Training in progress, step 8000, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4523108832
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2e7952d49f3b72dccae7df93d53fa89480a80bcf6a6ea965af4f805a6b2ee61
|
| 3 |
size 4523108832
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2912179275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c291b517be6ed9a8a8372cb6fdf80312291ffe3720046ec0442ccb876a2f941
|
| 3 |
size 2912179275
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8553d3987609520495680f36a4682a760a418a930648ce46178a1399cde440b0
|
| 3 |
size 14645
|
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 sha256:0c5dbcb730f9eb04aeb5f41b591529ac503b5347bbda8703b5a9a31804c671aa
|
| 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 sha256:2db53aadb4360cb98e71593df7e562c9eb432a4b511eb18cdf0621667cf3d66b
|
| 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": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -5258,6 +5258,356 @@
|
|
| 5258 |
"learning_rate": 5.0039999999999995e-05,
|
| 5259 |
"loss": 0.018700633943080903,
|
| 5260 |
"step": 7500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5261 |
}
|
| 5262 |
],
|
| 5263 |
"logging_steps": 10,
|
|
@@ -5277,7 +5627,7 @@
|
|
| 5277 |
"attributes": {}
|
| 5278 |
}
|
| 5279 |
},
|
| 5280 |
-
"total_flos": 6.
|
| 5281 |
"train_batch_size": 1,
|
| 5282 |
"trial_name": null,
|
| 5283 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0252,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 8000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 5258 |
"learning_rate": 5.0039999999999995e-05,
|
| 5259 |
"loss": 0.018700633943080903,
|
| 5260 |
"step": 7500
|
| 5261 |
+
},
|
| 5262 |
+
{
|
| 5263 |
+
"epoch": 4.0074,
|
| 5264 |
+
"grad_norm": 0.08042102307081223,
|
| 5265 |
+
"learning_rate": 4.9840000000000004e-05,
|
| 5266 |
+
"loss": 0.018940824270248412,
|
| 5267 |
+
"step": 7510
|
| 5268 |
+
},
|
| 5269 |
+
{
|
| 5270 |
+
"epoch": 4.0084,
|
| 5271 |
+
"grad_norm": 0.21641090512275696,
|
| 5272 |
+
"learning_rate": 4.9640000000000006e-05,
|
| 5273 |
+
"loss": 0.020995579659938812,
|
| 5274 |
+
"step": 7520
|
| 5275 |
+
},
|
| 5276 |
+
{
|
| 5277 |
+
"epoch": 4.0094,
|
| 5278 |
+
"grad_norm": 0.12526147067546844,
|
| 5279 |
+
"learning_rate": 4.944e-05,
|
| 5280 |
+
"loss": 0.019232508540153504,
|
| 5281 |
+
"step": 7530
|
| 5282 |
+
},
|
| 5283 |
+
{
|
| 5284 |
+
"epoch": 4.0104,
|
| 5285 |
+
"grad_norm": 0.10949081927537918,
|
| 5286 |
+
"learning_rate": 4.924e-05,
|
| 5287 |
+
"loss": 0.01903458535671234,
|
| 5288 |
+
"step": 7540
|
| 5289 |
+
},
|
| 5290 |
+
{
|
| 5291 |
+
"epoch": 4.0114,
|
| 5292 |
+
"grad_norm": 0.09758574515581131,
|
| 5293 |
+
"learning_rate": 4.9040000000000005e-05,
|
| 5294 |
+
"loss": 0.01993790864944458,
|
| 5295 |
+
"step": 7550
|
| 5296 |
+
},
|
| 5297 |
+
{
|
| 5298 |
+
"epoch": 4.0124,
|
| 5299 |
+
"grad_norm": 0.09602127224206924,
|
| 5300 |
+
"learning_rate": 4.884e-05,
|
| 5301 |
+
"loss": 0.021270765364170073,
|
| 5302 |
+
"step": 7560
|
| 5303 |
+
},
|
| 5304 |
+
{
|
| 5305 |
+
"epoch": 4.0134,
|
| 5306 |
+
"grad_norm": 0.08903051167726517,
|
| 5307 |
+
"learning_rate": 4.864e-05,
|
| 5308 |
+
"loss": 0.018507078289985657,
|
| 5309 |
+
"step": 7570
|
| 5310 |
+
},
|
| 5311 |
+
{
|
| 5312 |
+
"epoch": 4.0144,
|
| 5313 |
+
"grad_norm": 0.09527137875556946,
|
| 5314 |
+
"learning_rate": 4.8440000000000004e-05,
|
| 5315 |
+
"loss": 0.01924394816160202,
|
| 5316 |
+
"step": 7580
|
| 5317 |
+
},
|
| 5318 |
+
{
|
| 5319 |
+
"epoch": 4.0154,
|
| 5320 |
+
"grad_norm": 0.1218629702925682,
|
| 5321 |
+
"learning_rate": 4.824e-05,
|
| 5322 |
+
"loss": 0.02654995024204254,
|
| 5323 |
+
"step": 7590
|
| 5324 |
+
},
|
| 5325 |
+
{
|
| 5326 |
+
"epoch": 4.0164,
|
| 5327 |
+
"grad_norm": 0.0813978984951973,
|
| 5328 |
+
"learning_rate": 4.804e-05,
|
| 5329 |
+
"loss": 0.018305253982543946,
|
| 5330 |
+
"step": 7600
|
| 5331 |
+
},
|
| 5332 |
+
{
|
| 5333 |
+
"epoch": 4.0174,
|
| 5334 |
+
"grad_norm": 0.08926694095134735,
|
| 5335 |
+
"learning_rate": 4.784e-05,
|
| 5336 |
+
"loss": 0.01904298961162567,
|
| 5337 |
+
"step": 7610
|
| 5338 |
+
},
|
| 5339 |
+
{
|
| 5340 |
+
"epoch": 4.0184,
|
| 5341 |
+
"grad_norm": 0.084304079413414,
|
| 5342 |
+
"learning_rate": 4.7640000000000005e-05,
|
| 5343 |
+
"loss": 0.019423790276050568,
|
| 5344 |
+
"step": 7620
|
| 5345 |
+
},
|
| 5346 |
+
{
|
| 5347 |
+
"epoch": 4.0194,
|
| 5348 |
+
"grad_norm": 0.08805521577596664,
|
| 5349 |
+
"learning_rate": 4.744e-05,
|
| 5350 |
+
"loss": 0.019075404107570647,
|
| 5351 |
+
"step": 7630
|
| 5352 |
+
},
|
| 5353 |
+
{
|
| 5354 |
+
"epoch": 4.0204,
|
| 5355 |
+
"grad_norm": 0.07542338967323303,
|
| 5356 |
+
"learning_rate": 4.724e-05,
|
| 5357 |
+
"loss": 0.01921938955783844,
|
| 5358 |
+
"step": 7640
|
| 5359 |
+
},
|
| 5360 |
+
{
|
| 5361 |
+
"epoch": 4.0214,
|
| 5362 |
+
"grad_norm": 0.11425633728504181,
|
| 5363 |
+
"learning_rate": 4.7040000000000004e-05,
|
| 5364 |
+
"loss": 0.01996878832578659,
|
| 5365 |
+
"step": 7650
|
| 5366 |
+
},
|
| 5367 |
+
{
|
| 5368 |
+
"epoch": 4.0224,
|
| 5369 |
+
"grad_norm": 0.11465906351804733,
|
| 5370 |
+
"learning_rate": 4.684e-05,
|
| 5371 |
+
"loss": 0.019843164086341857,
|
| 5372 |
+
"step": 7660
|
| 5373 |
+
},
|
| 5374 |
+
{
|
| 5375 |
+
"epoch": 4.0234,
|
| 5376 |
+
"grad_norm": 0.12781135737895966,
|
| 5377 |
+
"learning_rate": 4.664e-05,
|
| 5378 |
+
"loss": 0.01992042511701584,
|
| 5379 |
+
"step": 7670
|
| 5380 |
+
},
|
| 5381 |
+
{
|
| 5382 |
+
"epoch": 4.0244,
|
| 5383 |
+
"grad_norm": 0.09651028364896774,
|
| 5384 |
+
"learning_rate": 4.644e-05,
|
| 5385 |
+
"loss": 0.019879813492298126,
|
| 5386 |
+
"step": 7680
|
| 5387 |
+
},
|
| 5388 |
+
{
|
| 5389 |
+
"epoch": 4.0254,
|
| 5390 |
+
"grad_norm": 0.10908477753400803,
|
| 5391 |
+
"learning_rate": 4.624e-05,
|
| 5392 |
+
"loss": 0.020540766417980194,
|
| 5393 |
+
"step": 7690
|
| 5394 |
+
},
|
| 5395 |
+
{
|
| 5396 |
+
"epoch": 4.0264,
|
| 5397 |
+
"grad_norm": 0.11591225117444992,
|
| 5398 |
+
"learning_rate": 4.604e-05,
|
| 5399 |
+
"loss": 0.020340165495872496,
|
| 5400 |
+
"step": 7700
|
| 5401 |
+
},
|
| 5402 |
+
{
|
| 5403 |
+
"epoch": 4.0274,
|
| 5404 |
+
"grad_norm": 0.0748429000377655,
|
| 5405 |
+
"learning_rate": 4.584e-05,
|
| 5406 |
+
"loss": 0.02119535654783249,
|
| 5407 |
+
"step": 7710
|
| 5408 |
+
},
|
| 5409 |
+
{
|
| 5410 |
+
"epoch": 4.0284,
|
| 5411 |
+
"grad_norm": 0.06569884717464447,
|
| 5412 |
+
"learning_rate": 4.564e-05,
|
| 5413 |
+
"loss": 0.019903770089149474,
|
| 5414 |
+
"step": 7720
|
| 5415 |
+
},
|
| 5416 |
+
{
|
| 5417 |
+
"epoch": 4.0294,
|
| 5418 |
+
"grad_norm": 0.08006953448057175,
|
| 5419 |
+
"learning_rate": 4.5440000000000005e-05,
|
| 5420 |
+
"loss": 0.020622527599334715,
|
| 5421 |
+
"step": 7730
|
| 5422 |
+
},
|
| 5423 |
+
{
|
| 5424 |
+
"epoch": 4.0304,
|
| 5425 |
+
"grad_norm": 0.12711670994758606,
|
| 5426 |
+
"learning_rate": 4.524000000000001e-05,
|
| 5427 |
+
"loss": 0.019541466236114503,
|
| 5428 |
+
"step": 7740
|
| 5429 |
+
},
|
| 5430 |
+
{
|
| 5431 |
+
"epoch": 5.0002,
|
| 5432 |
+
"grad_norm": 0.13896505534648895,
|
| 5433 |
+
"learning_rate": 4.504e-05,
|
| 5434 |
+
"loss": 0.020583076775074004,
|
| 5435 |
+
"step": 7750
|
| 5436 |
+
},
|
| 5437 |
+
{
|
| 5438 |
+
"epoch": 5.0012,
|
| 5439 |
+
"grad_norm": 0.08660919964313507,
|
| 5440 |
+
"learning_rate": 4.4840000000000004e-05,
|
| 5441 |
+
"loss": 0.016970406472682952,
|
| 5442 |
+
"step": 7760
|
| 5443 |
+
},
|
| 5444 |
+
{
|
| 5445 |
+
"epoch": 5.0022,
|
| 5446 |
+
"grad_norm": 0.11170201748609543,
|
| 5447 |
+
"learning_rate": 4.4640000000000006e-05,
|
| 5448 |
+
"loss": 0.018784117698669434,
|
| 5449 |
+
"step": 7770
|
| 5450 |
+
},
|
| 5451 |
+
{
|
| 5452 |
+
"epoch": 5.0032,
|
| 5453 |
+
"grad_norm": 0.08062026649713516,
|
| 5454 |
+
"learning_rate": 4.444e-05,
|
| 5455 |
+
"loss": 0.01944620758295059,
|
| 5456 |
+
"step": 7780
|
| 5457 |
+
},
|
| 5458 |
+
{
|
| 5459 |
+
"epoch": 5.0042,
|
| 5460 |
+
"grad_norm": 0.05483711510896683,
|
| 5461 |
+
"learning_rate": 4.424e-05,
|
| 5462 |
+
"loss": 0.01771644353866577,
|
| 5463 |
+
"step": 7790
|
| 5464 |
+
},
|
| 5465 |
+
{
|
| 5466 |
+
"epoch": 5.0052,
|
| 5467 |
+
"grad_norm": 0.08323200792074203,
|
| 5468 |
+
"learning_rate": 4.4040000000000005e-05,
|
| 5469 |
+
"loss": 0.01714482456445694,
|
| 5470 |
+
"step": 7800
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 5.0062,
|
| 5474 |
+
"grad_norm": 0.11725310981273651,
|
| 5475 |
+
"learning_rate": 4.384e-05,
|
| 5476 |
+
"loss": 0.01833456605672836,
|
| 5477 |
+
"step": 7810
|
| 5478 |
+
},
|
| 5479 |
+
{
|
| 5480 |
+
"epoch": 5.0072,
|
| 5481 |
+
"grad_norm": 0.08281055092811584,
|
| 5482 |
+
"learning_rate": 4.364e-05,
|
| 5483 |
+
"loss": 0.017857730388641357,
|
| 5484 |
+
"step": 7820
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 5.0082,
|
| 5488 |
+
"grad_norm": 0.08471981436014175,
|
| 5489 |
+
"learning_rate": 4.3440000000000004e-05,
|
| 5490 |
+
"loss": 0.0180238738656044,
|
| 5491 |
+
"step": 7830
|
| 5492 |
+
},
|
| 5493 |
+
{
|
| 5494 |
+
"epoch": 5.0092,
|
| 5495 |
+
"grad_norm": 0.07369330525398254,
|
| 5496 |
+
"learning_rate": 4.324e-05,
|
| 5497 |
+
"loss": 0.020367558300495147,
|
| 5498 |
+
"step": 7840
|
| 5499 |
+
},
|
| 5500 |
+
{
|
| 5501 |
+
"epoch": 5.0102,
|
| 5502 |
+
"grad_norm": 0.14078500866889954,
|
| 5503 |
+
"learning_rate": 4.304e-05,
|
| 5504 |
+
"loss": 0.018425200879573823,
|
| 5505 |
+
"step": 7850
|
| 5506 |
+
},
|
| 5507 |
+
{
|
| 5508 |
+
"epoch": 5.0112,
|
| 5509 |
+
"grad_norm": 0.09964534640312195,
|
| 5510 |
+
"learning_rate": 4.284e-05,
|
| 5511 |
+
"loss": 0.01857619732618332,
|
| 5512 |
+
"step": 7860
|
| 5513 |
+
},
|
| 5514 |
+
{
|
| 5515 |
+
"epoch": 5.0122,
|
| 5516 |
+
"grad_norm": 0.11316290497779846,
|
| 5517 |
+
"learning_rate": 4.2640000000000005e-05,
|
| 5518 |
+
"loss": 0.01818966567516327,
|
| 5519 |
+
"step": 7870
|
| 5520 |
+
},
|
| 5521 |
+
{
|
| 5522 |
+
"epoch": 5.0132,
|
| 5523 |
+
"grad_norm": 0.08515363186597824,
|
| 5524 |
+
"learning_rate": 4.244e-05,
|
| 5525 |
+
"loss": 0.01757376194000244,
|
| 5526 |
+
"step": 7880
|
| 5527 |
+
},
|
| 5528 |
+
{
|
| 5529 |
+
"epoch": 5.0142,
|
| 5530 |
+
"grad_norm": 0.07923150807619095,
|
| 5531 |
+
"learning_rate": 4.224e-05,
|
| 5532 |
+
"loss": 0.017273730039596556,
|
| 5533 |
+
"step": 7890
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 5.0152,
|
| 5537 |
+
"grad_norm": 0.13433587551116943,
|
| 5538 |
+
"learning_rate": 4.2040000000000004e-05,
|
| 5539 |
+
"loss": 0.02332073152065277,
|
| 5540 |
+
"step": 7900
|
| 5541 |
+
},
|
| 5542 |
+
{
|
| 5543 |
+
"epoch": 5.0162,
|
| 5544 |
+
"grad_norm": 0.08601813018321991,
|
| 5545 |
+
"learning_rate": 4.184e-05,
|
| 5546 |
+
"loss": 0.017112986743450166,
|
| 5547 |
+
"step": 7910
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 5.0172,
|
| 5551 |
+
"grad_norm": 0.09130693227052689,
|
| 5552 |
+
"learning_rate": 4.164e-05,
|
| 5553 |
+
"loss": 0.017460432648658753,
|
| 5554 |
+
"step": 7920
|
| 5555 |
+
},
|
| 5556 |
+
{
|
| 5557 |
+
"epoch": 5.0182,
|
| 5558 |
+
"grad_norm": 0.09017442166805267,
|
| 5559 |
+
"learning_rate": 4.144e-05,
|
| 5560 |
+
"loss": 0.01737082749605179,
|
| 5561 |
+
"step": 7930
|
| 5562 |
+
},
|
| 5563 |
+
{
|
| 5564 |
+
"epoch": 5.0192,
|
| 5565 |
+
"grad_norm": 0.08579915761947632,
|
| 5566 |
+
"learning_rate": 4.124e-05,
|
| 5567 |
+
"loss": 0.018616248667240144,
|
| 5568 |
+
"step": 7940
|
| 5569 |
+
},
|
| 5570 |
+
{
|
| 5571 |
+
"epoch": 5.0202,
|
| 5572 |
+
"grad_norm": 0.08709147572517395,
|
| 5573 |
+
"learning_rate": 4.104e-05,
|
| 5574 |
+
"loss": 0.018517845869064332,
|
| 5575 |
+
"step": 7950
|
| 5576 |
+
},
|
| 5577 |
+
{
|
| 5578 |
+
"epoch": 5.0212,
|
| 5579 |
+
"grad_norm": 0.1221432164311409,
|
| 5580 |
+
"learning_rate": 4.084e-05,
|
| 5581 |
+
"loss": 0.017741505801677705,
|
| 5582 |
+
"step": 7960
|
| 5583 |
+
},
|
| 5584 |
+
{
|
| 5585 |
+
"epoch": 5.0222,
|
| 5586 |
+
"grad_norm": 0.08435460925102234,
|
| 5587 |
+
"learning_rate": 4.064e-05,
|
| 5588 |
+
"loss": 0.018242837488651277,
|
| 5589 |
+
"step": 7970
|
| 5590 |
+
},
|
| 5591 |
+
{
|
| 5592 |
+
"epoch": 5.0232,
|
| 5593 |
+
"grad_norm": 0.07412870973348618,
|
| 5594 |
+
"learning_rate": 4.044e-05,
|
| 5595 |
+
"loss": 0.017411427199840547,
|
| 5596 |
+
"step": 7980
|
| 5597 |
+
},
|
| 5598 |
+
{
|
| 5599 |
+
"epoch": 5.0242,
|
| 5600 |
+
"grad_norm": 0.08610522001981735,
|
| 5601 |
+
"learning_rate": 4.024e-05,
|
| 5602 |
+
"loss": 0.01860019862651825,
|
| 5603 |
+
"step": 7990
|
| 5604 |
+
},
|
| 5605 |
+
{
|
| 5606 |
+
"epoch": 5.0252,
|
| 5607 |
+
"grad_norm": 0.06387773156166077,
|
| 5608 |
+
"learning_rate": 4.004e-05,
|
| 5609 |
+
"loss": 0.01812358796596527,
|
| 5610 |
+
"step": 8000
|
| 5611 |
}
|
| 5612 |
],
|
| 5613 |
"logging_steps": 10,
|
|
|
|
| 5627 |
"attributes": {}
|
| 5628 |
}
|
| 5629 |
},
|
| 5630 |
+
"total_flos": 6.771732972133908e+17,
|
| 5631 |
"train_batch_size": 1,
|
| 5632 |
"trial_name": null,
|
| 5633 |
"trial_params": null
|