Training in progress, epoch 1, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +445 -4
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 201892112
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5e26503ee5864af07ce4030836db51a4668d97e466902ce26562a2ae8f4dbd1
|
| 3 |
size 201892112
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 102864868
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7f87dfd3983905fc6f45e08cd5447e3695907754f080349935e996e48fe7470
|
| 3 |
size 102864868
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ba9dafcc54ce9ff084db79366d7b940d91787c2cfb616ba91a4bb22cd048faed
|
| 3 |
size 15984
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f5dfab0aba1f4849abd9625392cb4af7139fd3842b09daef021f142b39cff71
|
| 3 |
size 15984
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43a4c7a3e3866946bfa6c6f5a1716563f36029ceb622e5804a6f396508c55ea3
|
| 3 |
size 15984
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b39933ef00f6fb7daa29790fb90e8ff14fa4ec97393154716671cac132e30db
|
| 3 |
size 15984
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0c37e29f749e6da69d82af0a38f01b394a754fce56fccbdf27902a5d585e43f
|
| 3 |
size 15984
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33c091df1ce77bff431f3379561fab3489657fa2cde7f17e48d31b6b28bf660c
|
| 3 |
size 15984
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83e3de474f59434cf28a8c41406b1e8bd00ec6293d50ff81f9c75db3e70e91e0
|
| 3 |
size 15984
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6ca6a786fe6d9b25cb7742ed56fbbd3f68052ae10454dcdbf494e787b2682812
|
| 3 |
size 15984
|
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:b4ac31dcfba988823188c44099086577dd09e162577217892c3f4fd5ad8489c2
|
| 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":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -4466,6 +4466,447 @@
|
|
| 4466 |
"learning_rate": 5.445378758417925e-06,
|
| 4467 |
"loss": 1.063,
|
| 4468 |
"step": 637
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4469 |
}
|
| 4470 |
],
|
| 4471 |
"logging_steps": 1,
|
|
@@ -4480,12 +4921,12 @@
|
|
| 4480 |
"should_evaluate": false,
|
| 4481 |
"should_log": false,
|
| 4482 |
"should_save": true,
|
| 4483 |
-
"should_training_stop":
|
| 4484 |
},
|
| 4485 |
"attributes": {}
|
| 4486 |
}
|
| 4487 |
},
|
| 4488 |
-
"total_flos": 2.
|
| 4489 |
"train_batch_size": 4,
|
| 4490 |
"trial_name": null,
|
| 4491 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0984699882306788,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 700,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 4466 |
"learning_rate": 5.445378758417925e-06,
|
| 4467 |
"loss": 1.063,
|
| 4468 |
"step": 637
|
| 4469 |
+
},
|
| 4470 |
+
{
|
| 4471 |
+
"epoch": 1.001176932130247,
|
| 4472 |
+
"grad_norm": 3.3213553428649902,
|
| 4473 |
+
"learning_rate": 5.275390589031859e-06,
|
| 4474 |
+
"loss": 1.7375,
|
| 4475 |
+
"step": 638
|
| 4476 |
+
},
|
| 4477 |
+
{
|
| 4478 |
+
"epoch": 1.0027461749705766,
|
| 4479 |
+
"grad_norm": 1.4553793668746948,
|
| 4480 |
+
"learning_rate": 5.108026758167719e-06,
|
| 4481 |
+
"loss": 0.9496,
|
| 4482 |
+
"step": 639
|
| 4483 |
+
},
|
| 4484 |
+
{
|
| 4485 |
+
"epoch": 1.0043154178109062,
|
| 4486 |
+
"grad_norm": 1.3991262912750244,
|
| 4487 |
+
"learning_rate": 4.943291854189493e-06,
|
| 4488 |
+
"loss": 0.9158,
|
| 4489 |
+
"step": 640
|
| 4490 |
+
},
|
| 4491 |
+
{
|
| 4492 |
+
"epoch": 1.0058846606512357,
|
| 4493 |
+
"grad_norm": 1.5718517303466797,
|
| 4494 |
+
"learning_rate": 4.781190393387796e-06,
|
| 4495 |
+
"loss": 1.032,
|
| 4496 |
+
"step": 641
|
| 4497 |
+
},
|
| 4498 |
+
{
|
| 4499 |
+
"epoch": 1.0074539034915653,
|
| 4500 |
+
"grad_norm": 1.546706199645996,
|
| 4501 |
+
"learning_rate": 4.6217268198560404e-06,
|
| 4502 |
+
"loss": 0.9635,
|
| 4503 |
+
"step": 642
|
| 4504 |
+
},
|
| 4505 |
+
{
|
| 4506 |
+
"epoch": 1.0090231463318948,
|
| 4507 |
+
"grad_norm": 1.4677547216415405,
|
| 4508 |
+
"learning_rate": 4.464905505368658e-06,
|
| 4509 |
+
"loss": 0.9359,
|
| 4510 |
+
"step": 643
|
| 4511 |
+
},
|
| 4512 |
+
{
|
| 4513 |
+
"epoch": 1.0105923891722244,
|
| 4514 |
+
"grad_norm": 1.4244987964630127,
|
| 4515 |
+
"learning_rate": 4.3107307492612086e-06,
|
| 4516 |
+
"loss": 0.8894,
|
| 4517 |
+
"step": 644
|
| 4518 |
+
},
|
| 4519 |
+
{
|
| 4520 |
+
"epoch": 1.012161632012554,
|
| 4521 |
+
"grad_norm": 1.470013976097107,
|
| 4522 |
+
"learning_rate": 4.1592067783125015e-06,
|
| 4523 |
+
"loss": 1.0618,
|
| 4524 |
+
"step": 645
|
| 4525 |
+
},
|
| 4526 |
+
{
|
| 4527 |
+
"epoch": 1.0137308748528835,
|
| 4528 |
+
"grad_norm": 1.5071467161178589,
|
| 4529 |
+
"learning_rate": 4.010337746628751e-06,
|
| 4530 |
+
"loss": 0.9016,
|
| 4531 |
+
"step": 646
|
| 4532 |
+
},
|
| 4533 |
+
{
|
| 4534 |
+
"epoch": 1.015300117693213,
|
| 4535 |
+
"grad_norm": 1.406595230102539,
|
| 4536 |
+
"learning_rate": 3.864127735529656e-06,
|
| 4537 |
+
"loss": 0.9207,
|
| 4538 |
+
"step": 647
|
| 4539 |
+
},
|
| 4540 |
+
{
|
| 4541 |
+
"epoch": 1.0168693605335426,
|
| 4542 |
+
"grad_norm": 1.355204701423645,
|
| 4543 |
+
"learning_rate": 3.7205807534365315e-06,
|
| 4544 |
+
"loss": 0.9483,
|
| 4545 |
+
"step": 648
|
| 4546 |
+
},
|
| 4547 |
+
{
|
| 4548 |
+
"epoch": 1.0184386033738722,
|
| 4549 |
+
"grad_norm": 1.5814651250839233,
|
| 4550 |
+
"learning_rate": 3.5797007357623945e-06,
|
| 4551 |
+
"loss": 1.0056,
|
| 4552 |
+
"step": 649
|
| 4553 |
+
},
|
| 4554 |
+
{
|
| 4555 |
+
"epoch": 1.0200078462142017,
|
| 4556 |
+
"grad_norm": 1.5514806509017944,
|
| 4557 |
+
"learning_rate": 3.441491544804112e-06,
|
| 4558 |
+
"loss": 0.9282,
|
| 4559 |
+
"step": 650
|
| 4560 |
+
},
|
| 4561 |
+
{
|
| 4562 |
+
"epoch": 1.0215770890545313,
|
| 4563 |
+
"grad_norm": 1.3481324911117554,
|
| 4564 |
+
"learning_rate": 3.3059569696364502e-06,
|
| 4565 |
+
"loss": 0.853,
|
| 4566 |
+
"step": 651
|
| 4567 |
+
},
|
| 4568 |
+
{
|
| 4569 |
+
"epoch": 1.0231463318948608,
|
| 4570 |
+
"grad_norm": 1.4449180364608765,
|
| 4571 |
+
"learning_rate": 3.1731007260082616e-06,
|
| 4572 |
+
"loss": 0.9905,
|
| 4573 |
+
"step": 652
|
| 4574 |
+
},
|
| 4575 |
+
{
|
| 4576 |
+
"epoch": 1.0247155747351904,
|
| 4577 |
+
"grad_norm": 1.6003049612045288,
|
| 4578 |
+
"learning_rate": 3.0429264562405776e-06,
|
| 4579 |
+
"loss": 0.9784,
|
| 4580 |
+
"step": 653
|
| 4581 |
+
},
|
| 4582 |
+
{
|
| 4583 |
+
"epoch": 1.0262848175755197,
|
| 4584 |
+
"grad_norm": 1.501465916633606,
|
| 4585 |
+
"learning_rate": 2.9154377291267674e-06,
|
| 4586 |
+
"loss": 1.0185,
|
| 4587 |
+
"step": 654
|
| 4588 |
+
},
|
| 4589 |
+
{
|
| 4590 |
+
"epoch": 1.0278540604158493,
|
| 4591 |
+
"grad_norm": 1.373434066772461,
|
| 4592 |
+
"learning_rate": 2.790638039834668e-06,
|
| 4593 |
+
"loss": 0.9125,
|
| 4594 |
+
"step": 655
|
| 4595 |
+
},
|
| 4596 |
+
{
|
| 4597 |
+
"epoch": 1.0294233032561788,
|
| 4598 |
+
"grad_norm": 1.6120808124542236,
|
| 4599 |
+
"learning_rate": 2.6685308098108106e-06,
|
| 4600 |
+
"loss": 1.0961,
|
| 4601 |
+
"step": 656
|
| 4602 |
+
},
|
| 4603 |
+
{
|
| 4604 |
+
"epoch": 1.0309925460965084,
|
| 4605 |
+
"grad_norm": 1.455134630203247,
|
| 4606 |
+
"learning_rate": 2.5491193866866025e-06,
|
| 4607 |
+
"loss": 0.8963,
|
| 4608 |
+
"step": 657
|
| 4609 |
+
},
|
| 4610 |
+
{
|
| 4611 |
+
"epoch": 1.032561788936838,
|
| 4612 |
+
"grad_norm": 1.4574368000030518,
|
| 4613 |
+
"learning_rate": 2.432407044186509e-06,
|
| 4614 |
+
"loss": 0.9296,
|
| 4615 |
+
"step": 658
|
| 4616 |
+
},
|
| 4617 |
+
{
|
| 4618 |
+
"epoch": 1.0341310317771675,
|
| 4619 |
+
"grad_norm": 1.5060312747955322,
|
| 4620 |
+
"learning_rate": 2.3183969820383735e-06,
|
| 4621 |
+
"loss": 0.9674,
|
| 4622 |
+
"step": 659
|
| 4623 |
+
},
|
| 4624 |
+
{
|
| 4625 |
+
"epoch": 1.035700274617497,
|
| 4626 |
+
"grad_norm": 1.544988751411438,
|
| 4627 |
+
"learning_rate": 2.2070923258856255e-06,
|
| 4628 |
+
"loss": 0.9882,
|
| 4629 |
+
"step": 660
|
| 4630 |
+
},
|
| 4631 |
+
{
|
| 4632 |
+
"epoch": 1.0372695174578266,
|
| 4633 |
+
"grad_norm": 1.4960849285125732,
|
| 4634 |
+
"learning_rate": 2.098496127201648e-06,
|
| 4635 |
+
"loss": 0.966,
|
| 4636 |
+
"step": 661
|
| 4637 |
+
},
|
| 4638 |
+
{
|
| 4639 |
+
"epoch": 1.0388387602981561,
|
| 4640 |
+
"grad_norm": 1.5136550664901733,
|
| 4641 |
+
"learning_rate": 1.992611363206103e-06,
|
| 4642 |
+
"loss": 0.9574,
|
| 4643 |
+
"step": 662
|
| 4644 |
+
},
|
| 4645 |
+
{
|
| 4646 |
+
"epoch": 1.0404080031384857,
|
| 4647 |
+
"grad_norm": 1.4802120923995972,
|
| 4648 |
+
"learning_rate": 1.889440936783242e-06,
|
| 4649 |
+
"loss": 0.8649,
|
| 4650 |
+
"step": 663
|
| 4651 |
+
},
|
| 4652 |
+
{
|
| 4653 |
+
"epoch": 1.0419772459788152,
|
| 4654 |
+
"grad_norm": 1.5333938598632812,
|
| 4655 |
+
"learning_rate": 1.7889876764024505e-06,
|
| 4656 |
+
"loss": 1.0243,
|
| 4657 |
+
"step": 664
|
| 4658 |
+
},
|
| 4659 |
+
{
|
| 4660 |
+
"epoch": 1.0435464888191448,
|
| 4661 |
+
"grad_norm": 1.5968855619430542,
|
| 4662 |
+
"learning_rate": 1.691254336040595e-06,
|
| 4663 |
+
"loss": 0.9513,
|
| 4664 |
+
"step": 665
|
| 4665 |
+
},
|
| 4666 |
+
{
|
| 4667 |
+
"epoch": 1.0451157316594744,
|
| 4668 |
+
"grad_norm": 1.512230396270752,
|
| 4669 |
+
"learning_rate": 1.59624359510657e-06,
|
| 4670 |
+
"loss": 1.037,
|
| 4671 |
+
"step": 666
|
| 4672 |
+
},
|
| 4673 |
+
{
|
| 4674 |
+
"epoch": 1.046684974499804,
|
| 4675 |
+
"grad_norm": 1.5507651567459106,
|
| 4676 |
+
"learning_rate": 1.5039580583678393e-06,
|
| 4677 |
+
"loss": 0.9005,
|
| 4678 |
+
"step": 667
|
| 4679 |
+
},
|
| 4680 |
+
{
|
| 4681 |
+
"epoch": 1.0482542173401335,
|
| 4682 |
+
"grad_norm": 1.6249401569366455,
|
| 4683 |
+
"learning_rate": 1.414400255879008e-06,
|
| 4684 |
+
"loss": 0.9521,
|
| 4685 |
+
"step": 668
|
| 4686 |
+
},
|
| 4687 |
+
{
|
| 4688 |
+
"epoch": 1.049823460180463,
|
| 4689 |
+
"grad_norm": 1.5170681476593018,
|
| 4690 |
+
"learning_rate": 1.327572642912468e-06,
|
| 4691 |
+
"loss": 1.0033,
|
| 4692 |
+
"step": 669
|
| 4693 |
+
},
|
| 4694 |
+
{
|
| 4695 |
+
"epoch": 1.0513927030207926,
|
| 4696 |
+
"grad_norm": 1.505729079246521,
|
| 4697 |
+
"learning_rate": 1.2434775998910964e-06,
|
| 4698 |
+
"loss": 0.9384,
|
| 4699 |
+
"step": 670
|
| 4700 |
+
},
|
| 4701 |
+
{
|
| 4702 |
+
"epoch": 1.052961945861122,
|
| 4703 |
+
"grad_norm": 1.5404661893844604,
|
| 4704 |
+
"learning_rate": 1.1621174323229612e-06,
|
| 4705 |
+
"loss": 1.0742,
|
| 4706 |
+
"step": 671
|
| 4707 |
+
},
|
| 4708 |
+
{
|
| 4709 |
+
"epoch": 1.0545311887014515,
|
| 4710 |
+
"grad_norm": 1.6279956102371216,
|
| 4711 |
+
"learning_rate": 1.0834943707381784e-06,
|
| 4712 |
+
"loss": 0.8588,
|
| 4713 |
+
"step": 672
|
| 4714 |
+
},
|
| 4715 |
+
{
|
| 4716 |
+
"epoch": 1.056100431541781,
|
| 4717 |
+
"grad_norm": 1.5603866577148438,
|
| 4718 |
+
"learning_rate": 1.0076105706276888e-06,
|
| 4719 |
+
"loss": 1.0602,
|
| 4720 |
+
"step": 673
|
| 4721 |
+
},
|
| 4722 |
+
{
|
| 4723 |
+
"epoch": 1.0576696743821106,
|
| 4724 |
+
"grad_norm": 1.5433942079544067,
|
| 4725 |
+
"learning_rate": 9.344681123841967e-07,
|
| 4726 |
+
"loss": 0.9865,
|
| 4727 |
+
"step": 674
|
| 4728 |
+
},
|
| 4729 |
+
{
|
| 4730 |
+
"epoch": 1.0592389172224401,
|
| 4731 |
+
"grad_norm": 1.5365676879882812,
|
| 4732 |
+
"learning_rate": 8.640690012451515e-07,
|
| 4733 |
+
"loss": 0.967,
|
| 4734 |
+
"step": 675
|
| 4735 |
+
},
|
| 4736 |
+
{
|
| 4737 |
+
"epoch": 1.0608081600627697,
|
| 4738 |
+
"grad_norm": 1.5364326238632202,
|
| 4739 |
+
"learning_rate": 7.964151672377458e-07,
|
| 4740 |
+
"loss": 0.9036,
|
| 4741 |
+
"step": 676
|
| 4742 |
+
},
|
| 4743 |
+
{
|
| 4744 |
+
"epoch": 1.0623774029030992,
|
| 4745 |
+
"grad_norm": 1.4506745338439941,
|
| 4746 |
+
"learning_rate": 7.315084651260009e-07,
|
| 4747 |
+
"loss": 0.8917,
|
| 4748 |
+
"step": 677
|
| 4749 |
+
},
|
| 4750 |
+
{
|
| 4751 |
+
"epoch": 1.0639466457434288,
|
| 4752 |
+
"grad_norm": 1.42936372756958,
|
| 4753 |
+
"learning_rate": 6.69350674359959e-07,
|
| 4754 |
+
"loss": 0.8587,
|
| 4755 |
+
"step": 678
|
| 4756 |
+
},
|
| 4757 |
+
{
|
| 4758 |
+
"epoch": 1.0655158885837583,
|
| 4759 |
+
"grad_norm": 1.4997657537460327,
|
| 4760 |
+
"learning_rate": 6.099434990268609e-07,
|
| 4761 |
+
"loss": 1.0052,
|
| 4762 |
+
"step": 679
|
| 4763 |
+
},
|
| 4764 |
+
{
|
| 4765 |
+
"epoch": 1.067085131424088,
|
| 4766 |
+
"grad_norm": 1.4839539527893066,
|
| 4767 |
+
"learning_rate": 5.532885678043977e-07,
|
| 4768 |
+
"loss": 0.8442,
|
| 4769 |
+
"step": 680
|
| 4770 |
+
},
|
| 4771 |
+
{
|
| 4772 |
+
"epoch": 1.0686543742644175,
|
| 4773 |
+
"grad_norm": 1.5665620565414429,
|
| 4774 |
+
"learning_rate": 4.9938743391615e-07,
|
| 4775 |
+
"loss": 0.9402,
|
| 4776 |
+
"step": 681
|
| 4777 |
+
},
|
| 4778 |
+
{
|
| 4779 |
+
"epoch": 1.070223617104747,
|
| 4780 |
+
"grad_norm": 1.7468892335891724,
|
| 4781 |
+
"learning_rate": 4.482415750889204e-07,
|
| 4782 |
+
"loss": 0.9801,
|
| 4783 |
+
"step": 682
|
| 4784 |
+
},
|
| 4785 |
+
{
|
| 4786 |
+
"epoch": 1.0717928599450766,
|
| 4787 |
+
"grad_norm": 1.5470365285873413,
|
| 4788 |
+
"learning_rate": 3.998523935122772e-07,
|
| 4789 |
+
"loss": 0.9294,
|
| 4790 |
+
"step": 683
|
| 4791 |
+
},
|
| 4792 |
+
{
|
| 4793 |
+
"epoch": 1.0733621027854061,
|
| 4794 |
+
"grad_norm": 1.544100046157837,
|
| 4795 |
+
"learning_rate": 3.5422121580005864e-07,
|
| 4796 |
+
"loss": 0.9683,
|
| 4797 |
+
"step": 684
|
| 4798 |
+
},
|
| 4799 |
+
{
|
| 4800 |
+
"epoch": 1.0749313456257357,
|
| 4801 |
+
"grad_norm": 1.5746307373046875,
|
| 4802 |
+
"learning_rate": 3.1134929295407564e-07,
|
| 4803 |
+
"loss": 0.9611,
|
| 4804 |
+
"step": 685
|
| 4805 |
+
},
|
| 4806 |
+
{
|
| 4807 |
+
"epoch": 1.0765005884660652,
|
| 4808 |
+
"grad_norm": 1.612066388130188,
|
| 4809 |
+
"learning_rate": 2.7123780032973235e-07,
|
| 4810 |
+
"loss": 0.9273,
|
| 4811 |
+
"step": 686
|
| 4812 |
+
},
|
| 4813 |
+
{
|
| 4814 |
+
"epoch": 1.0780698313063946,
|
| 4815 |
+
"grad_norm": 1.5221633911132812,
|
| 4816 |
+
"learning_rate": 2.3388783760386601e-07,
|
| 4817 |
+
"loss": 0.9797,
|
| 4818 |
+
"step": 687
|
| 4819 |
+
},
|
| 4820 |
+
{
|
| 4821 |
+
"epoch": 1.079639074146724,
|
| 4822 |
+
"grad_norm": 1.4008647203445435,
|
| 4823 |
+
"learning_rate": 1.9930042874457254e-07,
|
| 4824 |
+
"loss": 0.9276,
|
| 4825 |
+
"step": 688
|
| 4826 |
+
},
|
| 4827 |
+
{
|
| 4828 |
+
"epoch": 1.0812083169870537,
|
| 4829 |
+
"grad_norm": 1.4528234004974365,
|
| 4830 |
+
"learning_rate": 1.6747652198313957e-07,
|
| 4831 |
+
"loss": 0.9828,
|
| 4832 |
+
"step": 689
|
| 4833 |
+
},
|
| 4834 |
+
{
|
| 4835 |
+
"epoch": 1.0827775598273832,
|
| 4836 |
+
"grad_norm": 1.5776127576828003,
|
| 4837 |
+
"learning_rate": 1.3841698978804285e-07,
|
| 4838 |
+
"loss": 0.9532,
|
| 4839 |
+
"step": 690
|
| 4840 |
+
},
|
| 4841 |
+
{
|
| 4842 |
+
"epoch": 1.0843468026677128,
|
| 4843 |
+
"grad_norm": 1.5421435832977295,
|
| 4844 |
+
"learning_rate": 1.1212262884103974e-07,
|
| 4845 |
+
"loss": 0.9135,
|
| 4846 |
+
"step": 691
|
| 4847 |
+
},
|
| 4848 |
+
{
|
| 4849 |
+
"epoch": 1.0859160455080423,
|
| 4850 |
+
"grad_norm": 1.5392076969146729,
|
| 4851 |
+
"learning_rate": 8.85941600153033e-08,
|
| 4852 |
+
"loss": 0.9884,
|
| 4853 |
+
"step": 692
|
| 4854 |
+
},
|
| 4855 |
+
{
|
| 4856 |
+
"epoch": 1.0874852883483719,
|
| 4857 |
+
"grad_norm": 1.4609785079956055,
|
| 4858 |
+
"learning_rate": 6.783222835572055e-08,
|
| 4859 |
+
"loss": 0.933,
|
| 4860 |
+
"step": 693
|
| 4861 |
+
},
|
| 4862 |
+
{
|
| 4863 |
+
"epoch": 1.0890545311887014,
|
| 4864 |
+
"grad_norm": 1.4286479949951172,
|
| 4865 |
+
"learning_rate": 4.98374030611084e-08,
|
| 4866 |
+
"loss": 0.8715,
|
| 4867 |
+
"step": 694
|
| 4868 |
+
},
|
| 4869 |
+
{
|
| 4870 |
+
"epoch": 1.090623774029031,
|
| 4871 |
+
"grad_norm": 1.6319226026535034,
|
| 4872 |
+
"learning_rate": 3.461017746871675e-08,
|
| 4873 |
+
"loss": 1.0219,
|
| 4874 |
+
"step": 695
|
| 4875 |
+
},
|
| 4876 |
+
{
|
| 4877 |
+
"epoch": 1.0921930168693605,
|
| 4878 |
+
"grad_norm": 1.5931543111801147,
|
| 4879 |
+
"learning_rate": 2.215096904060454e-08,
|
| 4880 |
+
"loss": 0.9608,
|
| 4881 |
+
"step": 696
|
| 4882 |
+
},
|
| 4883 |
+
{
|
| 4884 |
+
"epoch": 1.09376225970969,
|
| 4885 |
+
"grad_norm": 1.4178990125656128,
|
| 4886 |
+
"learning_rate": 1.246011935228064e-08,
|
| 4887 |
+
"loss": 0.8892,
|
| 4888 |
+
"step": 697
|
| 4889 |
+
},
|
| 4890 |
+
{
|
| 4891 |
+
"epoch": 1.0953315025500197,
|
| 4892 |
+
"grad_norm": 1.5912450551986694,
|
| 4893 |
+
"learning_rate": 5.537894083273543e-09,
|
| 4894 |
+
"loss": 0.9645,
|
| 4895 |
+
"step": 698
|
| 4896 |
+
},
|
| 4897 |
+
{
|
| 4898 |
+
"epoch": 1.0969007453903492,
|
| 4899 |
+
"grad_norm": 1.4889562129974365,
|
| 4900 |
+
"learning_rate": 1.384483009898796e-09,
|
| 4901 |
+
"loss": 0.87,
|
| 4902 |
+
"step": 699
|
| 4903 |
+
},
|
| 4904 |
+
{
|
| 4905 |
+
"epoch": 1.0984699882306788,
|
| 4906 |
+
"grad_norm": 1.4964066743850708,
|
| 4907 |
+
"learning_rate": 0.0,
|
| 4908 |
+
"loss": 0.9337,
|
| 4909 |
+
"step": 700
|
| 4910 |
}
|
| 4911 |
],
|
| 4912 |
"logging_steps": 1,
|
|
|
|
| 4921 |
"should_evaluate": false,
|
| 4922 |
"should_log": false,
|
| 4923 |
"should_save": true,
|
| 4924 |
+
"should_training_stop": true
|
| 4925 |
},
|
| 4926 |
"attributes": {}
|
| 4927 |
}
|
| 4928 |
},
|
| 4929 |
+
"total_flos": 2.524580473166889e+17,
|
| 4930 |
"train_batch_size": 4,
|
| 4931 |
"trial_name": null,
|
| 4932 |
"trial_params": null
|