Training in progress, step 17000, checkpoint
Browse files- last-checkpoint/global_step17000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step17000/mp_rank_00_model_states.pt +3 -0
- 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_step17000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9d5e5ee5de88ba37830911f0b46c57d6fb97ef78baef3bbcb369caedac1d2537
|
| 3 |
+
size 5117197020
|
last-checkpoint/global_step17000/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3508f7d0f16906096e5610ffe0590bca5cd155de85a6a8021b6e5bf9d8e2eabb
|
| 3 |
+
size 859127504
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step17000
|
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:aa4f82ea40fb0305db931cf7a54215d8c646ba708abad07172d476a907b2dad4
|
| 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:be21ff914d7590ad2180b18bca69f62255c4deee5c5c2b727794908b9d148dcc
|
| 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 sha256:4aa6830d6aa63edbea9a9fa4aac3b79365984a3d18eed4b014dcec7309b75dc2
|
| 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,
|
|
@@ -4357,6 +4357,586 @@
|
|
| 4357 |
"eval_steps_per_second": 1.552,
|
| 4358 |
"eval_wer": 84.67703913248468,
|
| 4359 |
"step": 15000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4360 |
}
|
| 4361 |
],
|
| 4362 |
"logging_steps": 25,
|
|
@@ -4376,7 +4956,7 @@
|
|
| 4376 |
"attributes": {}
|
| 4377 |
}
|
| 4378 |
},
|
| 4379 |
-
"total_flos": 2.
|
| 4380 |
"train_batch_size": 4,
|
| 4381 |
"trial_name": null,
|
| 4382 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 82.65912305516267,
|
| 3 |
+
"best_model_checkpoint": "./iteboshi_temp/checkpoint-16000",
|
| 4 |
+
"epoch": 18.722466960352424,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 17000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 4357 |
"eval_steps_per_second": 1.552,
|
| 4358 |
"eval_wer": 84.67703913248468,
|
| 4359 |
"step": 15000
|
| 4360 |
+
},
|
| 4361 |
+
{
|
| 4362 |
+
"epoch": 16.547356828193834,
|
| 4363 |
+
"grad_norm": 0.10291285067796707,
|
| 4364 |
+
"learning_rate": 5.1025641025641024e-06,
|
| 4365 |
+
"loss": 0.0041,
|
| 4366 |
+
"step": 15025
|
| 4367 |
+
},
|
| 4368 |
+
{
|
| 4369 |
+
"epoch": 16.57488986784141,
|
| 4370 |
+
"grad_norm": 0.1358381062746048,
|
| 4371 |
+
"learning_rate": 5.076923076923077e-06,
|
| 4372 |
+
"loss": 0.0066,
|
| 4373 |
+
"step": 15050
|
| 4374 |
+
},
|
| 4375 |
+
{
|
| 4376 |
+
"epoch": 16.602422907488986,
|
| 4377 |
+
"grad_norm": 0.020193297415971756,
|
| 4378 |
+
"learning_rate": 5.051282051282051e-06,
|
| 4379 |
+
"loss": 0.0054,
|
| 4380 |
+
"step": 15075
|
| 4381 |
+
},
|
| 4382 |
+
{
|
| 4383 |
+
"epoch": 16.629955947136565,
|
| 4384 |
+
"grad_norm": 0.03404547646641731,
|
| 4385 |
+
"learning_rate": 5.025641025641026e-06,
|
| 4386 |
+
"loss": 0.003,
|
| 4387 |
+
"step": 15100
|
| 4388 |
+
},
|
| 4389 |
+
{
|
| 4390 |
+
"epoch": 16.65748898678414,
|
| 4391 |
+
"grad_norm": 0.07087010145187378,
|
| 4392 |
+
"learning_rate": 5e-06,
|
| 4393 |
+
"loss": 0.0043,
|
| 4394 |
+
"step": 15125
|
| 4395 |
+
},
|
| 4396 |
+
{
|
| 4397 |
+
"epoch": 16.685022026431717,
|
| 4398 |
+
"grad_norm": 0.01731196418404579,
|
| 4399 |
+
"learning_rate": 4.974358974358975e-06,
|
| 4400 |
+
"loss": 0.0054,
|
| 4401 |
+
"step": 15150
|
| 4402 |
+
},
|
| 4403 |
+
{
|
| 4404 |
+
"epoch": 16.712555066079297,
|
| 4405 |
+
"grad_norm": 0.3552390933036804,
|
| 4406 |
+
"learning_rate": 4.948717948717949e-06,
|
| 4407 |
+
"loss": 0.0056,
|
| 4408 |
+
"step": 15175
|
| 4409 |
+
},
|
| 4410 |
+
{
|
| 4411 |
+
"epoch": 16.740088105726873,
|
| 4412 |
+
"grad_norm": 0.626363217830658,
|
| 4413 |
+
"learning_rate": 4.923076923076924e-06,
|
| 4414 |
+
"loss": 0.0079,
|
| 4415 |
+
"step": 15200
|
| 4416 |
+
},
|
| 4417 |
+
{
|
| 4418 |
+
"epoch": 16.76762114537445,
|
| 4419 |
+
"grad_norm": 0.0181206613779068,
|
| 4420 |
+
"learning_rate": 4.8974358974358975e-06,
|
| 4421 |
+
"loss": 0.006,
|
| 4422 |
+
"step": 15225
|
| 4423 |
+
},
|
| 4424 |
+
{
|
| 4425 |
+
"epoch": 16.795154185022028,
|
| 4426 |
+
"grad_norm": 0.09998787939548492,
|
| 4427 |
+
"learning_rate": 4.871794871794872e-06,
|
| 4428 |
+
"loss": 0.0055,
|
| 4429 |
+
"step": 15250
|
| 4430 |
+
},
|
| 4431 |
+
{
|
| 4432 |
+
"epoch": 16.822687224669604,
|
| 4433 |
+
"grad_norm": 0.1320696324110031,
|
| 4434 |
+
"learning_rate": 4.8461538461538465e-06,
|
| 4435 |
+
"loss": 0.004,
|
| 4436 |
+
"step": 15275
|
| 4437 |
+
},
|
| 4438 |
+
{
|
| 4439 |
+
"epoch": 16.85022026431718,
|
| 4440 |
+
"grad_norm": 0.09725293517112732,
|
| 4441 |
+
"learning_rate": 4.820512820512821e-06,
|
| 4442 |
+
"loss": 0.0047,
|
| 4443 |
+
"step": 15300
|
| 4444 |
+
},
|
| 4445 |
+
{
|
| 4446 |
+
"epoch": 16.87775330396476,
|
| 4447 |
+
"grad_norm": 0.11701034754514694,
|
| 4448 |
+
"learning_rate": 4.7948717948717955e-06,
|
| 4449 |
+
"loss": 0.0038,
|
| 4450 |
+
"step": 15325
|
| 4451 |
+
},
|
| 4452 |
+
{
|
| 4453 |
+
"epoch": 16.905286343612335,
|
| 4454 |
+
"grad_norm": 0.046660326421260834,
|
| 4455 |
+
"learning_rate": 4.76923076923077e-06,
|
| 4456 |
+
"loss": 0.0052,
|
| 4457 |
+
"step": 15350
|
| 4458 |
+
},
|
| 4459 |
+
{
|
| 4460 |
+
"epoch": 16.93281938325991,
|
| 4461 |
+
"grad_norm": 0.03657762333750725,
|
| 4462 |
+
"learning_rate": 4.743589743589744e-06,
|
| 4463 |
+
"loss": 0.004,
|
| 4464 |
+
"step": 15375
|
| 4465 |
+
},
|
| 4466 |
+
{
|
| 4467 |
+
"epoch": 16.96035242290749,
|
| 4468 |
+
"grad_norm": 0.13504423201084137,
|
| 4469 |
+
"learning_rate": 4.717948717948718e-06,
|
| 4470 |
+
"loss": 0.0038,
|
| 4471 |
+
"step": 15400
|
| 4472 |
+
},
|
| 4473 |
+
{
|
| 4474 |
+
"epoch": 16.987885462555067,
|
| 4475 |
+
"grad_norm": 0.08526286482810974,
|
| 4476 |
+
"learning_rate": 4.692307692307693e-06,
|
| 4477 |
+
"loss": 0.0066,
|
| 4478 |
+
"step": 15425
|
| 4479 |
+
},
|
| 4480 |
+
{
|
| 4481 |
+
"epoch": 17.015418502202643,
|
| 4482 |
+
"grad_norm": 0.02115774340927601,
|
| 4483 |
+
"learning_rate": 4.666666666666667e-06,
|
| 4484 |
+
"loss": 0.0046,
|
| 4485 |
+
"step": 15450
|
| 4486 |
+
},
|
| 4487 |
+
{
|
| 4488 |
+
"epoch": 17.042951541850222,
|
| 4489 |
+
"grad_norm": 0.2753286063671112,
|
| 4490 |
+
"learning_rate": 4.641025641025642e-06,
|
| 4491 |
+
"loss": 0.0045,
|
| 4492 |
+
"step": 15475
|
| 4493 |
+
},
|
| 4494 |
+
{
|
| 4495 |
+
"epoch": 17.070484581497798,
|
| 4496 |
+
"grad_norm": 0.03063320554792881,
|
| 4497 |
+
"learning_rate": 4.615384615384616e-06,
|
| 4498 |
+
"loss": 0.0045,
|
| 4499 |
+
"step": 15500
|
| 4500 |
+
},
|
| 4501 |
+
{
|
| 4502 |
+
"epoch": 17.098017621145374,
|
| 4503 |
+
"grad_norm": 0.010663969442248344,
|
| 4504 |
+
"learning_rate": 4.58974358974359e-06,
|
| 4505 |
+
"loss": 0.003,
|
| 4506 |
+
"step": 15525
|
| 4507 |
+
},
|
| 4508 |
+
{
|
| 4509 |
+
"epoch": 17.125550660792953,
|
| 4510 |
+
"grad_norm": 0.22800272703170776,
|
| 4511 |
+
"learning_rate": 4.564102564102564e-06,
|
| 4512 |
+
"loss": 0.002,
|
| 4513 |
+
"step": 15550
|
| 4514 |
+
},
|
| 4515 |
+
{
|
| 4516 |
+
"epoch": 17.15308370044053,
|
| 4517 |
+
"grad_norm": 0.22873614728450775,
|
| 4518 |
+
"learning_rate": 4.538461538461539e-06,
|
| 4519 |
+
"loss": 0.002,
|
| 4520 |
+
"step": 15575
|
| 4521 |
+
},
|
| 4522 |
+
{
|
| 4523 |
+
"epoch": 17.180616740088105,
|
| 4524 |
+
"grad_norm": 0.011216685175895691,
|
| 4525 |
+
"learning_rate": 4.512820512820513e-06,
|
| 4526 |
+
"loss": 0.0032,
|
| 4527 |
+
"step": 15600
|
| 4528 |
+
},
|
| 4529 |
+
{
|
| 4530 |
+
"epoch": 17.208149779735685,
|
| 4531 |
+
"grad_norm": 0.4962778091430664,
|
| 4532 |
+
"learning_rate": 4.487179487179488e-06,
|
| 4533 |
+
"loss": 0.0038,
|
| 4534 |
+
"step": 15625
|
| 4535 |
+
},
|
| 4536 |
+
{
|
| 4537 |
+
"epoch": 17.23568281938326,
|
| 4538 |
+
"grad_norm": 0.023572538048028946,
|
| 4539 |
+
"learning_rate": 4.461538461538462e-06,
|
| 4540 |
+
"loss": 0.0031,
|
| 4541 |
+
"step": 15650
|
| 4542 |
+
},
|
| 4543 |
+
{
|
| 4544 |
+
"epoch": 17.263215859030836,
|
| 4545 |
+
"grad_norm": 0.032316386699676514,
|
| 4546 |
+
"learning_rate": 4.435897435897436e-06,
|
| 4547 |
+
"loss": 0.0024,
|
| 4548 |
+
"step": 15675
|
| 4549 |
+
},
|
| 4550 |
+
{
|
| 4551 |
+
"epoch": 17.290748898678412,
|
| 4552 |
+
"grad_norm": 0.025533461943268776,
|
| 4553 |
+
"learning_rate": 4.4102564102564104e-06,
|
| 4554 |
+
"loss": 0.0024,
|
| 4555 |
+
"step": 15700
|
| 4556 |
+
},
|
| 4557 |
+
{
|
| 4558 |
+
"epoch": 17.318281938325992,
|
| 4559 |
+
"grad_norm": 0.02447574771940708,
|
| 4560 |
+
"learning_rate": 4.384615384615385e-06,
|
| 4561 |
+
"loss": 0.0057,
|
| 4562 |
+
"step": 15725
|
| 4563 |
+
},
|
| 4564 |
+
{
|
| 4565 |
+
"epoch": 17.345814977973568,
|
| 4566 |
+
"grad_norm": 0.021540969610214233,
|
| 4567 |
+
"learning_rate": 4.358974358974359e-06,
|
| 4568 |
+
"loss": 0.0035,
|
| 4569 |
+
"step": 15750
|
| 4570 |
+
},
|
| 4571 |
+
{
|
| 4572 |
+
"epoch": 17.373348017621144,
|
| 4573 |
+
"grad_norm": 0.022210588678717613,
|
| 4574 |
+
"learning_rate": 4.333333333333334e-06,
|
| 4575 |
+
"loss": 0.0029,
|
| 4576 |
+
"step": 15775
|
| 4577 |
+
},
|
| 4578 |
+
{
|
| 4579 |
+
"epoch": 17.400881057268723,
|
| 4580 |
+
"grad_norm": 0.01674061268568039,
|
| 4581 |
+
"learning_rate": 4.307692307692308e-06,
|
| 4582 |
+
"loss": 0.0037,
|
| 4583 |
+
"step": 15800
|
| 4584 |
+
},
|
| 4585 |
+
{
|
| 4586 |
+
"epoch": 17.4284140969163,
|
| 4587 |
+
"grad_norm": 0.013861955143511295,
|
| 4588 |
+
"learning_rate": 4.282051282051282e-06,
|
| 4589 |
+
"loss": 0.0023,
|
| 4590 |
+
"step": 15825
|
| 4591 |
+
},
|
| 4592 |
+
{
|
| 4593 |
+
"epoch": 17.455947136563875,
|
| 4594 |
+
"grad_norm": 0.023190615698695183,
|
| 4595 |
+
"learning_rate": 4.2564102564102566e-06,
|
| 4596 |
+
"loss": 0.0045,
|
| 4597 |
+
"step": 15850
|
| 4598 |
+
},
|
| 4599 |
+
{
|
| 4600 |
+
"epoch": 17.483480176211454,
|
| 4601 |
+
"grad_norm": 0.015583349391818047,
|
| 4602 |
+
"learning_rate": 4.230769230769231e-06,
|
| 4603 |
+
"loss": 0.0037,
|
| 4604 |
+
"step": 15875
|
| 4605 |
+
},
|
| 4606 |
+
{
|
| 4607 |
+
"epoch": 17.51101321585903,
|
| 4608 |
+
"grad_norm": 0.013716256245970726,
|
| 4609 |
+
"learning_rate": 4.2051282051282055e-06,
|
| 4610 |
+
"loss": 0.0035,
|
| 4611 |
+
"step": 15900
|
| 4612 |
+
},
|
| 4613 |
+
{
|
| 4614 |
+
"epoch": 17.538546255506606,
|
| 4615 |
+
"grad_norm": 0.016609976068139076,
|
| 4616 |
+
"learning_rate": 4.17948717948718e-06,
|
| 4617 |
+
"loss": 0.0041,
|
| 4618 |
+
"step": 15925
|
| 4619 |
+
},
|
| 4620 |
+
{
|
| 4621 |
+
"epoch": 17.566079295154186,
|
| 4622 |
+
"grad_norm": 0.01952126808464527,
|
| 4623 |
+
"learning_rate": 4.1538461538461545e-06,
|
| 4624 |
+
"loss": 0.0037,
|
| 4625 |
+
"step": 15950
|
| 4626 |
+
},
|
| 4627 |
+
{
|
| 4628 |
+
"epoch": 17.59361233480176,
|
| 4629 |
+
"grad_norm": 0.17321471869945526,
|
| 4630 |
+
"learning_rate": 4.128205128205128e-06,
|
| 4631 |
+
"loss": 0.0039,
|
| 4632 |
+
"step": 15975
|
| 4633 |
+
},
|
| 4634 |
+
{
|
| 4635 |
+
"epoch": 17.621145374449338,
|
| 4636 |
+
"grad_norm": 0.01522456482052803,
|
| 4637 |
+
"learning_rate": 4.102564102564103e-06,
|
| 4638 |
+
"loss": 0.0032,
|
| 4639 |
+
"step": 16000
|
| 4640 |
+
},
|
| 4641 |
+
{
|
| 4642 |
+
"epoch": 17.621145374449338,
|
| 4643 |
+
"eval_cer": 23.317437208395905,
|
| 4644 |
+
"eval_loss": 0.8810132741928101,
|
| 4645 |
+
"eval_runtime": 1715.3061,
|
| 4646 |
+
"eval_samples_per_second": 6.169,
|
| 4647 |
+
"eval_steps_per_second": 1.543,
|
| 4648 |
+
"eval_wer": 82.65912305516267,
|
| 4649 |
+
"step": 16000
|
| 4650 |
+
},
|
| 4651 |
+
{
|
| 4652 |
+
"epoch": 17.648678414096917,
|
| 4653 |
+
"grad_norm": 0.30109259486198425,
|
| 4654 |
+
"learning_rate": 4.076923076923077e-06,
|
| 4655 |
+
"loss": 0.0026,
|
| 4656 |
+
"step": 16025
|
| 4657 |
+
},
|
| 4658 |
+
{
|
| 4659 |
+
"epoch": 17.676211453744493,
|
| 4660 |
+
"grad_norm": 0.06414441019296646,
|
| 4661 |
+
"learning_rate": 4.051282051282052e-06,
|
| 4662 |
+
"loss": 0.0032,
|
| 4663 |
+
"step": 16050
|
| 4664 |
+
},
|
| 4665 |
+
{
|
| 4666 |
+
"epoch": 17.70374449339207,
|
| 4667 |
+
"grad_norm": 0.12120723724365234,
|
| 4668 |
+
"learning_rate": 4.025641025641026e-06,
|
| 4669 |
+
"loss": 0.0026,
|
| 4670 |
+
"step": 16075
|
| 4671 |
+
},
|
| 4672 |
+
{
|
| 4673 |
+
"epoch": 17.73127753303965,
|
| 4674 |
+
"grad_norm": 0.19122646749019623,
|
| 4675 |
+
"learning_rate": 4.000000000000001e-06,
|
| 4676 |
+
"loss": 0.0039,
|
| 4677 |
+
"step": 16100
|
| 4678 |
+
},
|
| 4679 |
+
{
|
| 4680 |
+
"epoch": 17.758810572687224,
|
| 4681 |
+
"grad_norm": 0.03467703238129616,
|
| 4682 |
+
"learning_rate": 3.974358974358974e-06,
|
| 4683 |
+
"loss": 0.0027,
|
| 4684 |
+
"step": 16125
|
| 4685 |
+
},
|
| 4686 |
+
{
|
| 4687 |
+
"epoch": 17.7863436123348,
|
| 4688 |
+
"grad_norm": 0.01050791796296835,
|
| 4689 |
+
"learning_rate": 3.948717948717949e-06,
|
| 4690 |
+
"loss": 0.0024,
|
| 4691 |
+
"step": 16150
|
| 4692 |
+
},
|
| 4693 |
+
{
|
| 4694 |
+
"epoch": 17.81387665198238,
|
| 4695 |
+
"grad_norm": 0.01552590075880289,
|
| 4696 |
+
"learning_rate": 3.923076923076923e-06,
|
| 4697 |
+
"loss": 0.003,
|
| 4698 |
+
"step": 16175
|
| 4699 |
+
},
|
| 4700 |
+
{
|
| 4701 |
+
"epoch": 17.841409691629956,
|
| 4702 |
+
"grad_norm": 0.021685760468244553,
|
| 4703 |
+
"learning_rate": 3.897435897435898e-06,
|
| 4704 |
+
"loss": 0.0028,
|
| 4705 |
+
"step": 16200
|
| 4706 |
+
},
|
| 4707 |
+
{
|
| 4708 |
+
"epoch": 17.86894273127753,
|
| 4709 |
+
"grad_norm": 0.012615197338163853,
|
| 4710 |
+
"learning_rate": 3.871794871794872e-06,
|
| 4711 |
+
"loss": 0.0018,
|
| 4712 |
+
"step": 16225
|
| 4713 |
+
},
|
| 4714 |
+
{
|
| 4715 |
+
"epoch": 17.89647577092511,
|
| 4716 |
+
"grad_norm": 0.024285893887281418,
|
| 4717 |
+
"learning_rate": 3.846153846153847e-06,
|
| 4718 |
+
"loss": 0.0034,
|
| 4719 |
+
"step": 16250
|
| 4720 |
+
},
|
| 4721 |
+
{
|
| 4722 |
+
"epoch": 17.924008810572687,
|
| 4723 |
+
"grad_norm": 0.019548427313566208,
|
| 4724 |
+
"learning_rate": 3.8205128205128204e-06,
|
| 4725 |
+
"loss": 0.0046,
|
| 4726 |
+
"step": 16275
|
| 4727 |
+
},
|
| 4728 |
+
{
|
| 4729 |
+
"epoch": 17.951541850220263,
|
| 4730 |
+
"grad_norm": 0.014185987412929535,
|
| 4731 |
+
"learning_rate": 3.794871794871795e-06,
|
| 4732 |
+
"loss": 0.0042,
|
| 4733 |
+
"step": 16300
|
| 4734 |
+
},
|
| 4735 |
+
{
|
| 4736 |
+
"epoch": 17.979074889867842,
|
| 4737 |
+
"grad_norm": 0.2013942152261734,
|
| 4738 |
+
"learning_rate": 3.7692307692307694e-06,
|
| 4739 |
+
"loss": 0.0025,
|
| 4740 |
+
"step": 16325
|
| 4741 |
+
},
|
| 4742 |
+
{
|
| 4743 |
+
"epoch": 18.006607929515418,
|
| 4744 |
+
"grad_norm": 0.01142708957195282,
|
| 4745 |
+
"learning_rate": 3.743589743589744e-06,
|
| 4746 |
+
"loss": 0.0028,
|
| 4747 |
+
"step": 16350
|
| 4748 |
+
},
|
| 4749 |
+
{
|
| 4750 |
+
"epoch": 18.034140969162994,
|
| 4751 |
+
"grad_norm": 0.1827182024717331,
|
| 4752 |
+
"learning_rate": 3.7179487179487184e-06,
|
| 4753 |
+
"loss": 0.0027,
|
| 4754 |
+
"step": 16375
|
| 4755 |
+
},
|
| 4756 |
+
{
|
| 4757 |
+
"epoch": 18.061674008810574,
|
| 4758 |
+
"grad_norm": 0.008858841843903065,
|
| 4759 |
+
"learning_rate": 3.692307692307693e-06,
|
| 4760 |
+
"loss": 0.0022,
|
| 4761 |
+
"step": 16400
|
| 4762 |
+
},
|
| 4763 |
+
{
|
| 4764 |
+
"epoch": 18.08920704845815,
|
| 4765 |
+
"grad_norm": 0.037348657846450806,
|
| 4766 |
+
"learning_rate": 3.6666666666666666e-06,
|
| 4767 |
+
"loss": 0.004,
|
| 4768 |
+
"step": 16425
|
| 4769 |
+
},
|
| 4770 |
+
{
|
| 4771 |
+
"epoch": 18.116740088105725,
|
| 4772 |
+
"grad_norm": 0.014842098578810692,
|
| 4773 |
+
"learning_rate": 3.641025641025641e-06,
|
| 4774 |
+
"loss": 0.003,
|
| 4775 |
+
"step": 16450
|
| 4776 |
+
},
|
| 4777 |
+
{
|
| 4778 |
+
"epoch": 18.144273127753305,
|
| 4779 |
+
"grad_norm": 0.012190734967589378,
|
| 4780 |
+
"learning_rate": 3.6153846153846156e-06,
|
| 4781 |
+
"loss": 0.0047,
|
| 4782 |
+
"step": 16475
|
| 4783 |
+
},
|
| 4784 |
+
{
|
| 4785 |
+
"epoch": 18.17180616740088,
|
| 4786 |
+
"grad_norm": 0.010254699736833572,
|
| 4787 |
+
"learning_rate": 3.58974358974359e-06,
|
| 4788 |
+
"loss": 0.0018,
|
| 4789 |
+
"step": 16500
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 18.199339207048457,
|
| 4793 |
+
"grad_norm": 0.012803646735846996,
|
| 4794 |
+
"learning_rate": 3.5641025641025646e-06,
|
| 4795 |
+
"loss": 0.0018,
|
| 4796 |
+
"step": 16525
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 18.226872246696036,
|
| 4800 |
+
"grad_norm": 0.010007087141275406,
|
| 4801 |
+
"learning_rate": 3.538461538461539e-06,
|
| 4802 |
+
"loss": 0.0037,
|
| 4803 |
+
"step": 16550
|
| 4804 |
+
},
|
| 4805 |
+
{
|
| 4806 |
+
"epoch": 18.254405286343612,
|
| 4807 |
+
"grad_norm": 0.010007468052208424,
|
| 4808 |
+
"learning_rate": 3.5128205128205127e-06,
|
| 4809 |
+
"loss": 0.0017,
|
| 4810 |
+
"step": 16575
|
| 4811 |
+
},
|
| 4812 |
+
{
|
| 4813 |
+
"epoch": 18.281938325991188,
|
| 4814 |
+
"grad_norm": 0.021304214373230934,
|
| 4815 |
+
"learning_rate": 3.487179487179487e-06,
|
| 4816 |
+
"loss": 0.0017,
|
| 4817 |
+
"step": 16600
|
| 4818 |
+
},
|
| 4819 |
+
{
|
| 4820 |
+
"epoch": 18.309471365638768,
|
| 4821 |
+
"grad_norm": 0.00610103365033865,
|
| 4822 |
+
"learning_rate": 3.4615384615384617e-06,
|
| 4823 |
+
"loss": 0.0017,
|
| 4824 |
+
"step": 16625
|
| 4825 |
+
},
|
| 4826 |
+
{
|
| 4827 |
+
"epoch": 18.337004405286343,
|
| 4828 |
+
"grad_norm": 0.17184419929981232,
|
| 4829 |
+
"learning_rate": 3.435897435897436e-06,
|
| 4830 |
+
"loss": 0.0023,
|
| 4831 |
+
"step": 16650
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 18.36453744493392,
|
| 4835 |
+
"grad_norm": 0.010224095545709133,
|
| 4836 |
+
"learning_rate": 3.4102564102564107e-06,
|
| 4837 |
+
"loss": 0.0028,
|
| 4838 |
+
"step": 16675
|
| 4839 |
+
},
|
| 4840 |
+
{
|
| 4841 |
+
"epoch": 18.3920704845815,
|
| 4842 |
+
"grad_norm": 0.016741087660193443,
|
| 4843 |
+
"learning_rate": 3.384615384615385e-06,
|
| 4844 |
+
"loss": 0.002,
|
| 4845 |
+
"step": 16700
|
| 4846 |
+
},
|
| 4847 |
+
{
|
| 4848 |
+
"epoch": 18.419603524229075,
|
| 4849 |
+
"grad_norm": 0.01324927993118763,
|
| 4850 |
+
"learning_rate": 3.358974358974359e-06,
|
| 4851 |
+
"loss": 0.0017,
|
| 4852 |
+
"step": 16725
|
| 4853 |
+
},
|
| 4854 |
+
{
|
| 4855 |
+
"epoch": 18.44713656387665,
|
| 4856 |
+
"grad_norm": 0.14577801525592804,
|
| 4857 |
+
"learning_rate": 3.3333333333333333e-06,
|
| 4858 |
+
"loss": 0.0025,
|
| 4859 |
+
"step": 16750
|
| 4860 |
+
},
|
| 4861 |
+
{
|
| 4862 |
+
"epoch": 18.47466960352423,
|
| 4863 |
+
"grad_norm": 0.0260769035667181,
|
| 4864 |
+
"learning_rate": 3.307692307692308e-06,
|
| 4865 |
+
"loss": 0.0018,
|
| 4866 |
+
"step": 16775
|
| 4867 |
+
},
|
| 4868 |
+
{
|
| 4869 |
+
"epoch": 18.502202643171806,
|
| 4870 |
+
"grad_norm": 0.01632179506123066,
|
| 4871 |
+
"learning_rate": 3.2820512820512823e-06,
|
| 4872 |
+
"loss": 0.0041,
|
| 4873 |
+
"step": 16800
|
| 4874 |
+
},
|
| 4875 |
+
{
|
| 4876 |
+
"epoch": 18.529735682819382,
|
| 4877 |
+
"grad_norm": 0.014896622858941555,
|
| 4878 |
+
"learning_rate": 3.256410256410257e-06,
|
| 4879 |
+
"loss": 0.0018,
|
| 4880 |
+
"step": 16825
|
| 4881 |
+
},
|
| 4882 |
+
{
|
| 4883 |
+
"epoch": 18.55726872246696,
|
| 4884 |
+
"grad_norm": 0.014535325579345226,
|
| 4885 |
+
"learning_rate": 3.2307692307692313e-06,
|
| 4886 |
+
"loss": 0.0022,
|
| 4887 |
+
"step": 16850
|
| 4888 |
+
},
|
| 4889 |
+
{
|
| 4890 |
+
"epoch": 18.584801762114537,
|
| 4891 |
+
"grad_norm": 0.011787498369812965,
|
| 4892 |
+
"learning_rate": 3.205128205128206e-06,
|
| 4893 |
+
"loss": 0.0016,
|
| 4894 |
+
"step": 16875
|
| 4895 |
+
},
|
| 4896 |
+
{
|
| 4897 |
+
"epoch": 18.612334801762113,
|
| 4898 |
+
"grad_norm": 0.04083514213562012,
|
| 4899 |
+
"learning_rate": 3.1794871794871795e-06,
|
| 4900 |
+
"loss": 0.0017,
|
| 4901 |
+
"step": 16900
|
| 4902 |
+
},
|
| 4903 |
+
{
|
| 4904 |
+
"epoch": 18.639867841409693,
|
| 4905 |
+
"grad_norm": 0.16764149069786072,
|
| 4906 |
+
"learning_rate": 3.153846153846154e-06,
|
| 4907 |
+
"loss": 0.0024,
|
| 4908 |
+
"step": 16925
|
| 4909 |
+
},
|
| 4910 |
+
{
|
| 4911 |
+
"epoch": 18.66740088105727,
|
| 4912 |
+
"grad_norm": 0.008704649284482002,
|
| 4913 |
+
"learning_rate": 3.1282051282051284e-06,
|
| 4914 |
+
"loss": 0.0015,
|
| 4915 |
+
"step": 16950
|
| 4916 |
+
},
|
| 4917 |
+
{
|
| 4918 |
+
"epoch": 18.694933920704845,
|
| 4919 |
+
"grad_norm": 0.007399390451610088,
|
| 4920 |
+
"learning_rate": 3.102564102564103e-06,
|
| 4921 |
+
"loss": 0.0014,
|
| 4922 |
+
"step": 16975
|
| 4923 |
+
},
|
| 4924 |
+
{
|
| 4925 |
+
"epoch": 18.722466960352424,
|
| 4926 |
+
"grad_norm": 0.016065089032053947,
|
| 4927 |
+
"learning_rate": 3.0769230769230774e-06,
|
| 4928 |
+
"loss": 0.0017,
|
| 4929 |
+
"step": 17000
|
| 4930 |
+
},
|
| 4931 |
+
{
|
| 4932 |
+
"epoch": 18.722466960352424,
|
| 4933 |
+
"eval_cer": 22.853163367074387,
|
| 4934 |
+
"eval_loss": 0.8870487809181213,
|
| 4935 |
+
"eval_runtime": 1702.6926,
|
| 4936 |
+
"eval_samples_per_second": 6.214,
|
| 4937 |
+
"eval_steps_per_second": 1.554,
|
| 4938 |
+
"eval_wer": 82.998585572843,
|
| 4939 |
+
"step": 17000
|
| 4940 |
}
|
| 4941 |
],
|
| 4942 |
"logging_steps": 25,
|
|
|
|
| 4956 |
"attributes": {}
|
| 4957 |
}
|
| 4958 |
},
|
| 4959 |
+
"total_flos": 2.920843417033166e+20,
|
| 4960 |
"train_batch_size": 4,
|
| 4961 |
"trial_name": null,
|
| 4962 |
"trial_params": null
|