Upload 10 files
Browse files- all_results.json +16 -0
- eval_results.json +10 -0
- model.safetensors +1 -1
- train_results.json +9 -0
- trainer_state.json +832 -4
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.5545159532889594,
|
| 4 |
+
"eval_loss": 2.853837490081787,
|
| 5 |
+
"eval_runtime": 100.33,
|
| 6 |
+
"eval_samples": 27721,
|
| 7 |
+
"eval_samples_per_second": 276.298,
|
| 8 |
+
"eval_steps_per_second": 8.641,
|
| 9 |
+
"perplexity": 17.354250987692804,
|
| 10 |
+
"total_flos": 3692222600970240.0,
|
| 11 |
+
"train_loss": 3.0354640246975806,
|
| 12 |
+
"train_runtime": 19788.2132,
|
| 13 |
+
"train_samples": 867980,
|
| 14 |
+
"train_samples_per_second": 131.59,
|
| 15 |
+
"train_steps_per_second": 4.112
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.5545159532889594,
|
| 4 |
+
"eval_loss": 2.853837490081787,
|
| 5 |
+
"eval_runtime": 100.33,
|
| 6 |
+
"eval_samples": 27721,
|
| 7 |
+
"eval_samples_per_second": 276.298,
|
| 8 |
+
"eval_steps_per_second": 8.641,
|
| 9 |
+
"perplexity": 17.354250987692804
|
| 10 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 26690208
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:79e8367cbafae9d789adc9218817dc7e19e5238a8cb24a0e7f8139803803710b
|
| 3 |
size 26690208
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 3692222600970240.0,
|
| 4 |
+
"train_loss": 3.0354640246975806,
|
| 5 |
+
"train_runtime": 19788.2132,
|
| 6 |
+
"train_samples": 867980,
|
| 7 |
+
"train_samples_per_second": 131.59,
|
| 8 |
+
"train_steps_per_second": 4.112
|
| 9 |
+
}
|
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,
|
|
@@ -323,6 +323,834 @@
|
|
| 323 |
"learning_rate": 3.617572964669739e-05,
|
| 324 |
"loss": 3.075,
|
| 325 |
"step": 22500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 326 |
}
|
| 327 |
],
|
| 328 |
"logging_steps": 500,
|
|
@@ -337,12 +1165,12 @@
|
|
| 337 |
"should_evaluate": false,
|
| 338 |
"should_log": false,
|
| 339 |
"should_save": true,
|
| 340 |
-
"should_training_stop":
|
| 341 |
},
|
| 342 |
"attributes": {}
|
| 343 |
}
|
| 344 |
},
|
| 345 |
-
"total_flos":
|
| 346 |
"train_batch_size": 32,
|
| 347 |
"trial_name": null,
|
| 348 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 81375,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 323 |
"learning_rate": 3.617572964669739e-05,
|
| 324 |
"loss": 3.075,
|
| 325 |
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.847926267281106,
|
| 329 |
+
"grad_norm": 2.820953130722046,
|
| 330 |
+
"learning_rate": 3.586850998463902e-05,
|
| 331 |
+
"loss": 3.068,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8663594470046083,
|
| 336 |
+
"grad_norm": 2.321009397506714,
|
| 337 |
+
"learning_rate": 3.556129032258065e-05,
|
| 338 |
+
"loss": 3.0588,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.8847926267281107,
|
| 343 |
+
"grad_norm": 2.54306697845459,
|
| 344 |
+
"learning_rate": 3.525407066052228e-05,
|
| 345 |
+
"loss": 3.0475,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9032258064516129,
|
| 350 |
+
"grad_norm": 2.3935065269470215,
|
| 351 |
+
"learning_rate": 3.4946850998463904e-05,
|
| 352 |
+
"loss": 3.0174,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9216589861751152,
|
| 357 |
+
"grad_norm": 2.3906099796295166,
|
| 358 |
+
"learning_rate": 3.463963133640553e-05,
|
| 359 |
+
"loss": 3.0189,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9400921658986175,
|
| 364 |
+
"grad_norm": 2.480583906173706,
|
| 365 |
+
"learning_rate": 3.433241167434716e-05,
|
| 366 |
+
"loss": 3.0467,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.9585253456221198,
|
| 371 |
+
"grad_norm": 2.1853816509246826,
|
| 372 |
+
"learning_rate": 3.402519201228879e-05,
|
| 373 |
+
"loss": 3.0324,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9769585253456221,
|
| 378 |
+
"grad_norm": 2.525022506713867,
|
| 379 |
+
"learning_rate": 3.371797235023041e-05,
|
| 380 |
+
"loss": 3.0046,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.9953917050691244,
|
| 385 |
+
"grad_norm": 2.310753345489502,
|
| 386 |
+
"learning_rate": 3.3410752688172044e-05,
|
| 387 |
+
"loss": 3.0123,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0138248847926268,
|
| 392 |
+
"grad_norm": 2.827805757522583,
|
| 393 |
+
"learning_rate": 3.3103533026113676e-05,
|
| 394 |
+
"loss": 3.0073,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.032258064516129,
|
| 399 |
+
"grad_norm": 2.4869062900543213,
|
| 400 |
+
"learning_rate": 3.27963133640553e-05,
|
| 401 |
+
"loss": 2.9999,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0506912442396312,
|
| 406 |
+
"grad_norm": 2.9428353309631348,
|
| 407 |
+
"learning_rate": 3.2489093701996933e-05,
|
| 408 |
+
"loss": 3.0065,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0691244239631337,
|
| 413 |
+
"grad_norm": 2.4092352390289307,
|
| 414 |
+
"learning_rate": 3.218187403993856e-05,
|
| 415 |
+
"loss": 2.97,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.087557603686636,
|
| 420 |
+
"grad_norm": 2.185153007507324,
|
| 421 |
+
"learning_rate": 3.1874654377880184e-05,
|
| 422 |
+
"loss": 2.9744,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1059907834101383,
|
| 427 |
+
"grad_norm": 2.547611713409424,
|
| 428 |
+
"learning_rate": 3.1567434715821816e-05,
|
| 429 |
+
"loss": 2.9755,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.1244239631336406,
|
| 434 |
+
"grad_norm": 2.3823814392089844,
|
| 435 |
+
"learning_rate": 3.126021505376344e-05,
|
| 436 |
+
"loss": 2.9597,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1428571428571428,
|
| 441 |
+
"grad_norm": 2.282871961593628,
|
| 442 |
+
"learning_rate": 3.095299539170507e-05,
|
| 443 |
+
"loss": 2.9871,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1612903225806452,
|
| 448 |
+
"grad_norm": 2.517770767211914,
|
| 449 |
+
"learning_rate": 3.06457757296467e-05,
|
| 450 |
+
"loss": 2.9971,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.1797235023041475,
|
| 455 |
+
"grad_norm": 2.8500301837921143,
|
| 456 |
+
"learning_rate": 3.0338556067588324e-05,
|
| 457 |
+
"loss": 2.9692,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.1981566820276497,
|
| 462 |
+
"grad_norm": 2.3024988174438477,
|
| 463 |
+
"learning_rate": 3.0031336405529953e-05,
|
| 464 |
+
"loss": 2.9512,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2165898617511521,
|
| 469 |
+
"grad_norm": 2.4389448165893555,
|
| 470 |
+
"learning_rate": 2.9724116743471585e-05,
|
| 471 |
+
"loss": 2.9743,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2350230414746544,
|
| 476 |
+
"grad_norm": 2.6087846755981445,
|
| 477 |
+
"learning_rate": 2.9416897081413213e-05,
|
| 478 |
+
"loss": 2.9634,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2534562211981566,
|
| 483 |
+
"grad_norm": 2.1963679790496826,
|
| 484 |
+
"learning_rate": 2.9109677419354842e-05,
|
| 485 |
+
"loss": 2.9408,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.271889400921659,
|
| 490 |
+
"grad_norm": 2.6434950828552246,
|
| 491 |
+
"learning_rate": 2.880245775729647e-05,
|
| 492 |
+
"loss": 2.9334,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.2903225806451613,
|
| 497 |
+
"grad_norm": 2.5725350379943848,
|
| 498 |
+
"learning_rate": 2.8495238095238096e-05,
|
| 499 |
+
"loss": 2.9443,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3087557603686637,
|
| 504 |
+
"grad_norm": 2.343334674835205,
|
| 505 |
+
"learning_rate": 2.8188018433179725e-05,
|
| 506 |
+
"loss": 2.9587,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.327188940092166,
|
| 511 |
+
"grad_norm": 2.673114776611328,
|
| 512 |
+
"learning_rate": 2.7880798771121353e-05,
|
| 513 |
+
"loss": 2.956,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.3456221198156681,
|
| 518 |
+
"grad_norm": 2.481757640838623,
|
| 519 |
+
"learning_rate": 2.757357910906298e-05,
|
| 520 |
+
"loss": 2.9332,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3640552995391704,
|
| 525 |
+
"grad_norm": 3.0299792289733887,
|
| 526 |
+
"learning_rate": 2.7266359447004607e-05,
|
| 527 |
+
"loss": 2.947,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.3824884792626728,
|
| 532 |
+
"grad_norm": 3.3357937335968018,
|
| 533 |
+
"learning_rate": 2.6959139784946236e-05,
|
| 534 |
+
"loss": 2.9236,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.400921658986175,
|
| 539 |
+
"grad_norm": 2.214954376220703,
|
| 540 |
+
"learning_rate": 2.6651920122887865e-05,
|
| 541 |
+
"loss": 2.9334,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4193548387096775,
|
| 546 |
+
"grad_norm": 2.7208831310272217,
|
| 547 |
+
"learning_rate": 2.6344700460829497e-05,
|
| 548 |
+
"loss": 2.9001,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4377880184331797,
|
| 553 |
+
"grad_norm": 2.822230577468872,
|
| 554 |
+
"learning_rate": 2.6037480798771125e-05,
|
| 555 |
+
"loss": 2.9563,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.456221198156682,
|
| 560 |
+
"grad_norm": 2.5907464027404785,
|
| 561 |
+
"learning_rate": 2.573026113671275e-05,
|
| 562 |
+
"loss": 2.9352,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.4746543778801844,
|
| 567 |
+
"grad_norm": 2.509422540664673,
|
| 568 |
+
"learning_rate": 2.542304147465438e-05,
|
| 569 |
+
"loss": 2.9388,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.4930875576036866,
|
| 574 |
+
"grad_norm": 2.8918466567993164,
|
| 575 |
+
"learning_rate": 2.5115821812596008e-05,
|
| 576 |
+
"loss": 2.9072,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.511520737327189,
|
| 581 |
+
"grad_norm": 2.3461146354675293,
|
| 582 |
+
"learning_rate": 2.4808602150537637e-05,
|
| 583 |
+
"loss": 2.9013,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5299539170506913,
|
| 588 |
+
"grad_norm": 2.3494646549224854,
|
| 589 |
+
"learning_rate": 2.4501382488479262e-05,
|
| 590 |
+
"loss": 2.8947,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5483870967741935,
|
| 595 |
+
"grad_norm": 2.2246921062469482,
|
| 596 |
+
"learning_rate": 2.419416282642089e-05,
|
| 597 |
+
"loss": 2.9193,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5668202764976957,
|
| 602 |
+
"grad_norm": 2.4895882606506348,
|
| 603 |
+
"learning_rate": 2.3886943164362523e-05,
|
| 604 |
+
"loss": 2.9209,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.5852534562211982,
|
| 609 |
+
"grad_norm": 2.234105110168457,
|
| 610 |
+
"learning_rate": 2.3579723502304148e-05,
|
| 611 |
+
"loss": 2.9104,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6036866359447006,
|
| 616 |
+
"grad_norm": 2.2471518516540527,
|
| 617 |
+
"learning_rate": 2.3272503840245777e-05,
|
| 618 |
+
"loss": 2.8924,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6221198156682028,
|
| 623 |
+
"grad_norm": 2.6903395652770996,
|
| 624 |
+
"learning_rate": 2.2965284178187405e-05,
|
| 625 |
+
"loss": 2.9108,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.640552995391705,
|
| 630 |
+
"grad_norm": 2.5113911628723145,
|
| 631 |
+
"learning_rate": 2.265806451612903e-05,
|
| 632 |
+
"loss": 2.9167,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.6589861751152073,
|
| 637 |
+
"grad_norm": 2.295367956161499,
|
| 638 |
+
"learning_rate": 2.2350844854070663e-05,
|
| 639 |
+
"loss": 2.91,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.6774193548387095,
|
| 644 |
+
"grad_norm": 2.705887794494629,
|
| 645 |
+
"learning_rate": 2.204362519201229e-05,
|
| 646 |
+
"loss": 2.9193,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.695852534562212,
|
| 651 |
+
"grad_norm": 2.490004777908325,
|
| 652 |
+
"learning_rate": 2.1736405529953917e-05,
|
| 653 |
+
"loss": 2.8773,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7142857142857144,
|
| 658 |
+
"grad_norm": 2.564751148223877,
|
| 659 |
+
"learning_rate": 2.1429185867895545e-05,
|
| 660 |
+
"loss": 2.8979,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7327188940092166,
|
| 665 |
+
"grad_norm": 2.5527286529541016,
|
| 666 |
+
"learning_rate": 2.1121966205837174e-05,
|
| 667 |
+
"loss": 2.9022,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.7511520737327189,
|
| 672 |
+
"grad_norm": 2.6402347087860107,
|
| 673 |
+
"learning_rate": 2.0814746543778803e-05,
|
| 674 |
+
"loss": 2.8751,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.769585253456221,
|
| 679 |
+
"grad_norm": 2.415748357772827,
|
| 680 |
+
"learning_rate": 2.050752688172043e-05,
|
| 681 |
+
"loss": 2.8786,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.7880184331797235,
|
| 686 |
+
"grad_norm": 2.6750245094299316,
|
| 687 |
+
"learning_rate": 2.020030721966206e-05,
|
| 688 |
+
"loss": 2.8853,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8064516129032258,
|
| 693 |
+
"grad_norm": 2.4245858192443848,
|
| 694 |
+
"learning_rate": 1.989308755760369e-05,
|
| 695 |
+
"loss": 2.875,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8248847926267282,
|
| 700 |
+
"grad_norm": 2.660170078277588,
|
| 701 |
+
"learning_rate": 1.9585867895545314e-05,
|
| 702 |
+
"loss": 2.8737,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8433179723502304,
|
| 707 |
+
"grad_norm": 2.4194977283477783,
|
| 708 |
+
"learning_rate": 1.9278648233486943e-05,
|
| 709 |
+
"loss": 2.8654,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8617511520737327,
|
| 714 |
+
"grad_norm": 2.4706435203552246,
|
| 715 |
+
"learning_rate": 1.8971428571428575e-05,
|
| 716 |
+
"loss": 2.8302,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.8801843317972349,
|
| 721 |
+
"grad_norm": 2.8965485095977783,
|
| 722 |
+
"learning_rate": 1.86642089093702e-05,
|
| 723 |
+
"loss": 2.874,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.8986175115207373,
|
| 728 |
+
"grad_norm": 2.812009811401367,
|
| 729 |
+
"learning_rate": 1.835698924731183e-05,
|
| 730 |
+
"loss": 2.8833,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.9170506912442398,
|
| 735 |
+
"grad_norm": 2.7252895832061768,
|
| 736 |
+
"learning_rate": 1.8049769585253457e-05,
|
| 737 |
+
"loss": 2.8639,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.935483870967742,
|
| 742 |
+
"grad_norm": 2.5407986640930176,
|
| 743 |
+
"learning_rate": 1.7742549923195083e-05,
|
| 744 |
+
"loss": 2.8637,
|
| 745 |
+
"step": 52500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9539170506912442,
|
| 749 |
+
"grad_norm": 2.5381031036376953,
|
| 750 |
+
"learning_rate": 1.7435330261136715e-05,
|
| 751 |
+
"loss": 2.8636,
|
| 752 |
+
"step": 53000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.9723502304147464,
|
| 756 |
+
"grad_norm": 2.5974888801574707,
|
| 757 |
+
"learning_rate": 1.7128110599078343e-05,
|
| 758 |
+
"loss": 2.8563,
|
| 759 |
+
"step": 53500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.9907834101382489,
|
| 763 |
+
"grad_norm": 2.5476796627044678,
|
| 764 |
+
"learning_rate": 1.682089093701997e-05,
|
| 765 |
+
"loss": 2.8651,
|
| 766 |
+
"step": 54000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.0092165898617513,
|
| 770 |
+
"grad_norm": 2.4536616802215576,
|
| 771 |
+
"learning_rate": 1.6513671274961597e-05,
|
| 772 |
+
"loss": 2.8788,
|
| 773 |
+
"step": 54500
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 2.0276497695852536,
|
| 777 |
+
"grad_norm": 3.109189510345459,
|
| 778 |
+
"learning_rate": 1.6206451612903226e-05,
|
| 779 |
+
"loss": 2.8355,
|
| 780 |
+
"step": 55000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.046082949308756,
|
| 784 |
+
"grad_norm": 2.727445363998413,
|
| 785 |
+
"learning_rate": 1.5899231950844855e-05,
|
| 786 |
+
"loss": 2.8582,
|
| 787 |
+
"step": 55500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 2.064516129032258,
|
| 791 |
+
"grad_norm": 2.809833288192749,
|
| 792 |
+
"learning_rate": 1.5592012288786483e-05,
|
| 793 |
+
"loss": 2.8763,
|
| 794 |
+
"step": 56000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.0829493087557602,
|
| 798 |
+
"grad_norm": 2.9285683631896973,
|
| 799 |
+
"learning_rate": 1.5284792626728112e-05,
|
| 800 |
+
"loss": 2.8524,
|
| 801 |
+
"step": 56500
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.1013824884792625,
|
| 805 |
+
"grad_norm": 2.599776268005371,
|
| 806 |
+
"learning_rate": 1.4977572964669739e-05,
|
| 807 |
+
"loss": 2.8235,
|
| 808 |
+
"step": 57000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.119815668202765,
|
| 812 |
+
"grad_norm": 2.367570638656616,
|
| 813 |
+
"learning_rate": 1.4670353302611368e-05,
|
| 814 |
+
"loss": 2.8624,
|
| 815 |
+
"step": 57500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.1382488479262673,
|
| 819 |
+
"grad_norm": 2.971496820449829,
|
| 820 |
+
"learning_rate": 1.4363133640552995e-05,
|
| 821 |
+
"loss": 2.8542,
|
| 822 |
+
"step": 58000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.1566820276497696,
|
| 826 |
+
"grad_norm": 2.530744791030884,
|
| 827 |
+
"learning_rate": 1.4055913978494625e-05,
|
| 828 |
+
"loss": 2.878,
|
| 829 |
+
"step": 58500
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.175115207373272,
|
| 833 |
+
"grad_norm": 2.3559248447418213,
|
| 834 |
+
"learning_rate": 1.3748694316436254e-05,
|
| 835 |
+
"loss": 2.8378,
|
| 836 |
+
"step": 59000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.193548387096774,
|
| 840 |
+
"grad_norm": 2.6017301082611084,
|
| 841 |
+
"learning_rate": 1.344147465437788e-05,
|
| 842 |
+
"loss": 2.8526,
|
| 843 |
+
"step": 59500
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.2119815668202767,
|
| 847 |
+
"grad_norm": 2.727224349975586,
|
| 848 |
+
"learning_rate": 1.313425499231951e-05,
|
| 849 |
+
"loss": 2.8457,
|
| 850 |
+
"step": 60000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.230414746543779,
|
| 854 |
+
"grad_norm": 2.7515804767608643,
|
| 855 |
+
"learning_rate": 1.2827035330261136e-05,
|
| 856 |
+
"loss": 2.8422,
|
| 857 |
+
"step": 60500
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.248847926267281,
|
| 861 |
+
"grad_norm": 2.1259450912475586,
|
| 862 |
+
"learning_rate": 1.2519815668202767e-05,
|
| 863 |
+
"loss": 2.8552,
|
| 864 |
+
"step": 61000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.2672811059907834,
|
| 868 |
+
"grad_norm": 2.3828954696655273,
|
| 869 |
+
"learning_rate": 1.2212596006144395e-05,
|
| 870 |
+
"loss": 2.833,
|
| 871 |
+
"step": 61500
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.2857142857142856,
|
| 875 |
+
"grad_norm": 2.588263988494873,
|
| 876 |
+
"learning_rate": 1.1905376344086022e-05,
|
| 877 |
+
"loss": 2.8259,
|
| 878 |
+
"step": 62000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.3041474654377883,
|
| 882 |
+
"grad_norm": 2.4910430908203125,
|
| 883 |
+
"learning_rate": 1.159815668202765e-05,
|
| 884 |
+
"loss": 2.8322,
|
| 885 |
+
"step": 62500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.3225806451612905,
|
| 889 |
+
"grad_norm": 2.4441442489624023,
|
| 890 |
+
"learning_rate": 1.129093701996928e-05,
|
| 891 |
+
"loss": 2.8368,
|
| 892 |
+
"step": 63000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.3410138248847927,
|
| 896 |
+
"grad_norm": 2.8292665481567383,
|
| 897 |
+
"learning_rate": 1.0983717357910907e-05,
|
| 898 |
+
"loss": 2.8541,
|
| 899 |
+
"step": 63500
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.359447004608295,
|
| 903 |
+
"grad_norm": 3.0737709999084473,
|
| 904 |
+
"learning_rate": 1.0676497695852535e-05,
|
| 905 |
+
"loss": 2.8302,
|
| 906 |
+
"step": 64000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.377880184331797,
|
| 910 |
+
"grad_norm": 2.6240386962890625,
|
| 911 |
+
"learning_rate": 1.0369278033794164e-05,
|
| 912 |
+
"loss": 2.8581,
|
| 913 |
+
"step": 64500
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.3963133640552994,
|
| 917 |
+
"grad_norm": 2.7089595794677734,
|
| 918 |
+
"learning_rate": 1.0062058371735791e-05,
|
| 919 |
+
"loss": 2.8291,
|
| 920 |
+
"step": 65000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.4147465437788016,
|
| 924 |
+
"grad_norm": 2.5590097904205322,
|
| 925 |
+
"learning_rate": 9.754838709677421e-06,
|
| 926 |
+
"loss": 2.8423,
|
| 927 |
+
"step": 65500
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.4331797235023043,
|
| 931 |
+
"grad_norm": 2.529129981994629,
|
| 932 |
+
"learning_rate": 9.447619047619048e-06,
|
| 933 |
+
"loss": 2.8313,
|
| 934 |
+
"step": 66000
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.4516129032258065,
|
| 938 |
+
"grad_norm": 2.3878464698791504,
|
| 939 |
+
"learning_rate": 9.140399385560675e-06,
|
| 940 |
+
"loss": 2.8396,
|
| 941 |
+
"step": 66500
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.4700460829493087,
|
| 945 |
+
"grad_norm": 2.324528217315674,
|
| 946 |
+
"learning_rate": 8.833179723502306e-06,
|
| 947 |
+
"loss": 2.8406,
|
| 948 |
+
"step": 67000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.488479262672811,
|
| 952 |
+
"grad_norm": 2.531818389892578,
|
| 953 |
+
"learning_rate": 8.525960061443933e-06,
|
| 954 |
+
"loss": 2.8283,
|
| 955 |
+
"step": 67500
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.506912442396313,
|
| 959 |
+
"grad_norm": 2.370063066482544,
|
| 960 |
+
"learning_rate": 8.218740399385561e-06,
|
| 961 |
+
"loss": 2.8218,
|
| 962 |
+
"step": 68000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.525345622119816,
|
| 966 |
+
"grad_norm": 2.7173168659210205,
|
| 967 |
+
"learning_rate": 7.91152073732719e-06,
|
| 968 |
+
"loss": 2.8193,
|
| 969 |
+
"step": 68500
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.543778801843318,
|
| 973 |
+
"grad_norm": 2.893047571182251,
|
| 974 |
+
"learning_rate": 7.604301075268818e-06,
|
| 975 |
+
"loss": 2.8789,
|
| 976 |
+
"step": 69000
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.5622119815668203,
|
| 980 |
+
"grad_norm": 2.3326709270477295,
|
| 981 |
+
"learning_rate": 7.297081413210446e-06,
|
| 982 |
+
"loss": 2.82,
|
| 983 |
+
"step": 69500
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.5806451612903225,
|
| 987 |
+
"grad_norm": 2.6681976318359375,
|
| 988 |
+
"learning_rate": 6.989861751152074e-06,
|
| 989 |
+
"loss": 2.8304,
|
| 990 |
+
"step": 70000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.5990783410138247,
|
| 994 |
+
"grad_norm": 2.398226261138916,
|
| 995 |
+
"learning_rate": 6.682642089093702e-06,
|
| 996 |
+
"loss": 2.8588,
|
| 997 |
+
"step": 70500
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.6175115207373274,
|
| 1001 |
+
"grad_norm": 2.898515462875366,
|
| 1002 |
+
"learning_rate": 6.375422427035331e-06,
|
| 1003 |
+
"loss": 2.8197,
|
| 1004 |
+
"step": 71000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.6359447004608296,
|
| 1008 |
+
"grad_norm": 2.739598035812378,
|
| 1009 |
+
"learning_rate": 6.0682027649769585e-06,
|
| 1010 |
+
"loss": 2.8118,
|
| 1011 |
+
"step": 71500
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.654377880184332,
|
| 1015 |
+
"grad_norm": 2.643958806991577,
|
| 1016 |
+
"learning_rate": 5.760983102918587e-06,
|
| 1017 |
+
"loss": 2.8237,
|
| 1018 |
+
"step": 72000
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.672811059907834,
|
| 1022 |
+
"grad_norm": 2.4359323978424072,
|
| 1023 |
+
"learning_rate": 5.453763440860216e-06,
|
| 1024 |
+
"loss": 2.849,
|
| 1025 |
+
"step": 72500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.6912442396313363,
|
| 1029 |
+
"grad_norm": 2.789459228515625,
|
| 1030 |
+
"learning_rate": 5.146543778801844e-06,
|
| 1031 |
+
"loss": 2.8295,
|
| 1032 |
+
"step": 73000
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.709677419354839,
|
| 1036 |
+
"grad_norm": 2.7510650157928467,
|
| 1037 |
+
"learning_rate": 4.8393241167434715e-06,
|
| 1038 |
+
"loss": 2.8261,
|
| 1039 |
+
"step": 73500
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.7281105990783407,
|
| 1043 |
+
"grad_norm": 2.687920570373535,
|
| 1044 |
+
"learning_rate": 4.5321044546851e-06,
|
| 1045 |
+
"loss": 2.8247,
|
| 1046 |
+
"step": 74000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.7465437788018434,
|
| 1050 |
+
"grad_norm": 2.563568592071533,
|
| 1051 |
+
"learning_rate": 4.224884792626729e-06,
|
| 1052 |
+
"loss": 2.8077,
|
| 1053 |
+
"step": 74500
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.7649769585253456,
|
| 1057 |
+
"grad_norm": 2.5233335494995117,
|
| 1058 |
+
"learning_rate": 3.917665130568357e-06,
|
| 1059 |
+
"loss": 2.8249,
|
| 1060 |
+
"step": 75000
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.783410138248848,
|
| 1064 |
+
"grad_norm": 2.1638145446777344,
|
| 1065 |
+
"learning_rate": 3.610445468509985e-06,
|
| 1066 |
+
"loss": 2.8117,
|
| 1067 |
+
"step": 75500
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.80184331797235,
|
| 1071 |
+
"grad_norm": 2.3863344192504883,
|
| 1072 |
+
"learning_rate": 3.303225806451613e-06,
|
| 1073 |
+
"loss": 2.8267,
|
| 1074 |
+
"step": 76000
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.8202764976958523,
|
| 1078 |
+
"grad_norm": 2.6310081481933594,
|
| 1079 |
+
"learning_rate": 2.9960061443932414e-06,
|
| 1080 |
+
"loss": 2.7984,
|
| 1081 |
+
"step": 76500
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.838709677419355,
|
| 1085 |
+
"grad_norm": 2.5833451747894287,
|
| 1086 |
+
"learning_rate": 2.6887864823348697e-06,
|
| 1087 |
+
"loss": 2.828,
|
| 1088 |
+
"step": 77000
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.857142857142857,
|
| 1092 |
+
"grad_norm": 2.796090602874756,
|
| 1093 |
+
"learning_rate": 2.381566820276498e-06,
|
| 1094 |
+
"loss": 2.8328,
|
| 1095 |
+
"step": 77500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.8755760368663594,
|
| 1099 |
+
"grad_norm": 2.4523568153381348,
|
| 1100 |
+
"learning_rate": 2.074347158218126e-06,
|
| 1101 |
+
"loss": 2.8262,
|
| 1102 |
+
"step": 78000
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.8940092165898617,
|
| 1106 |
+
"grad_norm": 2.7820804119110107,
|
| 1107 |
+
"learning_rate": 1.7671274961597542e-06,
|
| 1108 |
+
"loss": 2.8492,
|
| 1109 |
+
"step": 78500
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.912442396313364,
|
| 1113 |
+
"grad_norm": 2.6446750164031982,
|
| 1114 |
+
"learning_rate": 1.4599078341013825e-06,
|
| 1115 |
+
"loss": 2.8004,
|
| 1116 |
+
"step": 79000
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.9308755760368665,
|
| 1120 |
+
"grad_norm": 2.688173532485962,
|
| 1121 |
+
"learning_rate": 1.1526881720430107e-06,
|
| 1122 |
+
"loss": 2.8324,
|
| 1123 |
+
"step": 79500
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.9493087557603688,
|
| 1127 |
+
"grad_norm": 2.5094845294952393,
|
| 1128 |
+
"learning_rate": 8.454685099846391e-07,
|
| 1129 |
+
"loss": 2.8299,
|
| 1130 |
+
"step": 80000
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.967741935483871,
|
| 1134 |
+
"grad_norm": 3.0129997730255127,
|
| 1135 |
+
"learning_rate": 5.382488479262673e-07,
|
| 1136 |
+
"loss": 2.8397,
|
| 1137 |
+
"step": 80500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.986175115207373,
|
| 1141 |
+
"grad_norm": 2.273484706878662,
|
| 1142 |
+
"learning_rate": 2.3102918586789556e-07,
|
| 1143 |
+
"loss": 2.8117,
|
| 1144 |
+
"step": 81000
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 3.0,
|
| 1148 |
+
"step": 81375,
|
| 1149 |
+
"total_flos": 3692222600970240.0,
|
| 1150 |
+
"train_loss": 3.0354640246975806,
|
| 1151 |
+
"train_runtime": 19788.2132,
|
| 1152 |
+
"train_samples_per_second": 131.59,
|
| 1153 |
+
"train_steps_per_second": 4.112
|
| 1154 |
}
|
| 1155 |
],
|
| 1156 |
"logging_steps": 500,
|
|
|
|
| 1165 |
"should_evaluate": false,
|
| 1166 |
"should_log": false,
|
| 1167 |
"should_save": true,
|
| 1168 |
+
"should_training_stop": true
|
| 1169 |
},
|
| 1170 |
"attributes": {}
|
| 1171 |
}
|
| 1172 |
},
|
| 1173 |
+
"total_flos": 3692222600970240.0,
|
| 1174 |
"train_batch_size": 32,
|
| 1175 |
"trial_name": null,
|
| 1176 |
"trial_params": null
|