Training in progress, step 6600
Browse files- adapter_model.safetensors +1 -1
- last-checkpoint/adapter_config.json +2 -2
- 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/scaler.pt +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +5 -1781
- last-checkpoint/training_args.bin +1 -1
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14764744
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1722a8caea7ddbd73f25eed300ac43beb6708faa977770b5ce63e0420b94092d
|
| 3 |
size 14764744
|
last-checkpoint/adapter_config.json
CHANGED
|
@@ -32,9 +32,9 @@
|
|
| 32 |
"revision": null,
|
| 33 |
"target_modules": [
|
| 34 |
"q_proj",
|
|
|
|
| 35 |
"o_proj",
|
| 36 |
-
"k_proj"
|
| 37 |
-
"v_proj"
|
| 38 |
],
|
| 39 |
"target_parameters": null,
|
| 40 |
"task_type": "CAUSAL_LM",
|
|
|
|
| 32 |
"revision": null,
|
| 33 |
"target_modules": [
|
| 34 |
"q_proj",
|
| 35 |
+
"v_proj",
|
| 36 |
"o_proj",
|
| 37 |
+
"k_proj"
|
|
|
|
| 38 |
],
|
| 39 |
"target_parameters": null,
|
| 40 |
"task_type": "CAUSAL_LM",
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14764744
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:52217ccaabbb2c9a520ff6c890c15d98aa47a4c28fad51432c095d328a2a29be
|
| 3 |
size 14764744
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 29614667
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6af5e95ef3e872afad1bdf1fdf3ea5f5155675c6586239ad660ccd39c37347e
|
| 3 |
size 29614667
|
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 14917
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a82aeb7d859423fae1de432598feb0f08060628848416302157961cd81d904dc
|
| 3 |
size 14917
|
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 14917
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51933555f7d7492da069489fcbf900fea3e97204e288093912746efcc0733e6e
|
| 3 |
size 14917
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d223d319f4b8402bb6d092355111dac11a4bb47d6e23f14a8f4c5bc3c5b9c8f
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dd465d73b04f6c59912627375bc7905b039595a268513d7eb113da343a1d81e2
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -5336,1788 +5336,12 @@
|
|
| 5336 |
"eval_samples_per_second": 5.161,
|
| 5337 |
"eval_steps_per_second": 1.291,
|
| 5338 |
"step": 4800
|
| 5339 |
-
},
|
| 5340 |
-
{
|
| 5341 |
-
"entropy": 0.9813777908682824,
|
| 5342 |
-
"epoch": 1.252946922826441,
|
| 5343 |
-
"grad_norm": 0.44488945603370667,
|
| 5344 |
-
"learning_rate": 7.473300338629852e-06,
|
| 5345 |
-
"loss": 0.9760551452636719,
|
| 5346 |
-
"mean_token_accuracy": 0.7828159511089325,
|
| 5347 |
-
"num_tokens": 129388.0,
|
| 5348 |
-
"step": 4810
|
| 5349 |
-
},
|
| 5350 |
-
{
|
| 5351 |
-
"entropy": 1.028295949101448,
|
| 5352 |
-
"epoch": 1.2555519374796482,
|
| 5353 |
-
"grad_norm": 0.476211279630661,
|
| 5354 |
-
"learning_rate": 7.447251888512634e-06,
|
| 5355 |
-
"loss": 1.026865005493164,
|
| 5356 |
-
"mean_token_accuracy": 0.7739313766360283,
|
| 5357 |
-
"num_tokens": 256578.0,
|
| 5358 |
-
"step": 4820
|
| 5359 |
-
},
|
| 5360 |
-
{
|
| 5361 |
-
"entropy": 0.993380619585514,
|
| 5362 |
-
"epoch": 1.2581569521328557,
|
| 5363 |
-
"grad_norm": 0.5022156834602356,
|
| 5364 |
-
"learning_rate": 7.4212034383954155e-06,
|
| 5365 |
-
"loss": 0.9900455474853516,
|
| 5366 |
-
"mean_token_accuracy": 0.7809839487075806,
|
| 5367 |
-
"num_tokens": 379760.0,
|
| 5368 |
-
"step": 4830
|
| 5369 |
-
},
|
| 5370 |
-
{
|
| 5371 |
-
"entropy": 0.9588794216513634,
|
| 5372 |
-
"epoch": 1.2607619667860632,
|
| 5373 |
-
"grad_norm": 0.43847885727882385,
|
| 5374 |
-
"learning_rate": 7.395154988278198e-06,
|
| 5375 |
-
"loss": 0.9578507423400879,
|
| 5376 |
-
"mean_token_accuracy": 0.7875250384211541,
|
| 5377 |
-
"num_tokens": 506218.0,
|
| 5378 |
-
"step": 4840
|
| 5379 |
-
},
|
| 5380 |
-
{
|
| 5381 |
-
"entropy": 0.981762844324112,
|
| 5382 |
-
"epoch": 1.2633669814392707,
|
| 5383 |
-
"grad_norm": 0.44927820563316345,
|
| 5384 |
-
"learning_rate": 7.36910653816098e-06,
|
| 5385 |
-
"loss": 0.9703155517578125,
|
| 5386 |
-
"mean_token_accuracy": 0.7809026747941971,
|
| 5387 |
-
"num_tokens": 630994.0,
|
| 5388 |
-
"step": 4850
|
| 5389 |
-
},
|
| 5390 |
-
{
|
| 5391 |
-
"entropy": 0.9776830449700356,
|
| 5392 |
-
"epoch": 1.265971996092478,
|
| 5393 |
-
"grad_norm": 0.4348759055137634,
|
| 5394 |
-
"learning_rate": 7.343058088043762e-06,
|
| 5395 |
-
"loss": 0.9691827774047852,
|
| 5396 |
-
"mean_token_accuracy": 0.7856391072273254,
|
| 5397 |
-
"num_tokens": 757625.0,
|
| 5398 |
-
"step": 4860
|
| 5399 |
-
},
|
| 5400 |
-
{
|
| 5401 |
-
"entropy": 0.9467781752347946,
|
| 5402 |
-
"epoch": 1.2685770107456855,
|
| 5403 |
-
"grad_norm": 0.4643247127532959,
|
| 5404 |
-
"learning_rate": 7.3170096379265445e-06,
|
| 5405 |
-
"loss": 0.9435815811157227,
|
| 5406 |
-
"mean_token_accuracy": 0.7907302647829055,
|
| 5407 |
-
"num_tokens": 881836.0,
|
| 5408 |
-
"step": 4870
|
| 5409 |
-
},
|
| 5410 |
-
{
|
| 5411 |
-
"entropy": 0.9590273395180702,
|
| 5412 |
-
"epoch": 1.2711820253988928,
|
| 5413 |
-
"grad_norm": 0.45526987314224243,
|
| 5414 |
-
"learning_rate": 7.290961187809326e-06,
|
| 5415 |
-
"loss": 0.9674400329589844,
|
| 5416 |
-
"mean_token_accuracy": 0.785734324157238,
|
| 5417 |
-
"num_tokens": 1011144.0,
|
| 5418 |
-
"step": 4880
|
| 5419 |
-
},
|
| 5420 |
-
{
|
| 5421 |
-
"entropy": 1.017872877418995,
|
| 5422 |
-
"epoch": 1.2737870400521003,
|
| 5423 |
-
"grad_norm": 0.42657098174095154,
|
| 5424 |
-
"learning_rate": 7.264912737692108e-06,
|
| 5425 |
-
"loss": 1.0152156829833985,
|
| 5426 |
-
"mean_token_accuracy": 0.7740847021341324,
|
| 5427 |
-
"num_tokens": 1136047.0,
|
| 5428 |
-
"step": 4890
|
| 5429 |
-
},
|
| 5430 |
-
{
|
| 5431 |
-
"entropy": 0.993815653026104,
|
| 5432 |
-
"epoch": 1.2763920547053078,
|
| 5433 |
-
"grad_norm": 0.4295169711112976,
|
| 5434 |
-
"learning_rate": 7.23886428757489e-06,
|
| 5435 |
-
"loss": 0.9871231079101562,
|
| 5436 |
-
"mean_token_accuracy": 0.7813043370842934,
|
| 5437 |
-
"num_tokens": 1258878.0,
|
| 5438 |
-
"step": 4900
|
| 5439 |
-
},
|
| 5440 |
-
{
|
| 5441 |
-
"epoch": 1.2763920547053078,
|
| 5442 |
-
"eval_entropy": 0.9970914693472669,
|
| 5443 |
-
"eval_loss": 0.9884027242660522,
|
| 5444 |
-
"eval_mean_token_accuracy": 0.7808390200208133,
|
| 5445 |
-
"eval_num_tokens": 1258878.0,
|
| 5446 |
-
"eval_runtime": 1248.5811,
|
| 5447 |
-
"eval_samples_per_second": 5.451,
|
| 5448 |
-
"eval_steps_per_second": 1.363,
|
| 5449 |
-
"step": 4900
|
| 5450 |
-
},
|
| 5451 |
-
{
|
| 5452 |
-
"entropy": 1.0345384523272514,
|
| 5453 |
-
"epoch": 1.278997069358515,
|
| 5454 |
-
"grad_norm": 0.46928712725639343,
|
| 5455 |
-
"learning_rate": 7.212815837457672e-06,
|
| 5456 |
-
"loss": 1.0300195693969727,
|
| 5457 |
-
"mean_token_accuracy": 0.771189559996128,
|
| 5458 |
-
"num_tokens": 1386649.0,
|
| 5459 |
-
"step": 4910
|
| 5460 |
-
},
|
| 5461 |
-
{
|
| 5462 |
-
"entropy": 1.0029998540878295,
|
| 5463 |
-
"epoch": 1.2816020840117226,
|
| 5464 |
-
"grad_norm": 0.48189452290534973,
|
| 5465 |
-
"learning_rate": 7.186767387340454e-06,
|
| 5466 |
-
"loss": 0.9906700134277344,
|
| 5467 |
-
"mean_token_accuracy": 0.7785976707935334,
|
| 5468 |
-
"num_tokens": 1512338.0,
|
| 5469 |
-
"step": 4920
|
| 5470 |
-
},
|
| 5471 |
-
{
|
| 5472 |
-
"entropy": 1.0033065855503083,
|
| 5473 |
-
"epoch": 1.2842070986649299,
|
| 5474 |
-
"grad_norm": 0.4588858187198639,
|
| 5475 |
-
"learning_rate": 7.160718937223236e-06,
|
| 5476 |
-
"loss": 0.9927248001098633,
|
| 5477 |
-
"mean_token_accuracy": 0.7797123953700066,
|
| 5478 |
-
"num_tokens": 1642624.0,
|
| 5479 |
-
"step": 4930
|
| 5480 |
-
},
|
| 5481 |
-
{
|
| 5482 |
-
"entropy": 0.9549823224544525,
|
| 5483 |
-
"epoch": 1.2868121133181374,
|
| 5484 |
-
"grad_norm": 0.43219390511512756,
|
| 5485 |
-
"learning_rate": 7.134670487106018e-06,
|
| 5486 |
-
"loss": 0.9539813041687012,
|
| 5487 |
-
"mean_token_accuracy": 0.7869110450148582,
|
| 5488 |
-
"num_tokens": 1771847.0,
|
| 5489 |
-
"step": 4940
|
| 5490 |
-
},
|
| 5491 |
-
{
|
| 5492 |
-
"entropy": 1.0237484708428384,
|
| 5493 |
-
"epoch": 1.289417127971345,
|
| 5494 |
-
"grad_norm": 0.4907780885696411,
|
| 5495 |
-
"learning_rate": 7.108622036988799e-06,
|
| 5496 |
-
"loss": 1.0278969764709474,
|
| 5497 |
-
"mean_token_accuracy": 0.7715816020965576,
|
| 5498 |
-
"num_tokens": 1900902.0,
|
| 5499 |
-
"step": 4950
|
| 5500 |
-
},
|
| 5501 |
-
{
|
| 5502 |
-
"entropy": 1.006957694888115,
|
| 5503 |
-
"epoch": 1.2920221426245524,
|
| 5504 |
-
"grad_norm": 0.4545274078845978,
|
| 5505 |
-
"learning_rate": 7.082573586871582e-06,
|
| 5506 |
-
"loss": 1.0091020584106445,
|
| 5507 |
-
"mean_token_accuracy": 0.7763800770044327,
|
| 5508 |
-
"num_tokens": 2031777.0,
|
| 5509 |
-
"step": 4960
|
| 5510 |
-
},
|
| 5511 |
-
{
|
| 5512 |
-
"entropy": 0.9983339875936508,
|
| 5513 |
-
"epoch": 1.2946271572777597,
|
| 5514 |
-
"grad_norm": 0.4333556592464447,
|
| 5515 |
-
"learning_rate": 7.056525136754364e-06,
|
| 5516 |
-
"loss": 0.984560203552246,
|
| 5517 |
-
"mean_token_accuracy": 0.7822976499795914,
|
| 5518 |
-
"num_tokens": 2156527.0,
|
| 5519 |
-
"step": 4970
|
| 5520 |
-
},
|
| 5521 |
-
{
|
| 5522 |
-
"entropy": 0.9773911476135254,
|
| 5523 |
-
"epoch": 1.2972321719309672,
|
| 5524 |
-
"grad_norm": 0.4380306303501129,
|
| 5525 |
-
"learning_rate": 7.030476686637145e-06,
|
| 5526 |
-
"loss": 0.9723411560058594,
|
| 5527 |
-
"mean_token_accuracy": 0.7831587105989456,
|
| 5528 |
-
"num_tokens": 2280664.0,
|
| 5529 |
-
"step": 4980
|
| 5530 |
-
},
|
| 5531 |
-
{
|
| 5532 |
-
"entropy": 1.0176811680197715,
|
| 5533 |
-
"epoch": 1.2998371865841745,
|
| 5534 |
-
"grad_norm": 0.49647819995880127,
|
| 5535 |
-
"learning_rate": 7.004428236519928e-06,
|
| 5536 |
-
"loss": 1.0118232727050782,
|
| 5537 |
-
"mean_token_accuracy": 0.7755638837814331,
|
| 5538 |
-
"num_tokens": 2408726.0,
|
| 5539 |
-
"step": 4990
|
| 5540 |
-
},
|
| 5541 |
-
{
|
| 5542 |
-
"entropy": 1.003194873034954,
|
| 5543 |
-
"epoch": 1.302442201237382,
|
| 5544 |
-
"grad_norm": 0.4423670470714569,
|
| 5545 |
-
"learning_rate": 6.978379786402709e-06,
|
| 5546 |
-
"loss": 0.9979192733764648,
|
| 5547 |
-
"mean_token_accuracy": 0.7790161743760109,
|
| 5548 |
-
"num_tokens": 2538321.0,
|
| 5549 |
-
"step": 5000
|
| 5550 |
-
},
|
| 5551 |
-
{
|
| 5552 |
-
"epoch": 1.302442201237382,
|
| 5553 |
-
"eval_entropy": 0.9890262984991354,
|
| 5554 |
-
"eval_loss": 0.9879328608512878,
|
| 5555 |
-
"eval_mean_token_accuracy": 0.7809134237774671,
|
| 5556 |
-
"eval_num_tokens": 2538321.0,
|
| 5557 |
-
"eval_runtime": 1248.3546,
|
| 5558 |
-
"eval_samples_per_second": 5.452,
|
| 5559 |
-
"eval_steps_per_second": 1.363,
|
| 5560 |
-
"step": 5000
|
| 5561 |
-
},
|
| 5562 |
-
{
|
| 5563 |
-
"entropy": 0.9911179274320603,
|
| 5564 |
-
"epoch": 1.3050472158905895,
|
| 5565 |
-
"grad_norm": 0.4474965035915375,
|
| 5566 |
-
"learning_rate": 6.952331336285491e-06,
|
| 5567 |
-
"loss": 0.9846304893493653,
|
| 5568 |
-
"mean_token_accuracy": 0.7800104588270187,
|
| 5569 |
-
"num_tokens": 2664423.0,
|
| 5570 |
-
"step": 5010
|
| 5571 |
-
},
|
| 5572 |
-
{
|
| 5573 |
-
"entropy": 0.9809451669454574,
|
| 5574 |
-
"epoch": 1.3076522305437968,
|
| 5575 |
-
"grad_norm": 0.4665301442146301,
|
| 5576 |
-
"learning_rate": 6.926282886168274e-06,
|
| 5577 |
-
"loss": 0.9809820175170898,
|
| 5578 |
-
"mean_token_accuracy": 0.7841726258397103,
|
| 5579 |
-
"num_tokens": 2790315.0,
|
| 5580 |
-
"step": 5020
|
| 5581 |
-
},
|
| 5582 |
-
{
|
| 5583 |
-
"entropy": 0.9852373987436295,
|
| 5584 |
-
"epoch": 1.3102572451970043,
|
| 5585 |
-
"grad_norm": 0.429098516702652,
|
| 5586 |
-
"learning_rate": 6.900234436051055e-06,
|
| 5587 |
-
"loss": 0.9816387176513672,
|
| 5588 |
-
"mean_token_accuracy": 0.7819439515471458,
|
| 5589 |
-
"num_tokens": 2915995.0,
|
| 5590 |
-
"step": 5030
|
| 5591 |
-
},
|
| 5592 |
-
{
|
| 5593 |
-
"entropy": 0.986951345205307,
|
| 5594 |
-
"epoch": 1.3128622598502115,
|
| 5595 |
-
"grad_norm": 0.45305219292640686,
|
| 5596 |
-
"learning_rate": 6.874185985933837e-06,
|
| 5597 |
-
"loss": 0.9891517639160157,
|
| 5598 |
-
"mean_token_accuracy": 0.780052900314331,
|
| 5599 |
-
"num_tokens": 3041830.0,
|
| 5600 |
-
"step": 5040
|
| 5601 |
-
},
|
| 5602 |
-
{
|
| 5603 |
-
"entropy": 1.0265516340732574,
|
| 5604 |
-
"epoch": 1.315467274503419,
|
| 5605 |
-
"grad_norm": 0.4666403532028198,
|
| 5606 |
-
"learning_rate": 6.848137535816619e-06,
|
| 5607 |
-
"loss": 1.0243612289428712,
|
| 5608 |
-
"mean_token_accuracy": 0.7726057410240174,
|
| 5609 |
-
"num_tokens": 3167296.0,
|
| 5610 |
-
"step": 5050
|
| 5611 |
-
},
|
| 5612 |
-
{
|
| 5613 |
-
"entropy": 1.0306162551045417,
|
| 5614 |
-
"epoch": 1.3180722891566266,
|
| 5615 |
-
"grad_norm": 0.39735814929008484,
|
| 5616 |
-
"learning_rate": 6.8220890856994014e-06,
|
| 5617 |
-
"loss": 1.0236353874206543,
|
| 5618 |
-
"mean_token_accuracy": 0.7734616458415985,
|
| 5619 |
-
"num_tokens": 3292887.0,
|
| 5620 |
-
"step": 5060
|
| 5621 |
-
},
|
| 5622 |
-
{
|
| 5623 |
-
"entropy": 1.0042320653796195,
|
| 5624 |
-
"epoch": 1.320677303809834,
|
| 5625 |
-
"grad_norm": 0.39348119497299194,
|
| 5626 |
-
"learning_rate": 6.796040635582184e-06,
|
| 5627 |
-
"loss": 1.0001495361328125,
|
| 5628 |
-
"mean_token_accuracy": 0.7790237024426461,
|
| 5629 |
-
"num_tokens": 3418112.0,
|
| 5630 |
-
"step": 5070
|
| 5631 |
-
},
|
| 5632 |
-
{
|
| 5633 |
-
"entropy": 1.0145965054631234,
|
| 5634 |
-
"epoch": 1.3232823184630413,
|
| 5635 |
-
"grad_norm": 0.4478921592235565,
|
| 5636 |
-
"learning_rate": 6.7699921854649655e-06,
|
| 5637 |
-
"loss": 1.0137309074401855,
|
| 5638 |
-
"mean_token_accuracy": 0.7771003857254982,
|
| 5639 |
-
"num_tokens": 3548473.0,
|
| 5640 |
-
"step": 5080
|
| 5641 |
-
},
|
| 5642 |
-
{
|
| 5643 |
-
"entropy": 1.0103399023413657,
|
| 5644 |
-
"epoch": 1.3258873331162488,
|
| 5645 |
-
"grad_norm": 0.5050984621047974,
|
| 5646 |
-
"learning_rate": 6.7439437353477475e-06,
|
| 5647 |
-
"loss": 0.999553108215332,
|
| 5648 |
-
"mean_token_accuracy": 0.7785449594259262,
|
| 5649 |
-
"num_tokens": 3673127.0,
|
| 5650 |
-
"step": 5090
|
| 5651 |
-
},
|
| 5652 |
-
{
|
| 5653 |
-
"entropy": 0.966312225162983,
|
| 5654 |
-
"epoch": 1.3284923477694561,
|
| 5655 |
-
"grad_norm": 0.43490949273109436,
|
| 5656 |
-
"learning_rate": 6.717895285230529e-06,
|
| 5657 |
-
"loss": 0.9707491874694825,
|
| 5658 |
-
"mean_token_accuracy": 0.7834598645567894,
|
| 5659 |
-
"num_tokens": 3802026.0,
|
| 5660 |
-
"step": 5100
|
| 5661 |
-
},
|
| 5662 |
-
{
|
| 5663 |
-
"epoch": 1.3284923477694561,
|
| 5664 |
-
"eval_entropy": 0.9884775571971607,
|
| 5665 |
-
"eval_loss": 0.9877337217330933,
|
| 5666 |
-
"eval_mean_token_accuracy": 0.7809631075057805,
|
| 5667 |
-
"eval_num_tokens": 3802026.0,
|
| 5668 |
-
"eval_runtime": 1249.8046,
|
| 5669 |
-
"eval_samples_per_second": 5.446,
|
| 5670 |
-
"eval_steps_per_second": 1.362,
|
| 5671 |
-
"step": 5100
|
| 5672 |
-
},
|
| 5673 |
-
{
|
| 5674 |
-
"entropy": 0.9976687163114548,
|
| 5675 |
-
"epoch": 1.3310973624226636,
|
| 5676 |
-
"grad_norm": 0.43964383006095886,
|
| 5677 |
-
"learning_rate": 6.691846835113312e-06,
|
| 5678 |
-
"loss": 0.9837272644042969,
|
| 5679 |
-
"mean_token_accuracy": 0.7808806389570236,
|
| 5680 |
-
"num_tokens": 3930388.0,
|
| 5681 |
-
"step": 5110
|
| 5682 |
-
},
|
| 5683 |
-
{
|
| 5684 |
-
"entropy": 0.983721150457859,
|
| 5685 |
-
"epoch": 1.3337023770758711,
|
| 5686 |
-
"grad_norm": 0.46595776081085205,
|
| 5687 |
-
"learning_rate": 6.665798384996093e-06,
|
| 5688 |
-
"loss": 0.9791852951049804,
|
| 5689 |
-
"mean_token_accuracy": 0.7797340914607048,
|
| 5690 |
-
"num_tokens": 4057438.0,
|
| 5691 |
-
"step": 5120
|
| 5692 |
-
},
|
| 5693 |
-
{
|
| 5694 |
-
"entropy": 1.0189708486199378,
|
| 5695 |
-
"epoch": 1.3363073917290784,
|
| 5696 |
-
"grad_norm": 0.4384433329105377,
|
| 5697 |
-
"learning_rate": 6.639749934878875e-06,
|
| 5698 |
-
"loss": 1.0155632019042968,
|
| 5699 |
-
"mean_token_accuracy": 0.7752848476171493,
|
| 5700 |
-
"num_tokens": 4184226.0,
|
| 5701 |
-
"step": 5130
|
| 5702 |
-
},
|
| 5703 |
-
{
|
| 5704 |
-
"entropy": 0.9738220989704132,
|
| 5705 |
-
"epoch": 1.338912406382286,
|
| 5706 |
-
"grad_norm": 0.4892994463443756,
|
| 5707 |
-
"learning_rate": 6.613701484761658e-06,
|
| 5708 |
-
"loss": 0.9773561477661132,
|
| 5709 |
-
"mean_token_accuracy": 0.7810323163866997,
|
| 5710 |
-
"num_tokens": 4315468.0,
|
| 5711 |
-
"step": 5140
|
| 5712 |
-
},
|
| 5713 |
-
{
|
| 5714 |
-
"entropy": 0.9959368094801903,
|
| 5715 |
-
"epoch": 1.3415174210354932,
|
| 5716 |
-
"grad_norm": 0.4051533043384552,
|
| 5717 |
-
"learning_rate": 6.587653034644439e-06,
|
| 5718 |
-
"loss": 0.9895273208618164,
|
| 5719 |
-
"mean_token_accuracy": 0.7815038353204727,
|
| 5720 |
-
"num_tokens": 4439960.0,
|
| 5721 |
-
"step": 5150
|
| 5722 |
-
},
|
| 5723 |
-
{
|
| 5724 |
-
"entropy": 0.9800890773534775,
|
| 5725 |
-
"epoch": 1.3441224356887007,
|
| 5726 |
-
"grad_norm": 0.4586341679096222,
|
| 5727 |
-
"learning_rate": 6.561604584527222e-06,
|
| 5728 |
-
"loss": 0.9737453460693359,
|
| 5729 |
-
"mean_token_accuracy": 0.7814203977584839,
|
| 5730 |
-
"num_tokens": 4566056.0,
|
| 5731 |
-
"step": 5160
|
| 5732 |
-
},
|
| 5733 |
-
{
|
| 5734 |
-
"entropy": 1.0000663980841638,
|
| 5735 |
-
"epoch": 1.3467274503419082,
|
| 5736 |
-
"grad_norm": 0.4102465510368347,
|
| 5737 |
-
"learning_rate": 6.535556134410003e-06,
|
| 5738 |
-
"loss": 0.9939519882202148,
|
| 5739 |
-
"mean_token_accuracy": 0.7790467858314514,
|
| 5740 |
-
"num_tokens": 4694925.0,
|
| 5741 |
-
"step": 5170
|
| 5742 |
-
},
|
| 5743 |
-
{
|
| 5744 |
-
"entropy": 1.0039643123745918,
|
| 5745 |
-
"epoch": 1.3493324649951157,
|
| 5746 |
-
"grad_norm": 0.4231637418270111,
|
| 5747 |
-
"learning_rate": 6.509507684292785e-06,
|
| 5748 |
-
"loss": 1.001540756225586,
|
| 5749 |
-
"mean_token_accuracy": 0.7798496216535569,
|
| 5750 |
-
"num_tokens": 4817681.0,
|
| 5751 |
-
"step": 5180
|
| 5752 |
-
},
|
| 5753 |
-
{
|
| 5754 |
-
"entropy": 0.9824537605047226,
|
| 5755 |
-
"epoch": 1.351937479648323,
|
| 5756 |
-
"grad_norm": 0.4193444550037384,
|
| 5757 |
-
"learning_rate": 6.483459234175568e-06,
|
| 5758 |
-
"loss": 0.9902861595153809,
|
| 5759 |
-
"mean_token_accuracy": 0.7809780746698379,
|
| 5760 |
-
"num_tokens": 4946004.0,
|
| 5761 |
-
"step": 5190
|
| 5762 |
-
},
|
| 5763 |
-
{
|
| 5764 |
-
"entropy": 0.9934640824794769,
|
| 5765 |
-
"epoch": 1.3545424943015305,
|
| 5766 |
-
"grad_norm": 0.5171741843223572,
|
| 5767 |
-
"learning_rate": 6.457410784058349e-06,
|
| 5768 |
-
"loss": 0.9897693634033203,
|
| 5769 |
-
"mean_token_accuracy": 0.7789523780345917,
|
| 5770 |
-
"num_tokens": 5071552.0,
|
| 5771 |
-
"step": 5200
|
| 5772 |
-
},
|
| 5773 |
-
{
|
| 5774 |
-
"epoch": 1.3545424943015305,
|
| 5775 |
-
"eval_entropy": 0.9874165067941686,
|
| 5776 |
-
"eval_loss": 0.9873093962669373,
|
| 5777 |
-
"eval_mean_token_accuracy": 0.7809983336757409,
|
| 5778 |
-
"eval_num_tokens": 5071552.0,
|
| 5779 |
-
"eval_runtime": 1248.5888,
|
| 5780 |
-
"eval_samples_per_second": 5.451,
|
| 5781 |
-
"eval_steps_per_second": 1.363,
|
| 5782 |
-
"step": 5200
|
| 5783 |
-
},
|
| 5784 |
-
{
|
| 5785 |
-
"entropy": 0.9975464969873429,
|
| 5786 |
-
"epoch": 1.3571475089547378,
|
| 5787 |
-
"grad_norm": 0.43779075145721436,
|
| 5788 |
-
"learning_rate": 6.431362333941131e-06,
|
| 5789 |
-
"loss": 0.9847593307495117,
|
| 5790 |
-
"mean_token_accuracy": 0.7821353122591972,
|
| 5791 |
-
"num_tokens": 5199409.0,
|
| 5792 |
-
"step": 5210
|
| 5793 |
-
},
|
| 5794 |
-
{
|
| 5795 |
-
"entropy": 0.9890801250934601,
|
| 5796 |
-
"epoch": 1.3597525236079453,
|
| 5797 |
-
"grad_norm": 0.46792900562286377,
|
| 5798 |
-
"learning_rate": 6.405313883823912e-06,
|
| 5799 |
-
"loss": 0.9886545181274414,
|
| 5800 |
-
"mean_token_accuracy": 0.7808847963809967,
|
| 5801 |
-
"num_tokens": 5327018.0,
|
| 5802 |
-
"step": 5220
|
| 5803 |
-
},
|
| 5804 |
-
{
|
| 5805 |
-
"entropy": 1.0165737330913545,
|
| 5806 |
-
"epoch": 1.3623575382611528,
|
| 5807 |
-
"grad_norm": 0.422018438577652,
|
| 5808 |
-
"learning_rate": 6.379265433706695e-06,
|
| 5809 |
-
"loss": 1.0161245346069336,
|
| 5810 |
-
"mean_token_accuracy": 0.7733713835477829,
|
| 5811 |
-
"num_tokens": 5453326.0,
|
| 5812 |
-
"step": 5230
|
| 5813 |
-
},
|
| 5814 |
-
{
|
| 5815 |
-
"entropy": 0.9966152131557464,
|
| 5816 |
-
"epoch": 1.36496255291436,
|
| 5817 |
-
"grad_norm": 0.46802228689193726,
|
| 5818 |
-
"learning_rate": 6.353216983589477e-06,
|
| 5819 |
-
"loss": 0.9952875137329101,
|
| 5820 |
-
"mean_token_accuracy": 0.7801799476146698,
|
| 5821 |
-
"num_tokens": 5579340.0,
|
| 5822 |
-
"step": 5240
|
| 5823 |
-
},
|
| 5824 |
-
{
|
| 5825 |
-
"entropy": 1.013704551756382,
|
| 5826 |
-
"epoch": 1.3675675675675676,
|
| 5827 |
-
"grad_norm": 0.4642789363861084,
|
| 5828 |
-
"learning_rate": 6.327168533472258e-06,
|
| 5829 |
-
"loss": 1.0044668197631836,
|
| 5830 |
-
"mean_token_accuracy": 0.7750307902693748,
|
| 5831 |
-
"num_tokens": 5704206.0,
|
| 5832 |
-
"step": 5250
|
| 5833 |
-
},
|
| 5834 |
-
{
|
| 5835 |
-
"entropy": 0.9664336591959,
|
| 5836 |
-
"epoch": 1.3701725822207749,
|
| 5837 |
-
"grad_norm": 0.45772185921669006,
|
| 5838 |
-
"learning_rate": 6.301120083355041e-06,
|
| 5839 |
-
"loss": 0.9622204780578614,
|
| 5840 |
-
"mean_token_accuracy": 0.7862247511744499,
|
| 5841 |
-
"num_tokens": 5832067.0,
|
| 5842 |
-
"step": 5260
|
| 5843 |
-
},
|
| 5844 |
-
{
|
| 5845 |
-
"entropy": 0.9767679080367089,
|
| 5846 |
-
"epoch": 1.3727775968739824,
|
| 5847 |
-
"grad_norm": 0.4938032627105713,
|
| 5848 |
-
"learning_rate": 6.2750716332378225e-06,
|
| 5849 |
-
"loss": 0.9772611618041992,
|
| 5850 |
-
"mean_token_accuracy": 0.7834762305021286,
|
| 5851 |
-
"num_tokens": 5956416.0,
|
| 5852 |
-
"step": 5270
|
| 5853 |
-
},
|
| 5854 |
-
{
|
| 5855 |
-
"entropy": 0.997693844139576,
|
| 5856 |
-
"epoch": 1.3753826115271899,
|
| 5857 |
-
"grad_norm": 0.4085782766342163,
|
| 5858 |
-
"learning_rate": 6.249023183120605e-06,
|
| 5859 |
-
"loss": 0.993937873840332,
|
| 5860 |
-
"mean_token_accuracy": 0.7790059551596642,
|
| 5861 |
-
"num_tokens": 6080387.0,
|
| 5862 |
-
"step": 5280
|
| 5863 |
-
},
|
| 5864 |
-
{
|
| 5865 |
-
"entropy": 0.9551204591989517,
|
| 5866 |
-
"epoch": 1.3779876261803974,
|
| 5867 |
-
"grad_norm": 0.47124868631362915,
|
| 5868 |
-
"learning_rate": 6.2229747330033865e-06,
|
| 5869 |
-
"loss": 0.9600910186767578,
|
| 5870 |
-
"mean_token_accuracy": 0.7857999742031098,
|
| 5871 |
-
"num_tokens": 6210495.0,
|
| 5872 |
-
"step": 5290
|
| 5873 |
-
},
|
| 5874 |
-
{
|
| 5875 |
-
"entropy": 0.9895543649792671,
|
| 5876 |
-
"epoch": 1.3805926408336047,
|
| 5877 |
-
"grad_norm": 0.4795711934566498,
|
| 5878 |
-
"learning_rate": 6.196926282886169e-06,
|
| 5879 |
-
"loss": 0.980061149597168,
|
| 5880 |
-
"mean_token_accuracy": 0.7814426705241203,
|
| 5881 |
-
"num_tokens": 6338971.0,
|
| 5882 |
-
"step": 5300
|
| 5883 |
-
},
|
| 5884 |
-
{
|
| 5885 |
-
"epoch": 1.3805926408336047,
|
| 5886 |
-
"eval_entropy": 0.9903583783310532,
|
| 5887 |
-
"eval_loss": 0.9870790839195251,
|
| 5888 |
-
"eval_mean_token_accuracy": 0.781017405095868,
|
| 5889 |
-
"eval_num_tokens": 6338971.0,
|
| 5890 |
-
"eval_runtime": 1249.0996,
|
| 5891 |
-
"eval_samples_per_second": 5.449,
|
| 5892 |
-
"eval_steps_per_second": 1.363,
|
| 5893 |
-
"step": 5300
|
| 5894 |
-
},
|
| 5895 |
-
{
|
| 5896 |
-
"entropy": 1.0024344369769096,
|
| 5897 |
-
"epoch": 1.3831976554868122,
|
| 5898 |
-
"grad_norm": 0.44125476479530334,
|
| 5899 |
-
"learning_rate": 6.1708778327689515e-06,
|
| 5900 |
-
"loss": 1.0059074401855468,
|
| 5901 |
-
"mean_token_accuracy": 0.7774184092879295,
|
| 5902 |
-
"num_tokens": 6465148.0,
|
| 5903 |
-
"step": 5310
|
| 5904 |
-
},
|
| 5905 |
-
{
|
| 5906 |
-
"entropy": 0.9910042881965637,
|
| 5907 |
-
"epoch": 1.3858026701400195,
|
| 5908 |
-
"grad_norm": 0.4707449972629547,
|
| 5909 |
-
"learning_rate": 6.144829382651733e-06,
|
| 5910 |
-
"loss": 0.987660026550293,
|
| 5911 |
-
"mean_token_accuracy": 0.7800036445260048,
|
| 5912 |
-
"num_tokens": 6592528.0,
|
| 5913 |
-
"step": 5320
|
| 5914 |
-
},
|
| 5915 |
-
{
|
| 5916 |
-
"entropy": 0.9776637852191925,
|
| 5917 |
-
"epoch": 1.388407684793227,
|
| 5918 |
-
"grad_norm": 0.40415525436401367,
|
| 5919 |
-
"learning_rate": 6.118780932534515e-06,
|
| 5920 |
-
"loss": 0.9776216506958008,
|
| 5921 |
-
"mean_token_accuracy": 0.7837252721190453,
|
| 5922 |
-
"num_tokens": 6718090.0,
|
| 5923 |
-
"step": 5330
|
| 5924 |
-
},
|
| 5925 |
-
{
|
| 5926 |
-
"entropy": 0.9891742005944252,
|
| 5927 |
-
"epoch": 1.3910126994464345,
|
| 5928 |
-
"grad_norm": 0.4758605659008026,
|
| 5929 |
-
"learning_rate": 6.092732482417296e-06,
|
| 5930 |
-
"loss": 0.9910886764526368,
|
| 5931 |
-
"mean_token_accuracy": 0.7781824618577957,
|
| 5932 |
-
"num_tokens": 6845948.0,
|
| 5933 |
-
"step": 5340
|
| 5934 |
-
},
|
| 5935 |
-
{
|
| 5936 |
-
"entropy": 1.007858295738697,
|
| 5937 |
-
"epoch": 1.3936177140996417,
|
| 5938 |
-
"grad_norm": 0.47813597321510315,
|
| 5939 |
-
"learning_rate": 6.066684032300079e-06,
|
| 5940 |
-
"loss": 1.0086344718933105,
|
| 5941 |
-
"mean_token_accuracy": 0.7754964306950569,
|
| 5942 |
-
"num_tokens": 6973246.0,
|
| 5943 |
-
"step": 5350
|
| 5944 |
-
},
|
| 5945 |
-
{
|
| 5946 |
-
"entropy": 1.0145272225141526,
|
| 5947 |
-
"epoch": 1.3962227287528493,
|
| 5948 |
-
"grad_norm": 0.4583241939544678,
|
| 5949 |
-
"learning_rate": 6.040635582182861e-06,
|
| 5950 |
-
"loss": 1.0088977813720703,
|
| 5951 |
-
"mean_token_accuracy": 0.775096683204174,
|
| 5952 |
-
"num_tokens": 7099698.0,
|
| 5953 |
-
"step": 5360
|
| 5954 |
-
},
|
| 5955 |
-
{
|
| 5956 |
-
"entropy": 1.0434660747647286,
|
| 5957 |
-
"epoch": 1.3988277434060565,
|
| 5958 |
-
"grad_norm": 0.4353461265563965,
|
| 5959 |
-
"learning_rate": 6.014587132065642e-06,
|
| 5960 |
-
"loss": 1.0382314682006837,
|
| 5961 |
-
"mean_token_accuracy": 0.7711483478546143,
|
| 5962 |
-
"num_tokens": 7223734.0,
|
| 5963 |
-
"step": 5370
|
| 5964 |
-
},
|
| 5965 |
-
{
|
| 5966 |
-
"entropy": 0.990787972509861,
|
| 5967 |
-
"epoch": 1.401432758059264,
|
| 5968 |
-
"grad_norm": 0.47876155376434326,
|
| 5969 |
-
"learning_rate": 5.988538681948425e-06,
|
| 5970 |
-
"loss": 0.989743995666504,
|
| 5971 |
-
"mean_token_accuracy": 0.7813950344920159,
|
| 5972 |
-
"num_tokens": 7350371.0,
|
| 5973 |
-
"step": 5380
|
| 5974 |
-
},
|
| 5975 |
-
{
|
| 5976 |
-
"entropy": 1.0080674409866333,
|
| 5977 |
-
"epoch": 1.4040377727124715,
|
| 5978 |
-
"grad_norm": 0.5123191475868225,
|
| 5979 |
-
"learning_rate": 5.962490231831206e-06,
|
| 5980 |
-
"loss": 1.002880859375,
|
| 5981 |
-
"mean_token_accuracy": 0.7779936477541923,
|
| 5982 |
-
"num_tokens": 7477903.0,
|
| 5983 |
-
"step": 5390
|
| 5984 |
-
},
|
| 5985 |
-
{
|
| 5986 |
-
"entropy": 1.0048462897539139,
|
| 5987 |
-
"epoch": 1.406642787365679,
|
| 5988 |
-
"grad_norm": 0.4499814212322235,
|
| 5989 |
-
"learning_rate": 5.936441781713989e-06,
|
| 5990 |
-
"loss": 0.999144458770752,
|
| 5991 |
-
"mean_token_accuracy": 0.7778741553425789,
|
| 5992 |
-
"num_tokens": 7604005.0,
|
| 5993 |
-
"step": 5400
|
| 5994 |
-
},
|
| 5995 |
-
{
|
| 5996 |
-
"epoch": 1.406642787365679,
|
| 5997 |
-
"eval_entropy": 0.9904807468068586,
|
| 5998 |
-
"eval_loss": 0.9868780970573425,
|
| 5999 |
-
"eval_mean_token_accuracy": 0.7810666585290754,
|
| 6000 |
-
"eval_num_tokens": 7604005.0,
|
| 6001 |
-
"eval_runtime": 1248.7827,
|
| 6002 |
-
"eval_samples_per_second": 5.45,
|
| 6003 |
-
"eval_steps_per_second": 1.363,
|
| 6004 |
-
"step": 5400
|
| 6005 |
-
},
|
| 6006 |
-
{
|
| 6007 |
-
"entropy": 0.9735529109835624,
|
| 6008 |
-
"epoch": 1.4092478020188863,
|
| 6009 |
-
"grad_norm": 0.4465472996234894,
|
| 6010 |
-
"learning_rate": 5.910393331596771e-06,
|
| 6011 |
-
"loss": 0.9568084716796875,
|
| 6012 |
-
"mean_token_accuracy": 0.7862212345004082,
|
| 6013 |
-
"num_tokens": 7731776.0,
|
| 6014 |
-
"step": 5410
|
| 6015 |
-
},
|
| 6016 |
-
{
|
| 6017 |
-
"entropy": 0.9851446762681008,
|
| 6018 |
-
"epoch": 1.4118528166720938,
|
| 6019 |
-
"grad_norm": 0.45277079939842224,
|
| 6020 |
-
"learning_rate": 5.884344881479552e-06,
|
| 6021 |
-
"loss": 0.9776866912841797,
|
| 6022 |
-
"mean_token_accuracy": 0.7816747695207595,
|
| 6023 |
-
"num_tokens": 7858032.0,
|
| 6024 |
-
"step": 5420
|
| 6025 |
-
},
|
| 6026 |
-
{
|
| 6027 |
-
"entropy": 1.015771795809269,
|
| 6028 |
-
"epoch": 1.4144578313253011,
|
| 6029 |
-
"grad_norm": 0.4632008969783783,
|
| 6030 |
-
"learning_rate": 5.858296431362335e-06,
|
| 6031 |
-
"loss": 1.0171214103698731,
|
| 6032 |
-
"mean_token_accuracy": 0.7743904694914818,
|
| 6033 |
-
"num_tokens": 7985948.0,
|
| 6034 |
-
"step": 5430
|
| 6035 |
-
},
|
| 6036 |
-
{
|
| 6037 |
-
"entropy": 0.9892556384205818,
|
| 6038 |
-
"epoch": 1.4170628459785086,
|
| 6039 |
-
"grad_norm": 0.47347593307495117,
|
| 6040 |
-
"learning_rate": 5.832247981245116e-06,
|
| 6041 |
-
"loss": 0.9954731941223145,
|
| 6042 |
-
"mean_token_accuracy": 0.7805924400687217,
|
| 6043 |
-
"num_tokens": 8112546.0,
|
| 6044 |
-
"step": 5440
|
| 6045 |
-
},
|
| 6046 |
-
{
|
| 6047 |
-
"entropy": 1.0147403255105019,
|
| 6048 |
-
"epoch": 1.4196678606317161,
|
| 6049 |
-
"grad_norm": 0.4375954866409302,
|
| 6050 |
-
"learning_rate": 5.806199531127898e-06,
|
| 6051 |
-
"loss": 1.020146942138672,
|
| 6052 |
-
"mean_token_accuracy": 0.7747515097260476,
|
| 6053 |
-
"num_tokens": 8239855.0,
|
| 6054 |
-
"step": 5450
|
| 6055 |
-
},
|
| 6056 |
-
{
|
| 6057 |
-
"entropy": 1.010063646733761,
|
| 6058 |
-
"epoch": 1.4222728752849234,
|
| 6059 |
-
"grad_norm": 0.5044981241226196,
|
| 6060 |
-
"learning_rate": 5.7801510810106794e-06,
|
| 6061 |
-
"loss": 1.0016718864440919,
|
| 6062 |
-
"mean_token_accuracy": 0.7766346171498298,
|
| 6063 |
-
"num_tokens": 8367673.0,
|
| 6064 |
-
"step": 5460
|
| 6065 |
-
},
|
| 6066 |
-
{
|
| 6067 |
-
"entropy": 1.0042886927723884,
|
| 6068 |
-
"epoch": 1.424877889938131,
|
| 6069 |
-
"grad_norm": 0.4450787305831909,
|
| 6070 |
-
"learning_rate": 5.754102630893462e-06,
|
| 6071 |
-
"loss": 0.9996944427490234,
|
| 6072 |
-
"mean_token_accuracy": 0.7757628753781318,
|
| 6073 |
-
"num_tokens": 8498671.0,
|
| 6074 |
-
"step": 5470
|
| 6075 |
-
},
|
| 6076 |
-
{
|
| 6077 |
-
"entropy": 1.0018585577607155,
|
| 6078 |
-
"epoch": 1.4274829045913382,
|
| 6079 |
-
"grad_norm": 0.49529793858528137,
|
| 6080 |
-
"learning_rate": 5.728054180776244e-06,
|
| 6081 |
-
"loss": 0.9987139701843262,
|
| 6082 |
-
"mean_token_accuracy": 0.7768079906702041,
|
| 6083 |
-
"num_tokens": 8622797.0,
|
| 6084 |
-
"step": 5480
|
| 6085 |
-
},
|
| 6086 |
-
{
|
| 6087 |
-
"entropy": 1.0178197637200355,
|
| 6088 |
-
"epoch": 1.4300879192445457,
|
| 6089 |
-
"grad_norm": 0.43215155601501465,
|
| 6090 |
-
"learning_rate": 5.7020057306590255e-06,
|
| 6091 |
-
"loss": 1.0102587699890138,
|
| 6092 |
-
"mean_token_accuracy": 0.7749877154827118,
|
| 6093 |
-
"num_tokens": 8751670.0,
|
| 6094 |
-
"step": 5490
|
| 6095 |
-
},
|
| 6096 |
-
{
|
| 6097 |
-
"entropy": 0.9857858672738076,
|
| 6098 |
-
"epoch": 1.4326929338977532,
|
| 6099 |
-
"grad_norm": 0.46067577600479126,
|
| 6100 |
-
"learning_rate": 5.6759572805418084e-06,
|
| 6101 |
-
"loss": 0.9797880172729492,
|
| 6102 |
-
"mean_token_accuracy": 0.7822221577167511,
|
| 6103 |
-
"num_tokens": 8876670.0,
|
| 6104 |
-
"step": 5500
|
| 6105 |
-
},
|
| 6106 |
-
{
|
| 6107 |
-
"epoch": 1.4326929338977532,
|
| 6108 |
-
"eval_entropy": 0.9837744325994745,
|
| 6109 |
-
"eval_loss": 0.9865692257881165,
|
| 6110 |
-
"eval_mean_token_accuracy": 0.7811153954120417,
|
| 6111 |
-
"eval_num_tokens": 8876670.0,
|
| 6112 |
-
"eval_runtime": 1255.2558,
|
| 6113 |
-
"eval_samples_per_second": 5.422,
|
| 6114 |
-
"eval_steps_per_second": 1.356,
|
| 6115 |
-
"step": 5500
|
| 6116 |
-
},
|
| 6117 |
-
{
|
| 6118 |
-
"entropy": 0.9540723130106926,
|
| 6119 |
-
"epoch": 1.4352979485509607,
|
| 6120 |
-
"grad_norm": 0.5174602270126343,
|
| 6121 |
-
"learning_rate": 5.64990883042459e-06,
|
| 6122 |
-
"loss": 0.9584260940551758,
|
| 6123 |
-
"mean_token_accuracy": 0.7880000397562981,
|
| 6124 |
-
"num_tokens": 9002387.0,
|
| 6125 |
-
"step": 5510
|
| 6126 |
-
},
|
| 6127 |
-
{
|
| 6128 |
-
"entropy": 0.981420585513115,
|
| 6129 |
-
"epoch": 1.437902963204168,
|
| 6130 |
-
"grad_norm": 0.45851239562034607,
|
| 6131 |
-
"learning_rate": 5.6238603803073725e-06,
|
| 6132 |
-
"loss": 0.9809579849243164,
|
| 6133 |
-
"mean_token_accuracy": 0.7810369253158569,
|
| 6134 |
-
"num_tokens": 9130306.0,
|
| 6135 |
-
"step": 5520
|
| 6136 |
-
},
|
| 6137 |
-
{
|
| 6138 |
-
"entropy": 1.0064574167132379,
|
| 6139 |
-
"epoch": 1.4405079778573755,
|
| 6140 |
-
"grad_norm": 0.43128204345703125,
|
| 6141 |
-
"learning_rate": 5.5978119301901545e-06,
|
| 6142 |
-
"loss": 1.0060470581054688,
|
| 6143 |
-
"mean_token_accuracy": 0.7766813963651658,
|
| 6144 |
-
"num_tokens": 9253749.0,
|
| 6145 |
-
"step": 5530
|
| 6146 |
-
},
|
| 6147 |
-
{
|
| 6148 |
-
"entropy": 0.9711885288357734,
|
| 6149 |
-
"epoch": 1.4431129925105828,
|
| 6150 |
-
"grad_norm": 0.46980082988739014,
|
| 6151 |
-
"learning_rate": 5.571763480072936e-06,
|
| 6152 |
-
"loss": 0.9691987991333008,
|
| 6153 |
-
"mean_token_accuracy": 0.7856941938400268,
|
| 6154 |
-
"num_tokens": 9380142.0,
|
| 6155 |
-
"step": 5540
|
| 6156 |
-
},
|
| 6157 |
-
{
|
| 6158 |
-
"entropy": 0.9889684185385704,
|
| 6159 |
-
"epoch": 1.4457180071637903,
|
| 6160 |
-
"grad_norm": 0.46191513538360596,
|
| 6161 |
-
"learning_rate": 5.545715029955719e-06,
|
| 6162 |
-
"loss": 0.9874019622802734,
|
| 6163 |
-
"mean_token_accuracy": 0.7812291026115418,
|
| 6164 |
-
"num_tokens": 9510005.0,
|
| 6165 |
-
"step": 5550
|
| 6166 |
-
},
|
| 6167 |
-
{
|
| 6168 |
-
"entropy": 1.03478996604681,
|
| 6169 |
-
"epoch": 1.4483230218169978,
|
| 6170 |
-
"grad_norm": 0.4332844316959381,
|
| 6171 |
-
"learning_rate": 5.5196665798385e-06,
|
| 6172 |
-
"loss": 1.0199315071105957,
|
| 6173 |
-
"mean_token_accuracy": 0.77467340528965,
|
| 6174 |
-
"num_tokens": 9635497.0,
|
| 6175 |
-
"step": 5560
|
| 6176 |
-
},
|
| 6177 |
-
{
|
| 6178 |
-
"entropy": 1.011206428706646,
|
| 6179 |
-
"epoch": 1.450928036470205,
|
| 6180 |
-
"grad_norm": 0.48398062586784363,
|
| 6181 |
-
"learning_rate": 5.493618129721282e-06,
|
| 6182 |
-
"loss": 1.0067708969116211,
|
| 6183 |
-
"mean_token_accuracy": 0.7770825013518333,
|
| 6184 |
-
"num_tokens": 9760444.0,
|
| 6185 |
-
"step": 5570
|
| 6186 |
-
},
|
| 6187 |
-
{
|
| 6188 |
-
"entropy": 0.9973620474338531,
|
| 6189 |
-
"epoch": 1.4535330511234126,
|
| 6190 |
-
"grad_norm": 0.43173497915267944,
|
| 6191 |
-
"learning_rate": 5.467569679604065e-06,
|
| 6192 |
-
"loss": 0.9991565704345703,
|
| 6193 |
-
"mean_token_accuracy": 0.7782484918832779,
|
| 6194 |
-
"num_tokens": 9884456.0,
|
| 6195 |
-
"step": 5580
|
| 6196 |
-
},
|
| 6197 |
-
{
|
| 6198 |
-
"entropy": 1.021404606103897,
|
| 6199 |
-
"epoch": 1.4561380657766199,
|
| 6200 |
-
"grad_norm": 0.4523032605648041,
|
| 6201 |
-
"learning_rate": 5.441521229486846e-06,
|
| 6202 |
-
"loss": 1.0067808151245117,
|
| 6203 |
-
"mean_token_accuracy": 0.7764851897954941,
|
| 6204 |
-
"num_tokens": 10012245.0,
|
| 6205 |
-
"step": 5590
|
| 6206 |
-
},
|
| 6207 |
-
{
|
| 6208 |
-
"entropy": 0.9751285776495934,
|
| 6209 |
-
"epoch": 1.4587430804298274,
|
| 6210 |
-
"grad_norm": 0.4640898108482361,
|
| 6211 |
-
"learning_rate": 5.415472779369628e-06,
|
| 6212 |
-
"loss": 0.9704513549804688,
|
| 6213 |
-
"mean_token_accuracy": 0.7840812146663666,
|
| 6214 |
-
"num_tokens": 10138335.0,
|
| 6215 |
-
"step": 5600
|
| 6216 |
-
},
|
| 6217 |
-
{
|
| 6218 |
-
"epoch": 1.4587430804298274,
|
| 6219 |
-
"eval_entropy": 0.986843701815353,
|
| 6220 |
-
"eval_loss": 0.9863292574882507,
|
| 6221 |
-
"eval_mean_token_accuracy": 0.7811444691009163,
|
| 6222 |
-
"eval_num_tokens": 10138335.0,
|
| 6223 |
-
"eval_runtime": 1254.3085,
|
| 6224 |
-
"eval_samples_per_second": 5.426,
|
| 6225 |
-
"eval_steps_per_second": 1.357,
|
| 6226 |
-
"step": 5600
|
| 6227 |
-
},
|
| 6228 |
-
{
|
| 6229 |
-
"entropy": 1.0088901787996292,
|
| 6230 |
-
"epoch": 1.4613480950830349,
|
| 6231 |
-
"grad_norm": 0.4137093722820282,
|
| 6232 |
-
"learning_rate": 5.38942432925241e-06,
|
| 6233 |
-
"loss": 1.004704475402832,
|
| 6234 |
-
"mean_token_accuracy": 0.7777531415224075,
|
| 6235 |
-
"num_tokens": 10264667.0,
|
| 6236 |
-
"step": 5610
|
| 6237 |
-
},
|
| 6238 |
-
{
|
| 6239 |
-
"entropy": 0.9863568991422653,
|
| 6240 |
-
"epoch": 1.4639531097362424,
|
| 6241 |
-
"grad_norm": 0.495769739151001,
|
| 6242 |
-
"learning_rate": 5.363375879135192e-06,
|
| 6243 |
-
"loss": 0.9836648941040039,
|
| 6244 |
-
"mean_token_accuracy": 0.7818165495991707,
|
| 6245 |
-
"num_tokens": 10394357.0,
|
| 6246 |
-
"step": 5620
|
| 6247 |
-
},
|
| 6248 |
-
{
|
| 6249 |
-
"entropy": 1.0240776732563972,
|
| 6250 |
-
"epoch": 1.4665581243894497,
|
| 6251 |
-
"grad_norm": 0.4629792273044586,
|
| 6252 |
-
"learning_rate": 5.337327429017974e-06,
|
| 6253 |
-
"loss": 1.0142351150512696,
|
| 6254 |
-
"mean_token_accuracy": 0.7750792697072029,
|
| 6255 |
-
"num_tokens": 10518766.0,
|
| 6256 |
-
"step": 5630
|
| 6257 |
-
},
|
| 6258 |
-
{
|
| 6259 |
-
"entropy": 0.9978538602590561,
|
| 6260 |
-
"epoch": 1.4691631390426572,
|
| 6261 |
-
"grad_norm": 0.42182034254074097,
|
| 6262 |
-
"learning_rate": 5.311278978900756e-06,
|
| 6263 |
-
"loss": 0.9932327270507812,
|
| 6264 |
-
"mean_token_accuracy": 0.7780352964997291,
|
| 6265 |
-
"num_tokens": 10646092.0,
|
| 6266 |
-
"step": 5640
|
| 6267 |
-
},
|
| 6268 |
-
{
|
| 6269 |
-
"entropy": 0.970037293434143,
|
| 6270 |
-
"epoch": 1.4717681536958644,
|
| 6271 |
-
"grad_norm": 0.4570915699005127,
|
| 6272 |
-
"learning_rate": 5.285230528783538e-06,
|
| 6273 |
-
"loss": 0.9659971237182617,
|
| 6274 |
-
"mean_token_accuracy": 0.7846699148416519,
|
| 6275 |
-
"num_tokens": 10772533.0,
|
| 6276 |
-
"step": 5650
|
| 6277 |
-
},
|
| 6278 |
-
{
|
| 6279 |
-
"entropy": 0.9868507131934166,
|
| 6280 |
-
"epoch": 1.474373168349072,
|
| 6281 |
-
"grad_norm": 0.4327319860458374,
|
| 6282 |
-
"learning_rate": 5.259182078666319e-06,
|
| 6283 |
-
"loss": 0.9883245468139649,
|
| 6284 |
-
"mean_token_accuracy": 0.7823428332805633,
|
| 6285 |
-
"num_tokens": 10903811.0,
|
| 6286 |
-
"step": 5660
|
| 6287 |
-
},
|
| 6288 |
-
{
|
| 6289 |
-
"entropy": 0.9787280559539795,
|
| 6290 |
-
"epoch": 1.4769781830022795,
|
| 6291 |
-
"grad_norm": 0.45937690138816833,
|
| 6292 |
-
"learning_rate": 5.233133628549102e-06,
|
| 6293 |
-
"loss": 0.9710836410522461,
|
| 6294 |
-
"mean_token_accuracy": 0.7832621559500694,
|
| 6295 |
-
"num_tokens": 11030450.0,
|
| 6296 |
-
"step": 5670
|
| 6297 |
-
},
|
| 6298 |
-
{
|
| 6299 |
-
"entropy": 0.9945630744099617,
|
| 6300 |
-
"epoch": 1.4795831976554867,
|
| 6301 |
-
"grad_norm": 0.44801393151283264,
|
| 6302 |
-
"learning_rate": 5.207085178431883e-06,
|
| 6303 |
-
"loss": 0.9887495040893555,
|
| 6304 |
-
"mean_token_accuracy": 0.7813878715038299,
|
| 6305 |
-
"num_tokens": 11153343.0,
|
| 6306 |
-
"step": 5680
|
| 6307 |
-
},
|
| 6308 |
-
{
|
| 6309 |
-
"entropy": 0.9794335409998893,
|
| 6310 |
-
"epoch": 1.4821882123086942,
|
| 6311 |
-
"grad_norm": 0.4029315710067749,
|
| 6312 |
-
"learning_rate": 5.181036728314665e-06,
|
| 6313 |
-
"loss": 0.9747322082519532,
|
| 6314 |
-
"mean_token_accuracy": 0.7834831207990647,
|
| 6315 |
-
"num_tokens": 11284081.0,
|
| 6316 |
-
"step": 5690
|
| 6317 |
-
},
|
| 6318 |
-
{
|
| 6319 |
-
"entropy": 1.0130935430526733,
|
| 6320 |
-
"epoch": 1.4847932269619015,
|
| 6321 |
-
"grad_norm": 0.43846866488456726,
|
| 6322 |
-
"learning_rate": 5.154988278197448e-06,
|
| 6323 |
-
"loss": 1.005000114440918,
|
| 6324 |
-
"mean_token_accuracy": 0.7774733066558838,
|
| 6325 |
-
"num_tokens": 11412859.0,
|
| 6326 |
-
"step": 5700
|
| 6327 |
-
},
|
| 6328 |
-
{
|
| 6329 |
-
"epoch": 1.4847932269619015,
|
| 6330 |
-
"eval_entropy": 0.9864431050563391,
|
| 6331 |
-
"eval_loss": 0.9862220287322998,
|
| 6332 |
-
"eval_mean_token_accuracy": 0.7811710470291479,
|
| 6333 |
-
"eval_num_tokens": 11412859.0,
|
| 6334 |
-
"eval_runtime": 1255.1222,
|
| 6335 |
-
"eval_samples_per_second": 5.423,
|
| 6336 |
-
"eval_steps_per_second": 1.356,
|
| 6337 |
-
"step": 5700
|
| 6338 |
-
},
|
| 6339 |
-
{
|
| 6340 |
-
"entropy": 1.0116615876555444,
|
| 6341 |
-
"epoch": 1.487398241615109,
|
| 6342 |
-
"grad_norm": 0.43280670046806335,
|
| 6343 |
-
"learning_rate": 5.1289398280802295e-06,
|
| 6344 |
-
"loss": 1.0024726867675782,
|
| 6345 |
-
"mean_token_accuracy": 0.7771938517689705,
|
| 6346 |
-
"num_tokens": 11538936.0,
|
| 6347 |
-
"step": 5710
|
| 6348 |
-
},
|
| 6349 |
-
{
|
| 6350 |
-
"entropy": 0.9646360605955124,
|
| 6351 |
-
"epoch": 1.4900032562683165,
|
| 6352 |
-
"grad_norm": 0.45593976974487305,
|
| 6353 |
-
"learning_rate": 5.1028913779630115e-06,
|
| 6354 |
-
"loss": 0.9611331939697265,
|
| 6355 |
-
"mean_token_accuracy": 0.7832529351115227,
|
| 6356 |
-
"num_tokens": 11664929.0,
|
| 6357 |
-
"step": 5720
|
| 6358 |
-
},
|
| 6359 |
-
{
|
| 6360 |
-
"entropy": 0.9973645031452179,
|
| 6361 |
-
"epoch": 1.492608270921524,
|
| 6362 |
-
"grad_norm": 0.440469354391098,
|
| 6363 |
-
"learning_rate": 5.0768429278457935e-06,
|
| 6364 |
-
"loss": 0.9924856185913086,
|
| 6365 |
-
"mean_token_accuracy": 0.7780518174171448,
|
| 6366 |
-
"num_tokens": 11794319.0,
|
| 6367 |
-
"step": 5730
|
| 6368 |
-
},
|
| 6369 |
-
{
|
| 6370 |
-
"entropy": 0.9901895999908448,
|
| 6371 |
-
"epoch": 1.4952132855747313,
|
| 6372 |
-
"grad_norm": 0.42248594760894775,
|
| 6373 |
-
"learning_rate": 5.0507944777285756e-06,
|
| 6374 |
-
"loss": 0.9912988662719726,
|
| 6375 |
-
"mean_token_accuracy": 0.7825104221701622,
|
| 6376 |
-
"num_tokens": 11921665.0,
|
| 6377 |
-
"step": 5740
|
| 6378 |
-
},
|
| 6379 |
-
{
|
| 6380 |
-
"entropy": 0.9788355082273483,
|
| 6381 |
-
"epoch": 1.4978183002279388,
|
| 6382 |
-
"grad_norm": 0.5046400427818298,
|
| 6383 |
-
"learning_rate": 5.024746027611358e-06,
|
| 6384 |
-
"loss": 0.9787019729614258,
|
| 6385 |
-
"mean_token_accuracy": 0.7822767272591591,
|
| 6386 |
-
"num_tokens": 12043672.0,
|
| 6387 |
-
"step": 5750
|
| 6388 |
-
},
|
| 6389 |
-
{
|
| 6390 |
-
"entropy": 0.9165773242712021,
|
| 6391 |
-
"epoch": 1.500423314881146,
|
| 6392 |
-
"grad_norm": 0.47194966673851013,
|
| 6393 |
-
"learning_rate": 4.99869757749414e-06,
|
| 6394 |
-
"loss": 0.907895851135254,
|
| 6395 |
-
"mean_token_accuracy": 0.7939924985170365,
|
| 6396 |
-
"num_tokens": 12168158.0,
|
| 6397 |
-
"step": 5760
|
| 6398 |
-
},
|
| 6399 |
-
{
|
| 6400 |
-
"entropy": 0.9998183622956276,
|
| 6401 |
-
"epoch": 1.5030283295343536,
|
| 6402 |
-
"grad_norm": 0.48401686549186707,
|
| 6403 |
-
"learning_rate": 4.972649127376922e-06,
|
| 6404 |
-
"loss": 0.995943832397461,
|
| 6405 |
-
"mean_token_accuracy": 0.7780294433236122,
|
| 6406 |
-
"num_tokens": 12293055.0,
|
| 6407 |
-
"step": 5770
|
| 6408 |
-
},
|
| 6409 |
-
{
|
| 6410 |
-
"entropy": 0.9550597861409187,
|
| 6411 |
-
"epoch": 1.5056333441875611,
|
| 6412 |
-
"grad_norm": 0.4364171624183655,
|
| 6413 |
-
"learning_rate": 4.946600677259704e-06,
|
| 6414 |
-
"loss": 0.9600460052490234,
|
| 6415 |
-
"mean_token_accuracy": 0.7861277416348458,
|
| 6416 |
-
"num_tokens": 12423316.0,
|
| 6417 |
-
"step": 5780
|
| 6418 |
-
},
|
| 6419 |
-
{
|
| 6420 |
-
"entropy": 1.0095638766884805,
|
| 6421 |
-
"epoch": 1.5082383588407686,
|
| 6422 |
-
"grad_norm": 0.47353091835975647,
|
| 6423 |
-
"learning_rate": 4.920552227142486e-06,
|
| 6424 |
-
"loss": 1.0030521392822265,
|
| 6425 |
-
"mean_token_accuracy": 0.7775545686483383,
|
| 6426 |
-
"num_tokens": 12549491.0,
|
| 6427 |
-
"step": 5790
|
| 6428 |
-
},
|
| 6429 |
-
{
|
| 6430 |
-
"entropy": 1.0024770304560662,
|
| 6431 |
-
"epoch": 1.510843373493976,
|
| 6432 |
-
"grad_norm": 0.44944795966148376,
|
| 6433 |
-
"learning_rate": 4.894503777025268e-06,
|
| 6434 |
-
"loss": 0.9964849472045898,
|
| 6435 |
-
"mean_token_accuracy": 0.7789783880114556,
|
| 6436 |
-
"num_tokens": 12673196.0,
|
| 6437 |
-
"step": 5800
|
| 6438 |
-
},
|
| 6439 |
-
{
|
| 6440 |
-
"epoch": 1.510843373493976,
|
| 6441 |
-
"eval_entropy": 0.9865379387077238,
|
| 6442 |
-
"eval_loss": 0.9859896898269653,
|
| 6443 |
-
"eval_mean_token_accuracy": 0.78121930945634,
|
| 6444 |
-
"eval_num_tokens": 12673196.0,
|
| 6445 |
-
"eval_runtime": 1256.4729,
|
| 6446 |
-
"eval_samples_per_second": 5.417,
|
| 6447 |
-
"eval_steps_per_second": 1.355,
|
| 6448 |
-
"step": 5800
|
| 6449 |
-
},
|
| 6450 |
-
{
|
| 6451 |
-
"entropy": 0.9638326957821846,
|
| 6452 |
-
"epoch": 1.5134483881471832,
|
| 6453 |
-
"grad_norm": 0.5642250180244446,
|
| 6454 |
-
"learning_rate": 4.868455326908049e-06,
|
| 6455 |
-
"loss": 0.9621557235717774,
|
| 6456 |
-
"mean_token_accuracy": 0.785402400791645,
|
| 6457 |
-
"num_tokens": 12799063.0,
|
| 6458 |
-
"step": 5810
|
| 6459 |
-
},
|
| 6460 |
-
{
|
| 6461 |
-
"entropy": 0.9616957351565361,
|
| 6462 |
-
"epoch": 1.5160534028003907,
|
| 6463 |
-
"grad_norm": 0.4595557153224945,
|
| 6464 |
-
"learning_rate": 4.842406876790831e-06,
|
| 6465 |
-
"loss": 0.9615615844726563,
|
| 6466 |
-
"mean_token_accuracy": 0.7842398881912231,
|
| 6467 |
-
"num_tokens": 12928006.0,
|
| 6468 |
-
"step": 5820
|
| 6469 |
-
},
|
| 6470 |
-
{
|
| 6471 |
-
"entropy": 0.9759504601359368,
|
| 6472 |
-
"epoch": 1.5186584174535982,
|
| 6473 |
-
"grad_norm": 0.4274995028972626,
|
| 6474 |
-
"learning_rate": 4.816358426673614e-06,
|
| 6475 |
-
"loss": 0.9724794387817383,
|
| 6476 |
-
"mean_token_accuracy": 0.7829881265759469,
|
| 6477 |
-
"num_tokens": 13055699.0,
|
| 6478 |
-
"step": 5830
|
| 6479 |
-
},
|
| 6480 |
-
{
|
| 6481 |
-
"entropy": 0.9953486144542694,
|
| 6482 |
-
"epoch": 1.5212634321068057,
|
| 6483 |
-
"grad_norm": 0.4352840483188629,
|
| 6484 |
-
"learning_rate": 4.790309976556395e-06,
|
| 6485 |
-
"loss": 0.9870157241821289,
|
| 6486 |
-
"mean_token_accuracy": 0.7791727811098099,
|
| 6487 |
-
"num_tokens": 13180253.0,
|
| 6488 |
-
"step": 5840
|
| 6489 |
-
},
|
| 6490 |
-
{
|
| 6491 |
-
"entropy": 1.003223615884781,
|
| 6492 |
-
"epoch": 1.523868446760013,
|
| 6493 |
-
"grad_norm": 0.4743148684501648,
|
| 6494 |
-
"learning_rate": 4.764261526439177e-06,
|
| 6495 |
-
"loss": 1.0010628700256348,
|
| 6496 |
-
"mean_token_accuracy": 0.7773196116089821,
|
| 6497 |
-
"num_tokens": 13307671.0,
|
| 6498 |
-
"step": 5850
|
| 6499 |
-
},
|
| 6500 |
-
{
|
| 6501 |
-
"entropy": 1.0089710101485252,
|
| 6502 |
-
"epoch": 1.5264734614132205,
|
| 6503 |
-
"grad_norm": 0.49535682797431946,
|
| 6504 |
-
"learning_rate": 4.738213076321959e-06,
|
| 6505 |
-
"loss": 1.002498245239258,
|
| 6506 |
-
"mean_token_accuracy": 0.7766694515943527,
|
| 6507 |
-
"num_tokens": 13433253.0,
|
| 6508 |
-
"step": 5860
|
| 6509 |
-
},
|
| 6510 |
-
{
|
| 6511 |
-
"entropy": 1.0034620597958566,
|
| 6512 |
-
"epoch": 1.5290784760664278,
|
| 6513 |
-
"grad_norm": 0.4674288034439087,
|
| 6514 |
-
"learning_rate": 4.712164626204741e-06,
|
| 6515 |
-
"loss": 1.0025875091552734,
|
| 6516 |
-
"mean_token_accuracy": 0.777845348417759,
|
| 6517 |
-
"num_tokens": 13558835.0,
|
| 6518 |
-
"step": 5870
|
| 6519 |
-
},
|
| 6520 |
-
{
|
| 6521 |
-
"entropy": 1.0152205243706702,
|
| 6522 |
-
"epoch": 1.5316834907196353,
|
| 6523 |
-
"grad_norm": 0.4123161733150482,
|
| 6524 |
-
"learning_rate": 4.686116176087523e-06,
|
| 6525 |
-
"loss": 1.0081012725830079,
|
| 6526 |
-
"mean_token_accuracy": 0.7771851986646652,
|
| 6527 |
-
"num_tokens": 13684694.0,
|
| 6528 |
-
"step": 5880
|
| 6529 |
-
},
|
| 6530 |
-
{
|
| 6531 |
-
"entropy": 1.0014832943677903,
|
| 6532 |
-
"epoch": 1.5342885053728428,
|
| 6533 |
-
"grad_norm": 0.4308551847934723,
|
| 6534 |
-
"learning_rate": 4.660067725970305e-06,
|
| 6535 |
-
"loss": 1.0004742622375489,
|
| 6536 |
-
"mean_token_accuracy": 0.7791355445981025,
|
| 6537 |
-
"num_tokens": 13810512.0,
|
| 6538 |
-
"step": 5890
|
| 6539 |
-
},
|
| 6540 |
-
{
|
| 6541 |
-
"entropy": 0.9844494163990021,
|
| 6542 |
-
"epoch": 1.5368935200260503,
|
| 6543 |
-
"grad_norm": 0.41272395849227905,
|
| 6544 |
-
"learning_rate": 4.634019275853087e-06,
|
| 6545 |
-
"loss": 0.9907437324523926,
|
| 6546 |
-
"mean_token_accuracy": 0.7801890760660172,
|
| 6547 |
-
"num_tokens": 13938791.0,
|
| 6548 |
-
"step": 5900
|
| 6549 |
-
},
|
| 6550 |
-
{
|
| 6551 |
-
"epoch": 1.5368935200260503,
|
| 6552 |
-
"eval_entropy": 0.988329703218088,
|
| 6553 |
-
"eval_loss": 0.9858239889144897,
|
| 6554 |
-
"eval_mean_token_accuracy": 0.7812621929670753,
|
| 6555 |
-
"eval_num_tokens": 13938791.0,
|
| 6556 |
-
"eval_runtime": 1256.06,
|
| 6557 |
-
"eval_samples_per_second": 5.419,
|
| 6558 |
-
"eval_steps_per_second": 1.355,
|
| 6559 |
-
"step": 5900
|
| 6560 |
-
},
|
| 6561 |
-
{
|
| 6562 |
-
"entropy": 1.0191802248358726,
|
| 6563 |
-
"epoch": 1.5394985346792576,
|
| 6564 |
-
"grad_norm": 0.4569645822048187,
|
| 6565 |
-
"learning_rate": 4.607970825735869e-06,
|
| 6566 |
-
"loss": 1.0173967361450196,
|
| 6567 |
-
"mean_token_accuracy": 0.7746791526675224,
|
| 6568 |
-
"num_tokens": 14066692.0,
|
| 6569 |
-
"step": 5910
|
| 6570 |
-
},
|
| 6571 |
-
{
|
| 6572 |
-
"entropy": 1.0061603277921676,
|
| 6573 |
-
"epoch": 1.5421035493324649,
|
| 6574 |
-
"grad_norm": 0.49531760811805725,
|
| 6575 |
-
"learning_rate": 4.581922375618651e-06,
|
| 6576 |
-
"loss": 0.996528434753418,
|
| 6577 |
-
"mean_token_accuracy": 0.7769378647208214,
|
| 6578 |
-
"num_tokens": 14196108.0,
|
| 6579 |
-
"step": 5920
|
| 6580 |
-
},
|
| 6581 |
-
{
|
| 6582 |
-
"entropy": 1.0030234783887864,
|
| 6583 |
-
"epoch": 1.5447085639856724,
|
| 6584 |
-
"grad_norm": 0.46018144488334656,
|
| 6585 |
-
"learning_rate": 4.5558739255014325e-06,
|
| 6586 |
-
"loss": 0.9946203231811523,
|
| 6587 |
-
"mean_token_accuracy": 0.7782233148813248,
|
| 6588 |
-
"num_tokens": 14321406.0,
|
| 6589 |
-
"step": 5930
|
| 6590 |
-
},
|
| 6591 |
-
{
|
| 6592 |
-
"entropy": 0.9838744148612022,
|
| 6593 |
-
"epoch": 1.5473135786388799,
|
| 6594 |
-
"grad_norm": 0.4416678249835968,
|
| 6595 |
-
"learning_rate": 4.529825475384215e-06,
|
| 6596 |
-
"loss": 0.9903221130371094,
|
| 6597 |
-
"mean_token_accuracy": 0.7811186596751213,
|
| 6598 |
-
"num_tokens": 14447553.0,
|
| 6599 |
-
"step": 5940
|
| 6600 |
-
},
|
| 6601 |
-
{
|
| 6602 |
-
"entropy": 1.0334096074104309,
|
| 6603 |
-
"epoch": 1.5499185932920874,
|
| 6604 |
-
"grad_norm": 0.4901539087295532,
|
| 6605 |
-
"learning_rate": 4.5037770252669975e-06,
|
| 6606 |
-
"loss": 1.0264799118041992,
|
| 6607 |
-
"mean_token_accuracy": 0.7718588143587113,
|
| 6608 |
-
"num_tokens": 14572776.0,
|
| 6609 |
-
"step": 5950
|
| 6610 |
-
},
|
| 6611 |
-
{
|
| 6612 |
-
"entropy": 0.9825369924306869,
|
| 6613 |
-
"epoch": 1.5525236079452946,
|
| 6614 |
-
"grad_norm": 0.46482163667678833,
|
| 6615 |
-
"learning_rate": 4.477728575149779e-06,
|
| 6616 |
-
"loss": 0.9773822784423828,
|
| 6617 |
-
"mean_token_accuracy": 0.7825256258249282,
|
| 6618 |
-
"num_tokens": 14698771.0,
|
| 6619 |
-
"step": 5960
|
| 6620 |
-
},
|
| 6621 |
-
{
|
| 6622 |
-
"entropy": 1.0017751574516296,
|
| 6623 |
-
"epoch": 1.5551286225985022,
|
| 6624 |
-
"grad_norm": 0.4480389952659607,
|
| 6625 |
-
"learning_rate": 4.451680125032561e-06,
|
| 6626 |
-
"loss": 0.9876977920532226,
|
| 6627 |
-
"mean_token_accuracy": 0.7788993194699287,
|
| 6628 |
-
"num_tokens": 14828068.0,
|
| 6629 |
-
"step": 5970
|
| 6630 |
-
},
|
| 6631 |
-
{
|
| 6632 |
-
"entropy": 0.9880959033966065,
|
| 6633 |
-
"epoch": 1.5577336372517094,
|
| 6634 |
-
"grad_norm": 0.5405458807945251,
|
| 6635 |
-
"learning_rate": 4.425631674915343e-06,
|
| 6636 |
-
"loss": 0.9888273239135742,
|
| 6637 |
-
"mean_token_accuracy": 0.7789255991578102,
|
| 6638 |
-
"num_tokens": 14956897.0,
|
| 6639 |
-
"step": 5980
|
| 6640 |
-
},
|
| 6641 |
-
{
|
| 6642 |
-
"entropy": 0.9731021404266358,
|
| 6643 |
-
"epoch": 1.560338651904917,
|
| 6644 |
-
"grad_norm": 0.4288831353187561,
|
| 6645 |
-
"learning_rate": 4.399583224798125e-06,
|
| 6646 |
-
"loss": 0.9767970085144043,
|
| 6647 |
-
"mean_token_accuracy": 0.7836747691035271,
|
| 6648 |
-
"num_tokens": 15084027.0,
|
| 6649 |
-
"step": 5990
|
| 6650 |
-
},
|
| 6651 |
-
{
|
| 6652 |
-
"entropy": 0.9875722482800484,
|
| 6653 |
-
"epoch": 1.5629436665581244,
|
| 6654 |
-
"grad_norm": 0.4533047378063202,
|
| 6655 |
-
"learning_rate": 4.373534774680907e-06,
|
| 6656 |
-
"loss": 0.9807762145996094,
|
| 6657 |
-
"mean_token_accuracy": 0.7818003162741661,
|
| 6658 |
-
"num_tokens": 15213344.0,
|
| 6659 |
-
"step": 6000
|
| 6660 |
-
},
|
| 6661 |
-
{
|
| 6662 |
-
"epoch": 1.5629436665581244,
|
| 6663 |
-
"eval_entropy": 0.9917997027255954,
|
| 6664 |
-
"eval_loss": 0.9857071042060852,
|
| 6665 |
-
"eval_mean_token_accuracy": 0.7812367571576923,
|
| 6666 |
-
"eval_num_tokens": 15213344.0,
|
| 6667 |
-
"eval_runtime": 1255.4544,
|
| 6668 |
-
"eval_samples_per_second": 5.421,
|
| 6669 |
-
"eval_steps_per_second": 1.356,
|
| 6670 |
-
"step": 6000
|
| 6671 |
-
},
|
| 6672 |
-
{
|
| 6673 |
-
"entropy": 1.015291467308998,
|
| 6674 |
-
"epoch": 1.565548681211332,
|
| 6675 |
-
"grad_norm": 0.5271810293197632,
|
| 6676 |
-
"learning_rate": 4.347486324563689e-06,
|
| 6677 |
-
"loss": 1.0036104202270508,
|
| 6678 |
-
"mean_token_accuracy": 0.7787183627486229,
|
| 6679 |
-
"num_tokens": 15342378.0,
|
| 6680 |
-
"step": 6010
|
| 6681 |
-
},
|
| 6682 |
-
{
|
| 6683 |
-
"entropy": 0.9813457742333412,
|
| 6684 |
-
"epoch": 1.5681536958645392,
|
| 6685 |
-
"grad_norm": 0.5101106762886047,
|
| 6686 |
-
"learning_rate": 4.321437874446471e-06,
|
| 6687 |
-
"loss": 0.9781330108642579,
|
| 6688 |
-
"mean_token_accuracy": 0.782950209081173,
|
| 6689 |
-
"num_tokens": 15471856.0,
|
| 6690 |
-
"step": 6020
|
| 6691 |
-
},
|
| 6692 |
-
{
|
| 6693 |
-
"entropy": 0.9854018583893775,
|
| 6694 |
-
"epoch": 1.5707587105177465,
|
| 6695 |
-
"grad_norm": 0.47289198637008667,
|
| 6696 |
-
"learning_rate": 4.295389424329253e-06,
|
| 6697 |
-
"loss": 0.9898054122924804,
|
| 6698 |
-
"mean_token_accuracy": 0.7822852298617363,
|
| 6699 |
-
"num_tokens": 15597579.0,
|
| 6700 |
-
"step": 6030
|
| 6701 |
-
},
|
| 6702 |
-
{
|
| 6703 |
-
"entropy": 1.0115185648202896,
|
| 6704 |
-
"epoch": 1.573363725170954,
|
| 6705 |
-
"grad_norm": 0.6156825423240662,
|
| 6706 |
-
"learning_rate": 4.269340974212035e-06,
|
| 6707 |
-
"loss": 1.0101004600524903,
|
| 6708 |
-
"mean_token_accuracy": 0.7775449946522712,
|
| 6709 |
-
"num_tokens": 15723377.0,
|
| 6710 |
-
"step": 6040
|
| 6711 |
-
},
|
| 6712 |
-
{
|
| 6713 |
-
"entropy": 1.0091204077005387,
|
| 6714 |
-
"epoch": 1.5759687398241615,
|
| 6715 |
-
"grad_norm": 0.4316498637199402,
|
| 6716 |
-
"learning_rate": 4.243292524094817e-06,
|
| 6717 |
-
"loss": 1.004193115234375,
|
| 6718 |
-
"mean_token_accuracy": 0.7766278043389321,
|
| 6719 |
-
"num_tokens": 15850568.0,
|
| 6720 |
-
"step": 6050
|
| 6721 |
-
},
|
| 6722 |
-
{
|
| 6723 |
-
"entropy": 0.9943494066596031,
|
| 6724 |
-
"epoch": 1.578573754477369,
|
| 6725 |
-
"grad_norm": 0.43714702129364014,
|
| 6726 |
-
"learning_rate": 4.217244073977599e-06,
|
| 6727 |
-
"loss": 0.9877338409423828,
|
| 6728 |
-
"mean_token_accuracy": 0.7803116738796234,
|
| 6729 |
-
"num_tokens": 15975594.0,
|
| 6730 |
-
"step": 6060
|
| 6731 |
-
},
|
| 6732 |
-
{
|
| 6733 |
-
"entropy": 1.0086842522025108,
|
| 6734 |
-
"epoch": 1.5811787691305763,
|
| 6735 |
-
"grad_norm": 0.4606360197067261,
|
| 6736 |
-
"learning_rate": 4.191195623860381e-06,
|
| 6737 |
-
"loss": 1.0088016510009765,
|
| 6738 |
-
"mean_token_accuracy": 0.7768441960215569,
|
| 6739 |
-
"num_tokens": 16102721.0,
|
| 6740 |
-
"step": 6070
|
| 6741 |
-
},
|
| 6742 |
-
{
|
| 6743 |
-
"entropy": 0.9904811158776283,
|
| 6744 |
-
"epoch": 1.5837837837837838,
|
| 6745 |
-
"grad_norm": 0.4409731924533844,
|
| 6746 |
-
"learning_rate": 4.165147173743162e-06,
|
| 6747 |
-
"loss": 0.9882152557373047,
|
| 6748 |
-
"mean_token_accuracy": 0.7814568996429443,
|
| 6749 |
-
"num_tokens": 16228527.0,
|
| 6750 |
-
"step": 6080
|
| 6751 |
-
},
|
| 6752 |
-
{
|
| 6753 |
-
"entropy": 0.9894454151391983,
|
| 6754 |
-
"epoch": 1.586388798436991,
|
| 6755 |
-
"grad_norm": 0.43606066703796387,
|
| 6756 |
-
"learning_rate": 4.139098723625944e-06,
|
| 6757 |
-
"loss": 0.9862225532531739,
|
| 6758 |
-
"mean_token_accuracy": 0.7819684982299805,
|
| 6759 |
-
"num_tokens": 16355704.0,
|
| 6760 |
-
"step": 6090
|
| 6761 |
-
},
|
| 6762 |
-
{
|
| 6763 |
-
"entropy": 1.0019489586353303,
|
| 6764 |
-
"epoch": 1.5889938130901986,
|
| 6765 |
-
"grad_norm": 0.45068544149398804,
|
| 6766 |
-
"learning_rate": 4.113050273508726e-06,
|
| 6767 |
-
"loss": 0.9972606658935547,
|
| 6768 |
-
"mean_token_accuracy": 0.7783585876226425,
|
| 6769 |
-
"num_tokens": 16483250.0,
|
| 6770 |
-
"step": 6100
|
| 6771 |
-
},
|
| 6772 |
-
{
|
| 6773 |
-
"epoch": 1.5889938130901986,
|
| 6774 |
-
"eval_entropy": 0.9835532722265824,
|
| 6775 |
-
"eval_loss": 0.9855121374130249,
|
| 6776 |
-
"eval_mean_token_accuracy": 0.7813105090454519,
|
| 6777 |
-
"eval_num_tokens": 16483250.0,
|
| 6778 |
-
"eval_runtime": 1254.4678,
|
| 6779 |
-
"eval_samples_per_second": 5.425,
|
| 6780 |
-
"eval_steps_per_second": 1.357,
|
| 6781 |
-
"step": 6100
|
| 6782 |
-
},
|
| 6783 |
-
{
|
| 6784 |
-
"entropy": 0.976255078613758,
|
| 6785 |
-
"epoch": 1.591598827743406,
|
| 6786 |
-
"grad_norm": 0.4376319348812103,
|
| 6787 |
-
"learning_rate": 4.087001823391509e-06,
|
| 6788 |
-
"loss": 0.9689760208129883,
|
| 6789 |
-
"mean_token_accuracy": 0.7834189802408218,
|
| 6790 |
-
"num_tokens": 16608054.0,
|
| 6791 |
-
"step": 6110
|
| 6792 |
-
},
|
| 6793 |
-
{
|
| 6794 |
-
"entropy": 0.9631530359387398,
|
| 6795 |
-
"epoch": 1.5942038423966136,
|
| 6796 |
-
"grad_norm": 0.4029894471168518,
|
| 6797 |
-
"learning_rate": 4.06095337327429e-06,
|
| 6798 |
-
"loss": 0.9515939712524414,
|
| 6799 |
-
"mean_token_accuracy": 0.7867600157856941,
|
| 6800 |
-
"num_tokens": 16735703.0,
|
| 6801 |
-
"step": 6120
|
| 6802 |
-
},
|
| 6803 |
-
{
|
| 6804 |
-
"entropy": 0.9851865395903587,
|
| 6805 |
-
"epoch": 1.596808857049821,
|
| 6806 |
-
"grad_norm": 0.452918142080307,
|
| 6807 |
-
"learning_rate": 4.034904923157072e-06,
|
| 6808 |
-
"loss": 0.9814105033874512,
|
| 6809 |
-
"mean_token_accuracy": 0.7826128974556923,
|
| 6810 |
-
"num_tokens": 16859370.0,
|
| 6811 |
-
"step": 6130
|
| 6812 |
-
},
|
| 6813 |
-
{
|
| 6814 |
-
"entropy": 0.9910167038440705,
|
| 6815 |
-
"epoch": 1.5994138717030282,
|
| 6816 |
-
"grad_norm": 0.4704102575778961,
|
| 6817 |
-
"learning_rate": 4.008856473039854e-06,
|
| 6818 |
-
"loss": 0.9897844314575195,
|
| 6819 |
-
"mean_token_accuracy": 0.7792090073227882,
|
| 6820 |
-
"num_tokens": 16987280.0,
|
| 6821 |
-
"step": 6140
|
| 6822 |
-
},
|
| 6823 |
-
{
|
| 6824 |
-
"entropy": 0.9821922004222869,
|
| 6825 |
-
"epoch": 1.6020188863562357,
|
| 6826 |
-
"grad_norm": 0.43230006098747253,
|
| 6827 |
-
"learning_rate": 3.9828080229226365e-06,
|
| 6828 |
-
"loss": 0.9856078147888183,
|
| 6829 |
-
"mean_token_accuracy": 0.7809289380908012,
|
| 6830 |
-
"num_tokens": 17113280.0,
|
| 6831 |
-
"step": 6150
|
| 6832 |
-
},
|
| 6833 |
-
{
|
| 6834 |
-
"entropy": 0.9871781796216965,
|
| 6835 |
-
"epoch": 1.6046239010094432,
|
| 6836 |
-
"grad_norm": 0.5128525495529175,
|
| 6837 |
-
"learning_rate": 3.9567595728054185e-06,
|
| 6838 |
-
"loss": 0.9861786842346192,
|
| 6839 |
-
"mean_token_accuracy": 0.7808286741375923,
|
| 6840 |
-
"num_tokens": 17238030.0,
|
| 6841 |
-
"step": 6160
|
| 6842 |
-
},
|
| 6843 |
-
{
|
| 6844 |
-
"entropy": 0.973116934299469,
|
| 6845 |
-
"epoch": 1.6072289156626507,
|
| 6846 |
-
"grad_norm": 0.45384928584098816,
|
| 6847 |
-
"learning_rate": 3.9307111226882005e-06,
|
| 6848 |
-
"loss": 0.9666866302490235,
|
| 6849 |
-
"mean_token_accuracy": 0.784807562828064,
|
| 6850 |
-
"num_tokens": 17365602.0,
|
| 6851 |
-
"step": 6170
|
| 6852 |
-
},
|
| 6853 |
-
{
|
| 6854 |
-
"entropy": 0.9911901175975799,
|
| 6855 |
-
"epoch": 1.609833930315858,
|
| 6856 |
-
"grad_norm": 0.43043264746665955,
|
| 6857 |
-
"learning_rate": 3.9046626725709826e-06,
|
| 6858 |
-
"loss": 0.9839757919311524,
|
| 6859 |
-
"mean_token_accuracy": 0.7797017201781273,
|
| 6860 |
-
"num_tokens": 17496240.0,
|
| 6861 |
-
"step": 6180
|
| 6862 |
-
},
|
| 6863 |
-
{
|
| 6864 |
-
"entropy": 0.9626880213618279,
|
| 6865 |
-
"epoch": 1.6124389449690655,
|
| 6866 |
-
"grad_norm": 0.4411322772502899,
|
| 6867 |
-
"learning_rate": 3.878614222453765e-06,
|
| 6868 |
-
"loss": 0.9572731018066406,
|
| 6869 |
-
"mean_token_accuracy": 0.7839711874723434,
|
| 6870 |
-
"num_tokens": 17621635.0,
|
| 6871 |
-
"step": 6190
|
| 6872 |
-
},
|
| 6873 |
-
{
|
| 6874 |
-
"entropy": 1.0114291816949845,
|
| 6875 |
-
"epoch": 1.6150439596222728,
|
| 6876 |
-
"grad_norm": 0.4990837574005127,
|
| 6877 |
-
"learning_rate": 3.852565772336546e-06,
|
| 6878 |
-
"loss": 1.0094215393066406,
|
| 6879 |
-
"mean_token_accuracy": 0.7744873777031899,
|
| 6880 |
-
"num_tokens": 17748242.0,
|
| 6881 |
-
"step": 6200
|
| 6882 |
-
},
|
| 6883 |
-
{
|
| 6884 |
-
"epoch": 1.6150439596222728,
|
| 6885 |
-
"eval_entropy": 0.9851454944994419,
|
| 6886 |
-
"eval_loss": 0.9853406548500061,
|
| 6887 |
-
"eval_mean_token_accuracy": 0.781301787786282,
|
| 6888 |
-
"eval_num_tokens": 17748242.0,
|
| 6889 |
-
"eval_runtime": 1256.6923,
|
| 6890 |
-
"eval_samples_per_second": 5.416,
|
| 6891 |
-
"eval_steps_per_second": 1.354,
|
| 6892 |
-
"step": 6200
|
| 6893 |
-
},
|
| 6894 |
-
{
|
| 6895 |
-
"entropy": 0.982139666378498,
|
| 6896 |
-
"epoch": 1.6176489742754803,
|
| 6897 |
-
"grad_norm": 0.4867699146270752,
|
| 6898 |
-
"learning_rate": 3.826517322219328e-06,
|
| 6899 |
-
"loss": 0.9829599380493164,
|
| 6900 |
-
"mean_token_accuracy": 0.7825804129242897,
|
| 6901 |
-
"num_tokens": 17873657.0,
|
| 6902 |
-
"step": 6210
|
| 6903 |
-
},
|
| 6904 |
-
{
|
| 6905 |
-
"entropy": 0.9933568194508553,
|
| 6906 |
-
"epoch": 1.6202539889286878,
|
| 6907 |
-
"grad_norm": 0.4696708917617798,
|
| 6908 |
-
"learning_rate": 3.8004688721021103e-06,
|
| 6909 |
-
"loss": 0.9762285232543946,
|
| 6910 |
-
"mean_token_accuracy": 0.7807412505149841,
|
| 6911 |
-
"num_tokens": 18001335.0,
|
| 6912 |
-
"step": 6220
|
| 6913 |
-
},
|
| 6914 |
-
{
|
| 6915 |
-
"entropy": 0.9988187551498413,
|
| 6916 |
-
"epoch": 1.6228590035818953,
|
| 6917 |
-
"grad_norm": 0.45554623007774353,
|
| 6918 |
-
"learning_rate": 3.7744204219848923e-06,
|
| 6919 |
-
"loss": 1.0020133972167968,
|
| 6920 |
-
"mean_token_accuracy": 0.7782625585794449,
|
| 6921 |
-
"num_tokens": 18126670.0,
|
| 6922 |
-
"step": 6230
|
| 6923 |
-
},
|
| 6924 |
-
{
|
| 6925 |
-
"entropy": 0.9752851694822311,
|
| 6926 |
-
"epoch": 1.6254640182351026,
|
| 6927 |
-
"grad_norm": 0.43309473991394043,
|
| 6928 |
-
"learning_rate": 3.7483719718676743e-06,
|
| 6929 |
-
"loss": 0.9772072792053222,
|
| 6930 |
-
"mean_token_accuracy": 0.7809980019927025,
|
| 6931 |
-
"num_tokens": 18249535.0,
|
| 6932 |
-
"step": 6240
|
| 6933 |
-
},
|
| 6934 |
-
{
|
| 6935 |
-
"entropy": 0.9565880730748176,
|
| 6936 |
-
"epoch": 1.6280690328883098,
|
| 6937 |
-
"grad_norm": 0.484499990940094,
|
| 6938 |
-
"learning_rate": 3.722323521750456e-06,
|
| 6939 |
-
"loss": 0.9521919250488281,
|
| 6940 |
-
"mean_token_accuracy": 0.7870877295732498,
|
| 6941 |
-
"num_tokens": 18376177.0,
|
| 6942 |
-
"step": 6250
|
| 6943 |
-
},
|
| 6944 |
-
{
|
| 6945 |
-
"entropy": 0.9865999668836594,
|
| 6946 |
-
"epoch": 1.6306740475415173,
|
| 6947 |
-
"grad_norm": 0.4812857508659363,
|
| 6948 |
-
"learning_rate": 3.696275071633238e-06,
|
| 6949 |
-
"loss": 0.9790876388549805,
|
| 6950 |
-
"mean_token_accuracy": 0.7824391439557076,
|
| 6951 |
-
"num_tokens": 18502020.0,
|
| 6952 |
-
"step": 6260
|
| 6953 |
-
},
|
| 6954 |
-
{
|
| 6955 |
-
"entropy": 0.9682453617453575,
|
| 6956 |
-
"epoch": 1.6332790621947249,
|
| 6957 |
-
"grad_norm": 0.5014381408691406,
|
| 6958 |
-
"learning_rate": 3.67022662151602e-06,
|
| 6959 |
-
"loss": 0.9671321868896484,
|
| 6960 |
-
"mean_token_accuracy": 0.784609442949295,
|
| 6961 |
-
"num_tokens": 18628895.0,
|
| 6962 |
-
"step": 6270
|
| 6963 |
-
},
|
| 6964 |
-
{
|
| 6965 |
-
"entropy": 1.0102926969528199,
|
| 6966 |
-
"epoch": 1.6358840768479324,
|
| 6967 |
-
"grad_norm": 0.4856639504432678,
|
| 6968 |
-
"learning_rate": 3.644178171398802e-06,
|
| 6969 |
-
"loss": 1.004817008972168,
|
| 6970 |
-
"mean_token_accuracy": 0.7764660209417343,
|
| 6971 |
-
"num_tokens": 18755634.0,
|
| 6972 |
-
"step": 6280
|
| 6973 |
-
},
|
| 6974 |
-
{
|
| 6975 |
-
"entropy": 0.9895771816372871,
|
| 6976 |
-
"epoch": 1.6384890915011399,
|
| 6977 |
-
"grad_norm": 0.4235684275627136,
|
| 6978 |
-
"learning_rate": 3.618129721281584e-06,
|
| 6979 |
-
"loss": 0.9898414611816406,
|
| 6980 |
-
"mean_token_accuracy": 0.7806871175765991,
|
| 6981 |
-
"num_tokens": 18882454.0,
|
| 6982 |
-
"step": 6290
|
| 6983 |
-
},
|
| 6984 |
-
{
|
| 6985 |
-
"entropy": 0.9976322814822197,
|
| 6986 |
-
"epoch": 1.6410941061543471,
|
| 6987 |
-
"grad_norm": 0.4656156599521637,
|
| 6988 |
-
"learning_rate": 3.592081271164366e-06,
|
| 6989 |
-
"loss": 0.9955741882324218,
|
| 6990 |
-
"mean_token_accuracy": 0.7813329964876174,
|
| 6991 |
-
"num_tokens": 19012278.0,
|
| 6992 |
-
"step": 6300
|
| 6993 |
-
},
|
| 6994 |
-
{
|
| 6995 |
-
"epoch": 1.6410941061543471,
|
| 6996 |
-
"eval_entropy": 0.9860669597195403,
|
| 6997 |
-
"eval_loss": 0.985215961933136,
|
| 6998 |
-
"eval_mean_token_accuracy": 0.7812749157571064,
|
| 6999 |
-
"eval_num_tokens": 19012278.0,
|
| 7000 |
-
"eval_runtime": 1256.8998,
|
| 7001 |
-
"eval_samples_per_second": 5.415,
|
| 7002 |
-
"eval_steps_per_second": 1.354,
|
| 7003 |
-
"step": 6300
|
| 7004 |
-
},
|
| 7005 |
-
{
|
| 7006 |
-
"entropy": 0.9824397191405296,
|
| 7007 |
-
"epoch": 1.6436991208075544,
|
| 7008 |
-
"grad_norm": 0.4692520797252655,
|
| 7009 |
-
"learning_rate": 3.5660328210471477e-06,
|
| 7010 |
-
"loss": 0.9777761459350586,
|
| 7011 |
-
"mean_token_accuracy": 0.7832236424088478,
|
| 7012 |
-
"num_tokens": 19136646.0,
|
| 7013 |
-
"step": 6310
|
| 7014 |
-
},
|
| 7015 |
-
{
|
| 7016 |
-
"entropy": 0.9898837953805923,
|
| 7017 |
-
"epoch": 1.646304135460762,
|
| 7018 |
-
"grad_norm": 0.48879489302635193,
|
| 7019 |
-
"learning_rate": 3.5399843709299298e-06,
|
| 7020 |
-
"loss": 0.9906368255615234,
|
| 7021 |
-
"mean_token_accuracy": 0.7796260580420494,
|
| 7022 |
-
"num_tokens": 19263524.0,
|
| 7023 |
-
"step": 6320
|
| 7024 |
-
},
|
| 7025 |
-
{
|
| 7026 |
-
"entropy": 0.9903134703636169,
|
| 7027 |
-
"epoch": 1.6489091501139694,
|
| 7028 |
-
"grad_norm": 0.444400817155838,
|
| 7029 |
-
"learning_rate": 3.5139359208127122e-06,
|
| 7030 |
-
"loss": 0.9902142524719239,
|
| 7031 |
-
"mean_token_accuracy": 0.7802131101489067,
|
| 7032 |
-
"num_tokens": 19389562.0,
|
| 7033 |
-
"step": 6330
|
| 7034 |
-
},
|
| 7035 |
-
{
|
| 7036 |
-
"entropy": 0.9439644649624824,
|
| 7037 |
-
"epoch": 1.651514164767177,
|
| 7038 |
-
"grad_norm": 0.4586019814014435,
|
| 7039 |
-
"learning_rate": 3.487887470695494e-06,
|
| 7040 |
-
"loss": 0.9457040786743164,
|
| 7041 |
-
"mean_token_accuracy": 0.7893780574202538,
|
| 7042 |
-
"num_tokens": 19516575.0,
|
| 7043 |
-
"step": 6340
|
| 7044 |
-
},
|
| 7045 |
-
{
|
| 7046 |
-
"entropy": 0.9767524793744087,
|
| 7047 |
-
"epoch": 1.6541191794203842,
|
| 7048 |
-
"grad_norm": 0.4149959981441498,
|
| 7049 |
-
"learning_rate": 3.461839020578276e-06,
|
| 7050 |
-
"loss": 0.9651622772216797,
|
| 7051 |
-
"mean_token_accuracy": 0.7854742303490638,
|
| 7052 |
-
"num_tokens": 19644896.0,
|
| 7053 |
-
"step": 6350
|
| 7054 |
-
},
|
| 7055 |
-
{
|
| 7056 |
-
"entropy": 0.9634524628520011,
|
| 7057 |
-
"epoch": 1.6567241940735915,
|
| 7058 |
-
"grad_norm": 0.4312231242656708,
|
| 7059 |
-
"learning_rate": 3.435790570461058e-06,
|
| 7060 |
-
"loss": 0.9572691917419434,
|
| 7061 |
-
"mean_token_accuracy": 0.7873846724629402,
|
| 7062 |
-
"num_tokens": 19772864.0,
|
| 7063 |
-
"step": 6360
|
| 7064 |
-
},
|
| 7065 |
-
{
|
| 7066 |
-
"entropy": 0.9760506734251976,
|
| 7067 |
-
"epoch": 1.659329208726799,
|
| 7068 |
-
"grad_norm": 0.4425853490829468,
|
| 7069 |
-
"learning_rate": 3.4097421203438395e-06,
|
| 7070 |
-
"loss": 0.9743513107299805,
|
| 7071 |
-
"mean_token_accuracy": 0.7828145936131478,
|
| 7072 |
-
"num_tokens": 19899888.0,
|
| 7073 |
-
"step": 6370
|
| 7074 |
-
},
|
| 7075 |
-
{
|
| 7076 |
-
"entropy": 1.0453799113631248,
|
| 7077 |
-
"epoch": 1.6619342233800065,
|
| 7078 |
-
"grad_norm": 0.5294102430343628,
|
| 7079 |
-
"learning_rate": 3.3836936702266216e-06,
|
| 7080 |
-
"loss": 1.034913921356201,
|
| 7081 |
-
"mean_token_accuracy": 0.7686419218778611,
|
| 7082 |
-
"num_tokens": 20025043.0,
|
| 7083 |
-
"step": 6380
|
| 7084 |
-
},
|
| 7085 |
-
{
|
| 7086 |
-
"entropy": 0.9998988956212997,
|
| 7087 |
-
"epoch": 1.664539238033214,
|
| 7088 |
-
"grad_norm": 0.45683664083480835,
|
| 7089 |
-
"learning_rate": 3.357645220109404e-06,
|
| 7090 |
-
"loss": 0.9996541023254395,
|
| 7091 |
-
"mean_token_accuracy": 0.7783705160021782,
|
| 7092 |
-
"num_tokens": 20150902.0,
|
| 7093 |
-
"step": 6390
|
| 7094 |
-
},
|
| 7095 |
-
{
|
| 7096 |
-
"entropy": 1.027308452129364,
|
| 7097 |
-
"epoch": 1.6671442526864215,
|
| 7098 |
-
"grad_norm": 0.4325750172138214,
|
| 7099 |
-
"learning_rate": 3.331596769992186e-06,
|
| 7100 |
-
"loss": 1.0322607040405274,
|
| 7101 |
-
"mean_token_accuracy": 0.7724197462201119,
|
| 7102 |
-
"num_tokens": 20273299.0,
|
| 7103 |
-
"step": 6400
|
| 7104 |
-
},
|
| 7105 |
-
{
|
| 7106 |
-
"epoch": 1.6671442526864215,
|
| 7107 |
-
"eval_entropy": 0.9872279417921037,
|
| 7108 |
-
"eval_loss": 0.9850819706916809,
|
| 7109 |
-
"eval_mean_token_accuracy": 0.7812954522888192,
|
| 7110 |
-
"eval_num_tokens": 20273299.0,
|
| 7111 |
-
"eval_runtime": 1255.7356,
|
| 7112 |
-
"eval_samples_per_second": 5.42,
|
| 7113 |
-
"eval_steps_per_second": 1.355,
|
| 7114 |
-
"step": 6400
|
| 7115 |
}
|
| 7116 |
],
|
| 7117 |
"logging_steps": 10,
|
| 7118 |
-
"max_steps":
|
| 7119 |
"num_input_tokens_seen": 0,
|
| 7120 |
-
"num_train_epochs":
|
| 7121 |
"save_steps": 200,
|
| 7122 |
"stateful_callbacks": {
|
| 7123 |
"TrainerControl": {
|
|
@@ -7131,7 +5355,7 @@
|
|
| 7131 |
"attributes": {}
|
| 7132 |
}
|
| 7133 |
},
|
| 7134 |
-
"total_flos":
|
| 7135 |
"train_batch_size": 2,
|
| 7136 |
"trial_name": null,
|
| 7137 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.8267665676269216,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 4800,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 5336 |
"eval_samples_per_second": 5.161,
|
| 5337 |
"eval_steps_per_second": 1.291,
|
| 5338 |
"step": 4800
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5339 |
}
|
| 5340 |
],
|
| 5341 |
"logging_steps": 10,
|
| 5342 |
+
"max_steps": 5806,
|
| 5343 |
"num_input_tokens_seen": 0,
|
| 5344 |
+
"num_train_epochs": 1,
|
| 5345 |
"save_steps": 200,
|
| 5346 |
"stateful_callbacks": {
|
| 5347 |
"TrainerControl": {
|
|
|
|
| 5355 |
"attributes": {}
|
| 5356 |
}
|
| 5357 |
},
|
| 5358 |
+
"total_flos": 7.887012992552796e+17,
|
| 5359 |
"train_batch_size": 2,
|
| 5360 |
"trial_name": null,
|
| 5361 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e193b6114ef398d374d4a4437acc56d06a32e2263d65fa1d64344fb64467f82c
|
| 3 |
size 5777
|