Training in progress, step 3000, checkpoint
Browse files- last-checkpoint/global_step3000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +2 -2
- last-checkpoint/global_step3000/mp_rank_00_model_states.pt +2 -2
- last-checkpoint/latest +1 -1
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +585 -5
last-checkpoint/global_step3000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7072d40044ad4e8f11191d79b2aa90d677d09624e8e2ae612da9850b526cffca
|
| 3 |
+
size 5117197020
|
last-checkpoint/global_step3000/mp_rank_00_model_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e3a5126625cd292b4f4d5e948c09dd37887dbb4bf5ef92a93c5bc13ab40d2c37
|
| 3 |
+
size 859127504
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step3000
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 962205216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:610d4192b27e6309a44afb5d80a9c25b6f192f5989c5f8fb4e13c7b43939651e
|
| 3 |
size 962205216
|
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 sha256:5eaee0853f95d35cacfb932d41346ef50714e4ef121f4dd2abff57eddeebe889
|
| 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 sha256:536a9a8504fffb5687874aeab2eb4bb450e59d1e430c707280de4db4bc58c5a4
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "./iteboshi_temp/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -297,6 +297,586 @@
|
|
| 297 |
"eval_steps_per_second": 1.472,
|
| 298 |
"eval_wer": 97.56718528995756,
|
| 299 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 300 |
}
|
| 301 |
],
|
| 302 |
"logging_steps": 25,
|
|
@@ -316,7 +896,7 @@
|
|
| 316 |
"attributes": {}
|
| 317 |
}
|
| 318 |
},
|
| 319 |
-
"total_flos":
|
| 320 |
"train_batch_size": 4,
|
| 321 |
"trial_name": null,
|
| 322 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 87.76991984912777,
|
| 3 |
+
"best_model_checkpoint": "./iteboshi_temp/checkpoint-3000",
|
| 4 |
+
"epoch": 3.303964757709251,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 3000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 297 |
"eval_steps_per_second": 1.472,
|
| 298 |
"eval_wer": 97.56718528995756,
|
| 299 |
"step": 1000
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.1288546255506609,
|
| 303 |
+
"grad_norm": 1.21933913230896,
|
| 304 |
+
"learning_rate": 1.9461538461538462e-05,
|
| 305 |
+
"loss": 1.0609,
|
| 306 |
+
"step": 1025
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 1.1563876651982379,
|
| 310 |
+
"grad_norm": 1.4192328453063965,
|
| 311 |
+
"learning_rate": 1.943589743589744e-05,
|
| 312 |
+
"loss": 1.0849,
|
| 313 |
+
"step": 1050
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.183920704845815,
|
| 317 |
+
"grad_norm": 1.730343222618103,
|
| 318 |
+
"learning_rate": 1.9410256410256413e-05,
|
| 319 |
+
"loss": 1.0461,
|
| 320 |
+
"step": 1075
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.2114537444933922,
|
| 324 |
+
"grad_norm": 1.7515013217926025,
|
| 325 |
+
"learning_rate": 1.9384615384615386e-05,
|
| 326 |
+
"loss": 0.9945,
|
| 327 |
+
"step": 1100
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.2389867841409692,
|
| 331 |
+
"grad_norm": 2.047463893890381,
|
| 332 |
+
"learning_rate": 1.935897435897436e-05,
|
| 333 |
+
"loss": 1.0707,
|
| 334 |
+
"step": 1125
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 1.2665198237885462,
|
| 338 |
+
"grad_norm": 1.3279300928115845,
|
| 339 |
+
"learning_rate": 1.9333333333333333e-05,
|
| 340 |
+
"loss": 0.985,
|
| 341 |
+
"step": 1150
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 1.2940528634361232,
|
| 345 |
+
"grad_norm": 1.2352490425109863,
|
| 346 |
+
"learning_rate": 1.930769230769231e-05,
|
| 347 |
+
"loss": 0.9729,
|
| 348 |
+
"step": 1175
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 1.3215859030837005,
|
| 352 |
+
"grad_norm": 1.4669734239578247,
|
| 353 |
+
"learning_rate": 1.9282051282051284e-05,
|
| 354 |
+
"loss": 0.9921,
|
| 355 |
+
"step": 1200
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 1.3491189427312775,
|
| 359 |
+
"grad_norm": 1.233565092086792,
|
| 360 |
+
"learning_rate": 1.9256410256410258e-05,
|
| 361 |
+
"loss": 0.9546,
|
| 362 |
+
"step": 1225
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.3766519823788546,
|
| 366 |
+
"grad_norm": 1.4740595817565918,
|
| 367 |
+
"learning_rate": 1.923076923076923e-05,
|
| 368 |
+
"loss": 0.9798,
|
| 369 |
+
"step": 1250
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.4041850220264318,
|
| 373 |
+
"grad_norm": 1.0927783250808716,
|
| 374 |
+
"learning_rate": 1.920512820512821e-05,
|
| 375 |
+
"loss": 0.9465,
|
| 376 |
+
"step": 1275
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 1.4317180616740088,
|
| 380 |
+
"grad_norm": 1.5051064491271973,
|
| 381 |
+
"learning_rate": 1.9179487179487182e-05,
|
| 382 |
+
"loss": 0.9202,
|
| 383 |
+
"step": 1300
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.4592511013215859,
|
| 387 |
+
"grad_norm": 1.218206524848938,
|
| 388 |
+
"learning_rate": 1.9153846153846156e-05,
|
| 389 |
+
"loss": 0.9766,
|
| 390 |
+
"step": 1325
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.4867841409691631,
|
| 394 |
+
"grad_norm": 1.3299014568328857,
|
| 395 |
+
"learning_rate": 1.912820512820513e-05,
|
| 396 |
+
"loss": 0.9298,
|
| 397 |
+
"step": 1350
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.51431718061674,
|
| 401 |
+
"grad_norm": 2.0308456420898438,
|
| 402 |
+
"learning_rate": 1.9102564102564106e-05,
|
| 403 |
+
"loss": 0.9287,
|
| 404 |
+
"step": 1375
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.5418502202643172,
|
| 408 |
+
"grad_norm": 1.0478333234786987,
|
| 409 |
+
"learning_rate": 1.907692307692308e-05,
|
| 410 |
+
"loss": 0.9302,
|
| 411 |
+
"step": 1400
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.5693832599118944,
|
| 415 |
+
"grad_norm": 1.2070943117141724,
|
| 416 |
+
"learning_rate": 1.905128205128205e-05,
|
| 417 |
+
"loss": 0.9162,
|
| 418 |
+
"step": 1425
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 1.5969162995594712,
|
| 422 |
+
"grad_norm": 1.317423701286316,
|
| 423 |
+
"learning_rate": 1.9025641025641027e-05,
|
| 424 |
+
"loss": 0.9029,
|
| 425 |
+
"step": 1450
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.6244493392070485,
|
| 429 |
+
"grad_norm": 1.3458503484725952,
|
| 430 |
+
"learning_rate": 1.9e-05,
|
| 431 |
+
"loss": 0.8649,
|
| 432 |
+
"step": 1475
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.6519823788546255,
|
| 436 |
+
"grad_norm": 1.8415597677230835,
|
| 437 |
+
"learning_rate": 1.8974358974358975e-05,
|
| 438 |
+
"loss": 0.8413,
|
| 439 |
+
"step": 1500
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 1.6795154185022025,
|
| 443 |
+
"grad_norm": 1.2922658920288086,
|
| 444 |
+
"learning_rate": 1.894871794871795e-05,
|
| 445 |
+
"loss": 0.8506,
|
| 446 |
+
"step": 1525
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.7070484581497798,
|
| 450 |
+
"grad_norm": 1.3912965059280396,
|
| 451 |
+
"learning_rate": 1.8923076923076925e-05,
|
| 452 |
+
"loss": 0.8408,
|
| 453 |
+
"step": 1550
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.7345814977973568,
|
| 457 |
+
"grad_norm": 1.5371092557907104,
|
| 458 |
+
"learning_rate": 1.88974358974359e-05,
|
| 459 |
+
"loss": 0.8579,
|
| 460 |
+
"step": 1575
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 1.7621145374449338,
|
| 464 |
+
"grad_norm": 1.188888669013977,
|
| 465 |
+
"learning_rate": 1.8871794871794873e-05,
|
| 466 |
+
"loss": 0.8251,
|
| 467 |
+
"step": 1600
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 1.789647577092511,
|
| 471 |
+
"grad_norm": 1.2093167304992676,
|
| 472 |
+
"learning_rate": 1.8846153846153846e-05,
|
| 473 |
+
"loss": 0.8787,
|
| 474 |
+
"step": 1625
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 1.8171806167400881,
|
| 478 |
+
"grad_norm": 1.3911653757095337,
|
| 479 |
+
"learning_rate": 1.8820512820512823e-05,
|
| 480 |
+
"loss": 0.8652,
|
| 481 |
+
"step": 1650
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 1.8447136563876652,
|
| 485 |
+
"grad_norm": 1.707056999206543,
|
| 486 |
+
"learning_rate": 1.8794871794871797e-05,
|
| 487 |
+
"loss": 0.8693,
|
| 488 |
+
"step": 1675
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 1.8722466960352424,
|
| 492 |
+
"grad_norm": 1.1974895000457764,
|
| 493 |
+
"learning_rate": 1.876923076923077e-05,
|
| 494 |
+
"loss": 0.8092,
|
| 495 |
+
"step": 1700
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.8997797356828194,
|
| 499 |
+
"grad_norm": 1.9799768924713135,
|
| 500 |
+
"learning_rate": 1.8743589743589744e-05,
|
| 501 |
+
"loss": 0.8103,
|
| 502 |
+
"step": 1725
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.9273127753303965,
|
| 506 |
+
"grad_norm": 1.5621815919876099,
|
| 507 |
+
"learning_rate": 1.8717948717948718e-05,
|
| 508 |
+
"loss": 0.8212,
|
| 509 |
+
"step": 1750
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 1.9548458149779737,
|
| 513 |
+
"grad_norm": 1.3315322399139404,
|
| 514 |
+
"learning_rate": 1.8692307692307695e-05,
|
| 515 |
+
"loss": 0.813,
|
| 516 |
+
"step": 1775
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 1.9823788546255505,
|
| 520 |
+
"grad_norm": 0.9858968257904053,
|
| 521 |
+
"learning_rate": 1.866666666666667e-05,
|
| 522 |
+
"loss": 0.7959,
|
| 523 |
+
"step": 1800
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 2.0099118942731278,
|
| 527 |
+
"grad_norm": 1.0958722829818726,
|
| 528 |
+
"learning_rate": 1.8641025641025642e-05,
|
| 529 |
+
"loss": 0.7545,
|
| 530 |
+
"step": 1825
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.037444933920705,
|
| 534 |
+
"grad_norm": 1.0353975296020508,
|
| 535 |
+
"learning_rate": 1.8615384615384616e-05,
|
| 536 |
+
"loss": 0.5917,
|
| 537 |
+
"step": 1850
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.064977973568282,
|
| 541 |
+
"grad_norm": 2.9925360679626465,
|
| 542 |
+
"learning_rate": 1.8589743589743593e-05,
|
| 543 |
+
"loss": 0.5989,
|
| 544 |
+
"step": 1875
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 2.092511013215859,
|
| 548 |
+
"grad_norm": 1.3685253858566284,
|
| 549 |
+
"learning_rate": 1.8564102564102567e-05,
|
| 550 |
+
"loss": 0.605,
|
| 551 |
+
"step": 1900
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 2.1200440528634363,
|
| 555 |
+
"grad_norm": 1.0744121074676514,
|
| 556 |
+
"learning_rate": 1.853846153846154e-05,
|
| 557 |
+
"loss": 0.6191,
|
| 558 |
+
"step": 1925
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 2.147577092511013,
|
| 562 |
+
"grad_norm": 0.898098349571228,
|
| 563 |
+
"learning_rate": 1.8512820512820514e-05,
|
| 564 |
+
"loss": 0.5486,
|
| 565 |
+
"step": 1950
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 2.1751101321585904,
|
| 569 |
+
"grad_norm": 1.2373496294021606,
|
| 570 |
+
"learning_rate": 1.848717948717949e-05,
|
| 571 |
+
"loss": 0.5693,
|
| 572 |
+
"step": 1975
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.202643171806167,
|
| 576 |
+
"grad_norm": 1.3621195554733276,
|
| 577 |
+
"learning_rate": 1.8461538461538465e-05,
|
| 578 |
+
"loss": 0.5859,
|
| 579 |
+
"step": 2000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.202643171806167,
|
| 583 |
+
"eval_cer": 48.10966033496498,
|
| 584 |
+
"eval_loss": 0.8996243476867676,
|
| 585 |
+
"eval_runtime": 1860.7163,
|
| 586 |
+
"eval_samples_per_second": 5.687,
|
| 587 |
+
"eval_steps_per_second": 1.422,
|
| 588 |
+
"eval_wer": 90.94766619519095,
|
| 589 |
+
"step": 2000
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 2.2301762114537445,
|
| 593 |
+
"grad_norm": 0.9919908046722412,
|
| 594 |
+
"learning_rate": 1.8435897435897435e-05,
|
| 595 |
+
"loss": 0.5958,
|
| 596 |
+
"step": 2025
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.2577092511013217,
|
| 600 |
+
"grad_norm": 1.0579532384872437,
|
| 601 |
+
"learning_rate": 1.8410256410256412e-05,
|
| 602 |
+
"loss": 0.5828,
|
| 603 |
+
"step": 2050
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.2852422907488985,
|
| 607 |
+
"grad_norm": 0.8056641221046448,
|
| 608 |
+
"learning_rate": 1.8384615384615386e-05,
|
| 609 |
+
"loss": 0.5666,
|
| 610 |
+
"step": 2075
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 2.3127753303964758,
|
| 614 |
+
"grad_norm": 1.1343415975570679,
|
| 615 |
+
"learning_rate": 1.835897435897436e-05,
|
| 616 |
+
"loss": 0.577,
|
| 617 |
+
"step": 2100
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 2.340308370044053,
|
| 621 |
+
"grad_norm": 1.04411780834198,
|
| 622 |
+
"learning_rate": 1.8333333333333333e-05,
|
| 623 |
+
"loss": 0.5307,
|
| 624 |
+
"step": 2125
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.36784140969163,
|
| 628 |
+
"grad_norm": 1.0452271699905396,
|
| 629 |
+
"learning_rate": 1.830769230769231e-05,
|
| 630 |
+
"loss": 0.5545,
|
| 631 |
+
"step": 2150
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 2.395374449339207,
|
| 635 |
+
"grad_norm": 0.927592396736145,
|
| 636 |
+
"learning_rate": 1.8282051282051284e-05,
|
| 637 |
+
"loss": 0.5551,
|
| 638 |
+
"step": 2175
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.4229074889867843,
|
| 642 |
+
"grad_norm": 1.7057969570159912,
|
| 643 |
+
"learning_rate": 1.8256410256410257e-05,
|
| 644 |
+
"loss": 0.5326,
|
| 645 |
+
"step": 2200
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.450440528634361,
|
| 649 |
+
"grad_norm": 1.4575523138046265,
|
| 650 |
+
"learning_rate": 1.823076923076923e-05,
|
| 651 |
+
"loss": 0.5308,
|
| 652 |
+
"step": 2225
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 2.4779735682819384,
|
| 656 |
+
"grad_norm": 1.0914602279663086,
|
| 657 |
+
"learning_rate": 1.8205128205128208e-05,
|
| 658 |
+
"loss": 0.5547,
|
| 659 |
+
"step": 2250
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 2.505506607929515,
|
| 663 |
+
"grad_norm": 0.990104079246521,
|
| 664 |
+
"learning_rate": 1.817948717948718e-05,
|
| 665 |
+
"loss": 0.6072,
|
| 666 |
+
"step": 2275
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 2.5330396475770924,
|
| 670 |
+
"grad_norm": 0.8314220905303955,
|
| 671 |
+
"learning_rate": 1.8153846153846155e-05,
|
| 672 |
+
"loss": 0.5671,
|
| 673 |
+
"step": 2300
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 2.5605726872246697,
|
| 677 |
+
"grad_norm": 0.9760991334915161,
|
| 678 |
+
"learning_rate": 1.812820512820513e-05,
|
| 679 |
+
"loss": 0.5421,
|
| 680 |
+
"step": 2325
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.5881057268722465,
|
| 684 |
+
"grad_norm": 1.0801244974136353,
|
| 685 |
+
"learning_rate": 1.8102564102564102e-05,
|
| 686 |
+
"loss": 0.5558,
|
| 687 |
+
"step": 2350
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 2.6156387665198237,
|
| 691 |
+
"grad_norm": 0.8499842286109924,
|
| 692 |
+
"learning_rate": 1.807692307692308e-05,
|
| 693 |
+
"loss": 0.56,
|
| 694 |
+
"step": 2375
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 2.643171806167401,
|
| 698 |
+
"grad_norm": 1.1915971040725708,
|
| 699 |
+
"learning_rate": 1.8051282051282053e-05,
|
| 700 |
+
"loss": 0.5198,
|
| 701 |
+
"step": 2400
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 2.670704845814978,
|
| 705 |
+
"grad_norm": 1.093216061592102,
|
| 706 |
+
"learning_rate": 1.8025641025641027e-05,
|
| 707 |
+
"loss": 0.5156,
|
| 708 |
+
"step": 2425
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 2.698237885462555,
|
| 712 |
+
"grad_norm": 1.1357547044754028,
|
| 713 |
+
"learning_rate": 1.8e-05,
|
| 714 |
+
"loss": 0.5755,
|
| 715 |
+
"step": 2450
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 2.7257709251101323,
|
| 719 |
+
"grad_norm": 0.963991641998291,
|
| 720 |
+
"learning_rate": 1.7974358974358977e-05,
|
| 721 |
+
"loss": 0.5219,
|
| 722 |
+
"step": 2475
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 2.753303964757709,
|
| 726 |
+
"grad_norm": 2.196319341659546,
|
| 727 |
+
"learning_rate": 1.794871794871795e-05,
|
| 728 |
+
"loss": 0.5858,
|
| 729 |
+
"step": 2500
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 2.7808370044052864,
|
| 733 |
+
"grad_norm": 1.075908899307251,
|
| 734 |
+
"learning_rate": 1.7923076923076925e-05,
|
| 735 |
+
"loss": 0.5284,
|
| 736 |
+
"step": 2525
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 2.8083700440528636,
|
| 740 |
+
"grad_norm": 1.052140712738037,
|
| 741 |
+
"learning_rate": 1.78974358974359e-05,
|
| 742 |
+
"loss": 0.4964,
|
| 743 |
+
"step": 2550
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 2.8359030837004404,
|
| 747 |
+
"grad_norm": 0.9454672336578369,
|
| 748 |
+
"learning_rate": 1.7871794871794875e-05,
|
| 749 |
+
"loss": 0.5225,
|
| 750 |
+
"step": 2575
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 2.8634361233480177,
|
| 754 |
+
"grad_norm": 0.8262547850608826,
|
| 755 |
+
"learning_rate": 1.784615384615385e-05,
|
| 756 |
+
"loss": 0.5573,
|
| 757 |
+
"step": 2600
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 2.890969162995595,
|
| 761 |
+
"grad_norm": 1.0611587762832642,
|
| 762 |
+
"learning_rate": 1.7820512820512823e-05,
|
| 763 |
+
"loss": 0.5183,
|
| 764 |
+
"step": 2625
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 2.9185022026431717,
|
| 768 |
+
"grad_norm": 0.7847844958305359,
|
| 769 |
+
"learning_rate": 1.7794871794871796e-05,
|
| 770 |
+
"loss": 0.5333,
|
| 771 |
+
"step": 2650
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 2.946035242290749,
|
| 775 |
+
"grad_norm": 0.746285617351532,
|
| 776 |
+
"learning_rate": 1.776923076923077e-05,
|
| 777 |
+
"loss": 0.5264,
|
| 778 |
+
"step": 2675
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 2.9735682819383262,
|
| 782 |
+
"grad_norm": 1.381616234779358,
|
| 783 |
+
"learning_rate": 1.7743589743589744e-05,
|
| 784 |
+
"loss": 0.5074,
|
| 785 |
+
"step": 2700
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.001101321585903,
|
| 789 |
+
"grad_norm": 0.6723135113716125,
|
| 790 |
+
"learning_rate": 1.7717948717948717e-05,
|
| 791 |
+
"loss": 0.5631,
|
| 792 |
+
"step": 2725
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.0286343612334803,
|
| 796 |
+
"grad_norm": 0.9439449906349182,
|
| 797 |
+
"learning_rate": 1.7692307692307694e-05,
|
| 798 |
+
"loss": 0.3836,
|
| 799 |
+
"step": 2750
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.056167400881057,
|
| 803 |
+
"grad_norm": 0.9093062281608582,
|
| 804 |
+
"learning_rate": 1.7666666666666668e-05,
|
| 805 |
+
"loss": 0.342,
|
| 806 |
+
"step": 2775
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.0837004405286343,
|
| 810 |
+
"grad_norm": 0.7883495092391968,
|
| 811 |
+
"learning_rate": 1.7641025641025642e-05,
|
| 812 |
+
"loss": 0.3678,
|
| 813 |
+
"step": 2800
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.1112334801762116,
|
| 817 |
+
"grad_norm": 0.5074595808982849,
|
| 818 |
+
"learning_rate": 1.7615384615384615e-05,
|
| 819 |
+
"loss": 0.3406,
|
| 820 |
+
"step": 2825
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.1387665198237884,
|
| 824 |
+
"grad_norm": 1.330426812171936,
|
| 825 |
+
"learning_rate": 1.7589743589743592e-05,
|
| 826 |
+
"loss": 0.3432,
|
| 827 |
+
"step": 2850
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.1662995594713657,
|
| 831 |
+
"grad_norm": 1.008254051208496,
|
| 832 |
+
"learning_rate": 1.7564102564102566e-05,
|
| 833 |
+
"loss": 0.3722,
|
| 834 |
+
"step": 2875
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.193832599118943,
|
| 838 |
+
"grad_norm": 1.0520501136779785,
|
| 839 |
+
"learning_rate": 1.753846153846154e-05,
|
| 840 |
+
"loss": 0.3719,
|
| 841 |
+
"step": 2900
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.2213656387665197,
|
| 845 |
+
"grad_norm": 0.7822287082672119,
|
| 846 |
+
"learning_rate": 1.7512820512820513e-05,
|
| 847 |
+
"loss": 0.3729,
|
| 848 |
+
"step": 2925
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.248898678414097,
|
| 852 |
+
"grad_norm": 1.1690279245376587,
|
| 853 |
+
"learning_rate": 1.7487179487179487e-05,
|
| 854 |
+
"loss": 0.3723,
|
| 855 |
+
"step": 2950
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.2764317180616738,
|
| 859 |
+
"grad_norm": 0.8030567765235901,
|
| 860 |
+
"learning_rate": 1.7461538461538464e-05,
|
| 861 |
+
"loss": 0.362,
|
| 862 |
+
"step": 2975
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.303964757709251,
|
| 866 |
+
"grad_norm": 0.7881470918655396,
|
| 867 |
+
"learning_rate": 1.7435897435897438e-05,
|
| 868 |
+
"loss": 0.3373,
|
| 869 |
+
"step": 3000
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.303964757709251,
|
| 873 |
+
"eval_cer": 29.99495603727947,
|
| 874 |
+
"eval_loss": 0.7765971422195435,
|
| 875 |
+
"eval_runtime": 1749.0783,
|
| 876 |
+
"eval_samples_per_second": 6.049,
|
| 877 |
+
"eval_steps_per_second": 1.513,
|
| 878 |
+
"eval_wer": 87.76991984912777,
|
| 879 |
+
"step": 3000
|
| 880 |
}
|
| 881 |
],
|
| 882 |
"logging_steps": 25,
|
|
|
|
| 896 |
"attributes": {}
|
| 897 |
}
|
| 898 |
},
|
| 899 |
+
"total_flos": 5.154429559470293e+19,
|
| 900 |
"train_batch_size": 4,
|
| 901 |
"trial_name": null,
|
| 902 |
"trial_params": null
|