |
+
"step": 5996
|
| 41983 |
+
},
|
| 41984 |
+
{
|
| 41985 |
+
"epoch": 4.9276910435497125,
|
| 41986 |
+
"grad_norm": 0.5709188580513,
|
| 41987 |
+
"learning_rate": 0.0005,
|
| 41988 |
+
"loss": 1.414,
|
| 41989 |
+
"step": 5997
|
| 41990 |
+
},
|
| 41991 |
+
{
|
| 41992 |
+
"epoch": 4.928512736236647,
|
| 41993 |
+
"grad_norm": 0.5473321080207825,
|
| 41994 |
+
"learning_rate": 0.0005,
|
| 41995 |
+
"loss": 1.3689,
|
| 41996 |
+
"step": 5998
|
| 41997 |
+
},
|
| 41998 |
+
{
|
| 41999 |
+
"epoch": 4.929334428923583,
|
| 42000 |
+
"grad_norm": 0.5459017157554626,
|
| 42001 |
+
"learning_rate": 0.0005,
|
| 42002 |
+
"loss": 1.2876,
|
| 42003 |
+
"step": 5999
|
| 42004 |
+
},
|
| 42005 |
+
{
|
| 42006 |
+
"epoch": 4.930156121610517,
|
| 42007 |
+
"grad_norm": 0.5736708641052246,
|
| 42008 |
+
"learning_rate": 0.0005,
|
| 42009 |
+
"loss": 1.3623,
|
| 42010 |
+
"step": 6000
|
| 42011 |
}
|
| 42012 |
],
|
| 42013 |
"logging_steps": 1,
|
|
|
|
| 42022 |
"should_evaluate": false,
|
| 42023 |
"should_log": false,
|
| 42024 |
"should_save": true,
|
| 42025 |
+
"should_training_stop": true
|
| 42026 |
},
|
| 42027 |
"attributes": {}
|
| 42028 |
}
|
| 42029 |
},
|
| 42030 |
+
"total_flos": 1.2481056037404672e+18,
|
| 42031 |
"train_batch_size": 8,
|
| 42032 |
"trial_name": null,
|
| 42033 |
"trial_params": null
|