Training in progress, step 1000, 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 35237104
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6595bf55f02395c5cbd4666c9cb95f98302cd9ef023ebeb628cdc4cf4bf4caee
|
3 |
size 35237104
|
last-checkpoint/optimizer.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 18810356
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c4e6e0de1bcf5bda93ea3307cefce39a83cb8a6899981f858868c87622dba1fd
|
3 |
size 18810356
|
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:a7b6fd1c9514da6d7da36a4b4159526160fde1963c15f9d5ba39d98a761f42c0
|
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 sha256:4077036d99500a708f700f75da24d51b5300e184ad35fda49dc5a4df5596cca2
|
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": 250,
|
6 |
-
"global_step":
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
@@ -6339,6 +6339,714 @@
|
|
6339 |
"learning_rate": 2.7091379149682685e-06,
|
6340 |
"loss": 2.5194,
|
6341 |
"step": 900
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6342 |
}
|
6343 |
],
|
6344 |
"logging_steps": 1,
|
@@ -6353,12 +7061,12 @@
|
|
6353 |
"should_evaluate": false,
|
6354 |
"should_log": false,
|
6355 |
"should_save": true,
|
6356 |
-
"should_training_stop":
|
6357 |
},
|
6358 |
"attributes": {}
|
6359 |
}
|
6360 |
},
|
6361 |
-
"total_flos":
|
6362 |
"train_batch_size": 2,
|
6363 |
"trial_name": null,
|
6364 |
"trial_params": null
|
|
|
1 |
{
|
2 |
"best_metric": null,
|
3 |
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.3624173235480656,
|
5 |
"eval_steps": 250,
|
6 |
+
"global_step": 1000,
|
7 |
"is_hyper_param_search": false,
|
8 |
"is_local_process_zero": true,
|
9 |
"is_world_process_zero": true,
|
|
|
6339 |
"learning_rate": 2.7091379149682685e-06,
|
6340 |
"loss": 2.5194,
|
6341 |
"step": 900
|
6342 |
+
},
|
6343 |
+
{
|
6344 |
+
"epoch": 0.3265380085168071,
|
6345 |
+
"grad_norm": 6.8190178871154785,
|
6346 |
+
"learning_rate": 2.6557085182532582e-06,
|
6347 |
+
"loss": 2.4827,
|
6348 |
+
"step": 901
|
6349 |
+
},
|
6350 |
+
{
|
6351 |
+
"epoch": 0.32690042584035517,
|
6352 |
+
"grad_norm": 8.026456832885742,
|
6353 |
+
"learning_rate": 2.602796871124663e-06,
|
6354 |
+
"loss": 3.6636,
|
6355 |
+
"step": 902
|
6356 |
+
},
|
6357 |
+
{
|
6358 |
+
"epoch": 0.32726284316390325,
|
6359 |
+
"grad_norm": 6.956340789794922,
|
6360 |
+
"learning_rate": 2.5504035522157854e-06,
|
6361 |
+
"loss": 2.8476,
|
6362 |
+
"step": 903
|
6363 |
+
},
|
6364 |
+
{
|
6365 |
+
"epoch": 0.3276252604874513,
|
6366 |
+
"grad_norm": 5.77375602722168,
|
6367 |
+
"learning_rate": 2.4985291344915674e-06,
|
6368 |
+
"loss": 2.6426,
|
6369 |
+
"step": 904
|
6370 |
+
},
|
6371 |
+
{
|
6372 |
+
"epoch": 0.32798767781099936,
|
6373 |
+
"grad_norm": 6.200889587402344,
|
6374 |
+
"learning_rate": 2.4471741852423237e-06,
|
6375 |
+
"loss": 3.4329,
|
6376 |
+
"step": 905
|
6377 |
+
},
|
6378 |
+
{
|
6379 |
+
"epoch": 0.32835009513454744,
|
6380 |
+
"grad_norm": 6.780580043792725,
|
6381 |
+
"learning_rate": 2.3963392660775575e-06,
|
6382 |
+
"loss": 3.2439,
|
6383 |
+
"step": 906
|
6384 |
+
},
|
6385 |
+
{
|
6386 |
+
"epoch": 0.3287125124580955,
|
6387 |
+
"grad_norm": 6.5239644050598145,
|
6388 |
+
"learning_rate": 2.3460249329197824e-06,
|
6389 |
+
"loss": 2.5931,
|
6390 |
+
"step": 907
|
6391 |
+
},
|
6392 |
+
{
|
6393 |
+
"epoch": 0.32907492978164354,
|
6394 |
+
"grad_norm": 7.130517482757568,
|
6395 |
+
"learning_rate": 2.296231735998511e-06,
|
6396 |
+
"loss": 2.6862,
|
6397 |
+
"step": 908
|
6398 |
+
},
|
6399 |
+
{
|
6400 |
+
"epoch": 0.3294373471051916,
|
6401 |
+
"grad_norm": 6.246675968170166,
|
6402 |
+
"learning_rate": 2.2469602198441573e-06,
|
6403 |
+
"loss": 2.8948,
|
6404 |
+
"step": 909
|
6405 |
+
},
|
6406 |
+
{
|
6407 |
+
"epoch": 0.3297997644287397,
|
6408 |
+
"grad_norm": 7.630551338195801,
|
6409 |
+
"learning_rate": 2.1982109232821178e-06,
|
6410 |
+
"loss": 2.615,
|
6411 |
+
"step": 910
|
6412 |
+
},
|
6413 |
+
{
|
6414 |
+
"epoch": 0.3301621817522878,
|
6415 |
+
"grad_norm": 4.334846019744873,
|
6416 |
+
"learning_rate": 2.149984379426906e-06,
|
6417 |
+
"loss": 2.2025,
|
6418 |
+
"step": 911
|
6419 |
+
},
|
6420 |
+
{
|
6421 |
+
"epoch": 0.3305245990758358,
|
6422 |
+
"grad_norm": 6.460603713989258,
|
6423 |
+
"learning_rate": 2.102281115676258e-06,
|
6424 |
+
"loss": 2.6511,
|
6425 |
+
"step": 912
|
6426 |
+
},
|
6427 |
+
{
|
6428 |
+
"epoch": 0.3308870163993839,
|
6429 |
+
"grad_norm": 7.5470991134643555,
|
6430 |
+
"learning_rate": 2.0551016537054493e-06,
|
6431 |
+
"loss": 2.6114,
|
6432 |
+
"step": 913
|
6433 |
+
},
|
6434 |
+
{
|
6435 |
+
"epoch": 0.331249433722932,
|
6436 |
+
"grad_norm": 6.667960166931152,
|
6437 |
+
"learning_rate": 2.008446509461498e-06,
|
6438 |
+
"loss": 2.8898,
|
6439 |
+
"step": 914
|
6440 |
+
},
|
6441 |
+
{
|
6442 |
+
"epoch": 0.33161185104648,
|
6443 |
+
"grad_norm": 7.1564836502075195,
|
6444 |
+
"learning_rate": 1.962316193157593e-06,
|
6445 |
+
"loss": 2.7416,
|
6446 |
+
"step": 915
|
6447 |
+
},
|
6448 |
+
{
|
6449 |
+
"epoch": 0.3319742683700281,
|
6450 |
+
"grad_norm": 6.149023056030273,
|
6451 |
+
"learning_rate": 1.91671120926748e-06,
|
6452 |
+
"loss": 2.5522,
|
6453 |
+
"step": 916
|
6454 |
+
},
|
6455 |
+
{
|
6456 |
+
"epoch": 0.33233668569357616,
|
6457 |
+
"grad_norm": 8.589316368103027,
|
6458 |
+
"learning_rate": 1.8716320565199618e-06,
|
6459 |
+
"loss": 1.8536,
|
6460 |
+
"step": 917
|
6461 |
+
},
|
6462 |
+
{
|
6463 |
+
"epoch": 0.33269910301712424,
|
6464 |
+
"grad_norm": 5.793923377990723,
|
6465 |
+
"learning_rate": 1.8270792278934302e-06,
|
6466 |
+
"loss": 2.9145,
|
6467 |
+
"step": 918
|
6468 |
+
},
|
6469 |
+
{
|
6470 |
+
"epoch": 0.33306152034067227,
|
6471 |
+
"grad_norm": 6.867883682250977,
|
6472 |
+
"learning_rate": 1.7830532106104747e-06,
|
6473 |
+
"loss": 3.1564,
|
6474 |
+
"step": 919
|
6475 |
+
},
|
6476 |
+
{
|
6477 |
+
"epoch": 0.33342393766422035,
|
6478 |
+
"grad_norm": 5.381682872772217,
|
6479 |
+
"learning_rate": 1.7395544861325718e-06,
|
6480 |
+
"loss": 2.5914,
|
6481 |
+
"step": 920
|
6482 |
+
},
|
6483 |
+
{
|
6484 |
+
"epoch": 0.33378635498776843,
|
6485 |
+
"grad_norm": 6.665421009063721,
|
6486 |
+
"learning_rate": 1.696583530154794e-06,
|
6487 |
+
"loss": 2.2271,
|
6488 |
+
"step": 921
|
6489 |
+
},
|
6490 |
+
{
|
6491 |
+
"epoch": 0.33414877231131646,
|
6492 |
+
"grad_norm": 9.370551109313965,
|
6493 |
+
"learning_rate": 1.6541408126006463e-06,
|
6494 |
+
"loss": 2.8091,
|
6495 |
+
"step": 922
|
6496 |
+
},
|
6497 |
+
{
|
6498 |
+
"epoch": 0.33451118963486454,
|
6499 |
+
"grad_norm": 7.959995746612549,
|
6500 |
+
"learning_rate": 1.6122267976168781e-06,
|
6501 |
+
"loss": 3.6784,
|
6502 |
+
"step": 923
|
6503 |
+
},
|
6504 |
+
{
|
6505 |
+
"epoch": 0.3348736069584126,
|
6506 |
+
"grad_norm": 7.0754475593566895,
|
6507 |
+
"learning_rate": 1.5708419435684462e-06,
|
6508 |
+
"loss": 2.9994,
|
6509 |
+
"step": 924
|
6510 |
+
},
|
6511 |
+
{
|
6512 |
+
"epoch": 0.3352360242819607,
|
6513 |
+
"grad_norm": 6.9350361824035645,
|
6514 |
+
"learning_rate": 1.5299867030334814e-06,
|
6515 |
+
"loss": 1.8455,
|
6516 |
+
"step": 925
|
6517 |
+
},
|
6518 |
+
{
|
6519 |
+
"epoch": 0.3355984416055087,
|
6520 |
+
"grad_norm": 6.33629846572876,
|
6521 |
+
"learning_rate": 1.4896615227983468e-06,
|
6522 |
+
"loss": 2.2597,
|
6523 |
+
"step": 926
|
6524 |
+
},
|
6525 |
+
{
|
6526 |
+
"epoch": 0.3359608589290568,
|
6527 |
+
"grad_norm": 6.027073383331299,
|
6528 |
+
"learning_rate": 1.4498668438527597e-06,
|
6529 |
+
"loss": 2.548,
|
6530 |
+
"step": 927
|
6531 |
+
},
|
6532 |
+
{
|
6533 |
+
"epoch": 0.3363232762526049,
|
6534 |
+
"grad_norm": 6.1353044509887695,
|
6535 |
+
"learning_rate": 1.4106031013849496e-06,
|
6536 |
+
"loss": 2.6825,
|
6537 |
+
"step": 928
|
6538 |
+
},
|
6539 |
+
{
|
6540 |
+
"epoch": 0.33668569357615297,
|
6541 |
+
"grad_norm": 4.776915550231934,
|
6542 |
+
"learning_rate": 1.3718707247769135e-06,
|
6543 |
+
"loss": 3.3943,
|
6544 |
+
"step": 929
|
6545 |
+
},
|
6546 |
+
{
|
6547 |
+
"epoch": 0.337048110899701,
|
6548 |
+
"grad_norm": 7.03324031829834,
|
6549 |
+
"learning_rate": 1.333670137599713e-06,
|
6550 |
+
"loss": 2.7431,
|
6551 |
+
"step": 930
|
6552 |
+
},
|
6553 |
+
{
|
6554 |
+
"epoch": 0.3374105282232491,
|
6555 |
+
"grad_norm": 7.59218692779541,
|
6556 |
+
"learning_rate": 1.2960017576088446e-06,
|
6557 |
+
"loss": 2.9679,
|
6558 |
+
"step": 931
|
6559 |
+
},
|
6560 |
+
{
|
6561 |
+
"epoch": 0.33777294554679715,
|
6562 |
+
"grad_norm": 6.613968372344971,
|
6563 |
+
"learning_rate": 1.2588659967397e-06,
|
6564 |
+
"loss": 2.2011,
|
6565 |
+
"step": 932
|
6566 |
+
},
|
6567 |
+
{
|
6568 |
+
"epoch": 0.3381353628703452,
|
6569 |
+
"grad_norm": 5.920099258422852,
|
6570 |
+
"learning_rate": 1.222263261102985e-06,
|
6571 |
+
"loss": 2.3399,
|
6572 |
+
"step": 933
|
6573 |
+
},
|
6574 |
+
{
|
6575 |
+
"epoch": 0.33849778019389326,
|
6576 |
+
"grad_norm": 9.415234565734863,
|
6577 |
+
"learning_rate": 1.1861939509803687e-06,
|
6578 |
+
"loss": 2.8695,
|
6579 |
+
"step": 934
|
6580 |
+
},
|
6581 |
+
{
|
6582 |
+
"epoch": 0.33886019751744134,
|
6583 |
+
"grad_norm": 5.54183292388916,
|
6584 |
+
"learning_rate": 1.1506584608200367e-06,
|
6585 |
+
"loss": 2.8506,
|
6586 |
+
"step": 935
|
6587 |
+
},
|
6588 |
+
{
|
6589 |
+
"epoch": 0.3392226148409894,
|
6590 |
+
"grad_norm": 6.957115650177002,
|
6591 |
+
"learning_rate": 1.1156571792324211e-06,
|
6592 |
+
"loss": 2.3371,
|
6593 |
+
"step": 936
|
6594 |
+
},
|
6595 |
+
{
|
6596 |
+
"epoch": 0.33958503216453745,
|
6597 |
+
"grad_norm": 4.749189376831055,
|
6598 |
+
"learning_rate": 1.0811904889859336e-06,
|
6599 |
+
"loss": 2.3668,
|
6600 |
+
"step": 937
|
6601 |
+
},
|
6602 |
+
{
|
6603 |
+
"epoch": 0.33994744948808553,
|
6604 |
+
"grad_norm": 7.631181716918945,
|
6605 |
+
"learning_rate": 1.0472587670027678e-06,
|
6606 |
+
"loss": 3.3614,
|
6607 |
+
"step": 938
|
6608 |
+
},
|
6609 |
+
{
|
6610 |
+
"epoch": 0.3403098668116336,
|
6611 |
+
"grad_norm": 5.728475570678711,
|
6612 |
+
"learning_rate": 1.0138623843548078e-06,
|
6613 |
+
"loss": 2.0496,
|
6614 |
+
"step": 939
|
6615 |
+
},
|
6616 |
+
{
|
6617 |
+
"epoch": 0.34067228413518164,
|
6618 |
+
"grad_norm": 6.60684871673584,
|
6619 |
+
"learning_rate": 9.810017062595322e-07,
|
6620 |
+
"loss": 2.7532,
|
6621 |
+
"step": 940
|
6622 |
+
},
|
6623 |
+
{
|
6624 |
+
"epoch": 0.3410347014587297,
|
6625 |
+
"grad_norm": 6.326074123382568,
|
6626 |
+
"learning_rate": 9.486770920760668e-07,
|
6627 |
+
"loss": 2.5077,
|
6628 |
+
"step": 941
|
6629 |
+
},
|
6630 |
+
{
|
6631 |
+
"epoch": 0.3413971187822778,
|
6632 |
+
"grad_norm": 7.019238471984863,
|
6633 |
+
"learning_rate": 9.168888953011989e-07,
|
6634 |
+
"loss": 2.0309,
|
6635 |
+
"step": 942
|
6636 |
+
},
|
6637 |
+
{
|
6638 |
+
"epoch": 0.3417595361058259,
|
6639 |
+
"grad_norm": 8.484062194824219,
|
6640 |
+
"learning_rate": 8.856374635655695e-07,
|
6641 |
+
"loss": 3.3332,
|
6642 |
+
"step": 943
|
6643 |
+
},
|
6644 |
+
{
|
6645 |
+
"epoch": 0.3421219534293739,
|
6646 |
+
"grad_norm": 7.0083842277526855,
|
6647 |
+
"learning_rate": 8.549231386298151e-07,
|
6648 |
+
"loss": 2.9812,
|
6649 |
+
"step": 944
|
6650 |
+
},
|
6651 |
+
{
|
6652 |
+
"epoch": 0.342484370752922,
|
6653 |
+
"grad_norm": 6.267253875732422,
|
6654 |
+
"learning_rate": 8.247462563808817e-07,
|
6655 |
+
"loss": 3.0255,
|
6656 |
+
"step": 945
|
6657 |
+
},
|
6658 |
+
{
|
6659 |
+
"epoch": 0.34284678807647007,
|
6660 |
+
"grad_norm": 8.533939361572266,
|
6661 |
+
"learning_rate": 7.951071468283167e-07,
|
6662 |
+
"loss": 3.2914,
|
6663 |
+
"step": 946
|
6664 |
+
},
|
6665 |
+
{
|
6666 |
+
"epoch": 0.34320920540001815,
|
6667 |
+
"grad_norm": 6.0240373611450195,
|
6668 |
+
"learning_rate": 7.66006134100672e-07,
|
6669 |
+
"loss": 3.0139,
|
6670 |
+
"step": 947
|
6671 |
+
},
|
6672 |
+
{
|
6673 |
+
"epoch": 0.3435716227235662,
|
6674 |
+
"grad_norm": 6.486016750335693,
|
6675 |
+
"learning_rate": 7.374435364419674e-07,
|
6676 |
+
"loss": 2.9023,
|
6677 |
+
"step": 948
|
6678 |
+
},
|
6679 |
+
{
|
6680 |
+
"epoch": 0.34393404004711425,
|
6681 |
+
"grad_norm": 8.553296089172363,
|
6682 |
+
"learning_rate": 7.094196662081831e-07,
|
6683 |
+
"loss": 2.9581,
|
6684 |
+
"step": 949
|
6685 |
+
},
|
6686 |
+
{
|
6687 |
+
"epoch": 0.34429645737066233,
|
6688 |
+
"grad_norm": 6.792461395263672,
|
6689 |
+
"learning_rate": 6.819348298638839e-07,
|
6690 |
+
"loss": 3.0948,
|
6691 |
+
"step": 950
|
6692 |
+
},
|
6693 |
+
{
|
6694 |
+
"epoch": 0.34465887469421036,
|
6695 |
+
"grad_norm": 6.591213703155518,
|
6696 |
+
"learning_rate": 6.549893279788277e-07,
|
6697 |
+
"loss": 2.7468,
|
6698 |
+
"step": 951
|
6699 |
+
},
|
6700 |
+
{
|
6701 |
+
"epoch": 0.34502129201775844,
|
6702 |
+
"grad_norm": 7.372961044311523,
|
6703 |
+
"learning_rate": 6.285834552247128e-07,
|
6704 |
+
"loss": 2.6526,
|
6705 |
+
"step": 952
|
6706 |
+
},
|
6707 |
+
{
|
6708 |
+
"epoch": 0.3453837093413065,
|
6709 |
+
"grad_norm": 5.669467926025391,
|
6710 |
+
"learning_rate": 6.027175003719354e-07,
|
6711 |
+
"loss": 2.7276,
|
6712 |
+
"step": 953
|
6713 |
+
},
|
6714 |
+
{
|
6715 |
+
"epoch": 0.3457461266648546,
|
6716 |
+
"grad_norm": 6.340987682342529,
|
6717 |
+
"learning_rate": 5.773917462864264e-07,
|
6718 |
+
"loss": 2.5295,
|
6719 |
+
"step": 954
|
6720 |
+
},
|
6721 |
+
{
|
6722 |
+
"epoch": 0.34610854398840263,
|
6723 |
+
"grad_norm": 6.276302814483643,
|
6724 |
+
"learning_rate": 5.526064699265753e-07,
|
6725 |
+
"loss": 3.6131,
|
6726 |
+
"step": 955
|
6727 |
+
},
|
6728 |
+
{
|
6729 |
+
"epoch": 0.3464709613119507,
|
6730 |
+
"grad_norm": 5.535942077636719,
|
6731 |
+
"learning_rate": 5.283619423401998e-07,
|
6732 |
+
"loss": 2.4283,
|
6733 |
+
"step": 956
|
6734 |
+
},
|
6735 |
+
{
|
6736 |
+
"epoch": 0.3468333786354988,
|
6737 |
+
"grad_norm": 7.801699638366699,
|
6738 |
+
"learning_rate": 5.046584286615697e-07,
|
6739 |
+
"loss": 2.8357,
|
6740 |
+
"step": 957
|
6741 |
+
},
|
6742 |
+
{
|
6743 |
+
"epoch": 0.3471957959590468,
|
6744 |
+
"grad_norm": 8.31689167022705,
|
6745 |
+
"learning_rate": 4.814961881085045e-07,
|
6746 |
+
"loss": 3.3685,
|
6747 |
+
"step": 958
|
6748 |
+
},
|
6749 |
+
{
|
6750 |
+
"epoch": 0.3475582132825949,
|
6751 |
+
"grad_norm": 6.9537577629089355,
|
6752 |
+
"learning_rate": 4.5887547397955864e-07,
|
6753 |
+
"loss": 3.3186,
|
6754 |
+
"step": 959
|
6755 |
+
},
|
6756 |
+
{
|
6757 |
+
"epoch": 0.347920630606143,
|
6758 |
+
"grad_norm": 6.808418273925781,
|
6759 |
+
"learning_rate": 4.367965336512403e-07,
|
6760 |
+
"loss": 2.6928,
|
6761 |
+
"step": 960
|
6762 |
+
},
|
6763 |
+
{
|
6764 |
+
"epoch": 0.34828304792969106,
|
6765 |
+
"grad_norm": 5.8220086097717285,
|
6766 |
+
"learning_rate": 4.1525960857530243e-07,
|
6767 |
+
"loss": 2.7677,
|
6768 |
+
"step": 961
|
6769 |
+
},
|
6770 |
+
{
|
6771 |
+
"epoch": 0.3486454652532391,
|
6772 |
+
"grad_norm": 5.832417011260986,
|
6773 |
+
"learning_rate": 3.9426493427611177e-07,
|
6774 |
+
"loss": 2.819,
|
6775 |
+
"step": 962
|
6776 |
+
},
|
6777 |
+
{
|
6778 |
+
"epoch": 0.34900788257678717,
|
6779 |
+
"grad_norm": 7.183162689208984,
|
6780 |
+
"learning_rate": 3.738127403480507e-07,
|
6781 |
+
"loss": 2.7156,
|
6782 |
+
"step": 963
|
6783 |
+
},
|
6784 |
+
{
|
6785 |
+
"epoch": 0.34937029990033525,
|
6786 |
+
"grad_norm": 7.9874677658081055,
|
6787 |
+
"learning_rate": 3.5390325045304706e-07,
|
6788 |
+
"loss": 3.0892,
|
6789 |
+
"step": 964
|
6790 |
+
},
|
6791 |
+
{
|
6792 |
+
"epoch": 0.3497327172238833,
|
6793 |
+
"grad_norm": 7.457160949707031,
|
6794 |
+
"learning_rate": 3.3453668231809286e-07,
|
6795 |
+
"loss": 3.4502,
|
6796 |
+
"step": 965
|
6797 |
+
},
|
6798 |
+
{
|
6799 |
+
"epoch": 0.35009513454743135,
|
6800 |
+
"grad_norm": 7.874043941497803,
|
6801 |
+
"learning_rate": 3.157132477328628e-07,
|
6802 |
+
"loss": 2.7757,
|
6803 |
+
"step": 966
|
6804 |
+
},
|
6805 |
+
{
|
6806 |
+
"epoch": 0.35045755187097943,
|
6807 |
+
"grad_norm": 7.999273777008057,
|
6808 |
+
"learning_rate": 2.9743315254743833e-07,
|
6809 |
+
"loss": 3.0866,
|
6810 |
+
"step": 967
|
6811 |
+
},
|
6812 |
+
{
|
6813 |
+
"epoch": 0.3508199691945275,
|
6814 |
+
"grad_norm": 8.767258644104004,
|
6815 |
+
"learning_rate": 2.796965966699927e-07,
|
6816 |
+
"loss": 2.5532,
|
6817 |
+
"step": 968
|
6818 |
+
},
|
6819 |
+
{
|
6820 |
+
"epoch": 0.35118238651807554,
|
6821 |
+
"grad_norm": 6.00533390045166,
|
6822 |
+
"learning_rate": 2.625037740646763e-07,
|
6823 |
+
"loss": 2.9059,
|
6824 |
+
"step": 969
|
6825 |
+
},
|
6826 |
+
{
|
6827 |
+
"epoch": 0.3515448038416236,
|
6828 |
+
"grad_norm": 5.833471298217773,
|
6829 |
+
"learning_rate": 2.458548727494292e-07,
|
6830 |
+
"loss": 3.3388,
|
6831 |
+
"step": 970
|
6832 |
+
},
|
6833 |
+
{
|
6834 |
+
"epoch": 0.3519072211651717,
|
6835 |
+
"grad_norm": 6.837587833404541,
|
6836 |
+
"learning_rate": 2.2975007479397738e-07,
|
6837 |
+
"loss": 2.9694,
|
6838 |
+
"step": 971
|
6839 |
+
},
|
6840 |
+
{
|
6841 |
+
"epoch": 0.3522696384887198,
|
6842 |
+
"grad_norm": 7.076719760894775,
|
6843 |
+
"learning_rate": 2.1418955631781202e-07,
|
6844 |
+
"loss": 2.2279,
|
6845 |
+
"step": 972
|
6846 |
+
},
|
6847 |
+
{
|
6848 |
+
"epoch": 0.3526320558122678,
|
6849 |
+
"grad_norm": 5.842435359954834,
|
6850 |
+
"learning_rate": 1.9917348748826335e-07,
|
6851 |
+
"loss": 2.7922,
|
6852 |
+
"step": 973
|
6853 |
+
},
|
6854 |
+
{
|
6855 |
+
"epoch": 0.3529944731358159,
|
6856 |
+
"grad_norm": 8.36257266998291,
|
6857 |
+
"learning_rate": 1.847020325186577e-07,
|
6858 |
+
"loss": 3.7132,
|
6859 |
+
"step": 974
|
6860 |
+
},
|
6861 |
+
{
|
6862 |
+
"epoch": 0.35335689045936397,
|
6863 |
+
"grad_norm": 7.264481067657471,
|
6864 |
+
"learning_rate": 1.7077534966650766e-07,
|
6865 |
+
"loss": 2.8945,
|
6866 |
+
"step": 975
|
6867 |
+
},
|
6868 |
+
{
|
6869 |
+
"epoch": 0.353719307782912,
|
6870 |
+
"grad_norm": 5.050163269042969,
|
6871 |
+
"learning_rate": 1.5739359123178587e-07,
|
6872 |
+
"loss": 2.8021,
|
6873 |
+
"step": 976
|
6874 |
+
},
|
6875 |
+
{
|
6876 |
+
"epoch": 0.3540817251064601,
|
6877 |
+
"grad_norm": 6.886010646820068,
|
6878 |
+
"learning_rate": 1.4455690355525964e-07,
|
6879 |
+
"loss": 3.1517,
|
6880 |
+
"step": 977
|
6881 |
+
},
|
6882 |
+
{
|
6883 |
+
"epoch": 0.35444414243000816,
|
6884 |
+
"grad_norm": 5.721440315246582,
|
6885 |
+
"learning_rate": 1.3226542701689215e-07,
|
6886 |
+
"loss": 2.6086,
|
6887 |
+
"step": 978
|
6888 |
+
},
|
6889 |
+
{
|
6890 |
+
"epoch": 0.35480655975355624,
|
6891 |
+
"grad_norm": 5.630771160125732,
|
6892 |
+
"learning_rate": 1.2051929603428825e-07,
|
6893 |
+
"loss": 2.9555,
|
6894 |
+
"step": 979
|
6895 |
+
},
|
6896 |
+
{
|
6897 |
+
"epoch": 0.35516897707710426,
|
6898 |
+
"grad_norm": 6.419443130493164,
|
6899 |
+
"learning_rate": 1.0931863906127327e-07,
|
6900 |
+
"loss": 2.4218,
|
6901 |
+
"step": 980
|
6902 |
+
},
|
6903 |
+
{
|
6904 |
+
"epoch": 0.35553139440065235,
|
6905 |
+
"grad_norm": 8.198994636535645,
|
6906 |
+
"learning_rate": 9.866357858642205e-08,
|
6907 |
+
"loss": 2.5117,
|
6908 |
+
"step": 981
|
6909 |
+
},
|
6910 |
+
{
|
6911 |
+
"epoch": 0.3558938117242004,
|
6912 |
+
"grad_norm": 6.984527587890625,
|
6913 |
+
"learning_rate": 8.855423113177664e-08,
|
6914 |
+
"loss": 2.5872,
|
6915 |
+
"step": 982
|
6916 |
+
},
|
6917 |
+
{
|
6918 |
+
"epoch": 0.3562562290477485,
|
6919 |
+
"grad_norm": 6.198144435882568,
|
6920 |
+
"learning_rate": 7.899070725153613e-08,
|
6921 |
+
"loss": 2.5865,
|
6922 |
+
"step": 983
|
6923 |
+
},
|
6924 |
+
{
|
6925 |
+
"epoch": 0.35661864637129653,
|
6926 |
+
"grad_norm": 5.657758712768555,
|
6927 |
+
"learning_rate": 6.997311153086883e-08,
|
6928 |
+
"loss": 2.7767,
|
6929 |
+
"step": 984
|
6930 |
+
},
|
6931 |
+
{
|
6932 |
+
"epoch": 0.3569810636948446,
|
6933 |
+
"grad_norm": 8.617463111877441,
|
6934 |
+
"learning_rate": 6.150154258476315e-08,
|
6935 |
+
"loss": 2.5342,
|
6936 |
+
"step": 985
|
6937 |
+
},
|
6938 |
+
{
|
6939 |
+
"epoch": 0.3573434810183927,
|
6940 |
+
"grad_norm": 6.265313625335693,
|
6941 |
+
"learning_rate": 5.3576093056922906e-08,
|
6942 |
+
"loss": 1.9735,
|
6943 |
+
"step": 986
|
6944 |
+
},
|
6945 |
+
{
|
6946 |
+
"epoch": 0.3577058983419407,
|
6947 |
+
"grad_norm": 4.9574875831604,
|
6948 |
+
"learning_rate": 4.619684961881254e-08,
|
6949 |
+
"loss": 1.8526,
|
6950 |
+
"step": 987
|
6951 |
+
},
|
6952 |
+
{
|
6953 |
+
"epoch": 0.3580683156654888,
|
6954 |
+
"grad_norm": 6.38330602645874,
|
6955 |
+
"learning_rate": 3.936389296864129e-08,
|
6956 |
+
"loss": 3.3058,
|
6957 |
+
"step": 988
|
6958 |
+
},
|
6959 |
+
{
|
6960 |
+
"epoch": 0.3584307329890369,
|
6961 |
+
"grad_norm": 6.539586067199707,
|
6962 |
+
"learning_rate": 3.3077297830541584e-08,
|
6963 |
+
"loss": 3.1111,
|
6964 |
+
"step": 989
|
6965 |
+
},
|
6966 |
+
{
|
6967 |
+
"epoch": 0.35879315031258496,
|
6968 |
+
"grad_norm": 6.351942539215088,
|
6969 |
+
"learning_rate": 2.7337132953697554e-08,
|
6970 |
+
"loss": 2.7353,
|
6971 |
+
"step": 990
|
6972 |
+
},
|
6973 |
+
{
|
6974 |
+
"epoch": 0.359155567636133,
|
6975 |
+
"grad_norm": 8.30229377746582,
|
6976 |
+
"learning_rate": 2.214346111164556e-08,
|
6977 |
+
"loss": 2.9746,
|
6978 |
+
"step": 991
|
6979 |
+
},
|
6980 |
+
{
|
6981 |
+
"epoch": 0.35951798495968107,
|
6982 |
+
"grad_norm": 6.088867664337158,
|
6983 |
+
"learning_rate": 1.749633910153592e-08,
|
6984 |
+
"loss": 2.5657,
|
6985 |
+
"step": 992
|
6986 |
+
},
|
6987 |
+
{
|
6988 |
+
"epoch": 0.35988040228322915,
|
6989 |
+
"grad_norm": 5.72982931137085,
|
6990 |
+
"learning_rate": 1.3395817743561134e-08,
|
6991 |
+
"loss": 2.762,
|
6992 |
+
"step": 993
|
6993 |
+
},
|
6994 |
+
{
|
6995 |
+
"epoch": 0.36024281960677723,
|
6996 |
+
"grad_norm": 7.2047576904296875,
|
6997 |
+
"learning_rate": 9.841941880361916e-09,
|
6998 |
+
"loss": 2.6464,
|
6999 |
+
"step": 994
|
7000 |
+
},
|
7001 |
+
{
|
7002 |
+
"epoch": 0.36060523693032526,
|
7003 |
+
"grad_norm": 4.944203853607178,
|
7004 |
+
"learning_rate": 6.834750376549792e-09,
|
7005 |
+
"loss": 2.7518,
|
7006 |
+
"step": 995
|
7007 |
+
},
|
7008 |
+
{
|
7009 |
+
"epoch": 0.36096765425387334,
|
7010 |
+
"grad_norm": 6.469346523284912,
|
7011 |
+
"learning_rate": 4.3742761183018784e-09,
|
7012 |
+
"loss": 3.3189,
|
7013 |
+
"step": 996
|
7014 |
+
},
|
7015 |
+
{
|
7016 |
+
"epoch": 0.3613300715774214,
|
7017 |
+
"grad_norm": 5.523458480834961,
|
7018 |
+
"learning_rate": 2.4605460129556445e-09,
|
7019 |
+
"loss": 2.5312,
|
7020 |
+
"step": 997
|
7021 |
+
},
|
7022 |
+
{
|
7023 |
+
"epoch": 0.36169248890096944,
|
7024 |
+
"grad_norm": 6.366145133972168,
|
7025 |
+
"learning_rate": 1.0935809887702154e-09,
|
7026 |
+
"loss": 2.7652,
|
7027 |
+
"step": 998
|
7028 |
+
},
|
7029 |
+
{
|
7030 |
+
"epoch": 0.3620549062245175,
|
7031 |
+
"grad_norm": 6.140265941619873,
|
7032 |
+
"learning_rate": 2.7339599464326627e-10,
|
7033 |
+
"loss": 2.759,
|
7034 |
+
"step": 999
|
7035 |
+
},
|
7036 |
+
{
|
7037 |
+
"epoch": 0.3624173235480656,
|
7038 |
+
"grad_norm": 6.8456130027771,
|
7039 |
+
"learning_rate": 0.0,
|
7040 |
+
"loss": 2.8016,
|
7041 |
+
"step": 1000
|
7042 |
+
},
|
7043 |
+
{
|
7044 |
+
"epoch": 0.3624173235480656,
|
7045 |
+
"eval_loss": 2.871779203414917,
|
7046 |
+
"eval_runtime": 179.6734,
|
7047 |
+
"eval_samples_per_second": 6.467,
|
7048 |
+
"eval_steps_per_second": 3.234,
|
7049 |
+
"step": 1000
|
7050 |
}
|
7051 |
],
|
7052 |
"logging_steps": 1,
|
|
|
7061 |
"should_evaluate": false,
|
7062 |
"should_log": false,
|
7063 |
"should_save": true,
|
7064 |
+
"should_training_stop": true
|
7065 |
},
|
7066 |
"attributes": {}
|
7067 |
}
|
7068 |
},
|
7069 |
+
"total_flos": 7.209543008256e+16,
|
7070 |
"train_batch_size": 2,
|
7071 |
"trial_name": null,
|
7072 |
"trial_params": null
|