Training in progress, step 3530, checkpoint
Browse files
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1657
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f9ce6262a4b3ff08115ba8e8f08a9c74f5fea4b5bae3255980926b1a843542b4
|
| 3 |
size 1657
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14709
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a54f7db1a5b83ed41bc98fe2a70e6c4692c07e8927f05dc6fcf2ed715eb8ad6
|
| 3 |
size 14709
|
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 sha256:87efa915b3e6a2b66e6a0ba5eee7ed0f75d7c7db6f2bd23f50b128171e6b248f
|
| 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": 27.
|
| 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,
|
|
@@ -6344,6 +6344,24 @@
|
|
| 6344 |
"mean_token_accuracy": 0.6476831644773483,
|
| 6345 |
"num_tokens": 12837657.0,
|
| 6346 |
"step": 3520
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6347 |
}
|
| 6348 |
],
|
| 6349 |
"logging_steps": 5,
|
|
@@ -6363,7 +6381,7 @@
|
|
| 6363 |
"attributes": {}
|
| 6364 |
}
|
| 6365 |
},
|
| 6366 |
-
"total_flos": 6.
|
| 6367 |
"train_batch_size": 4,
|
| 6368 |
"trial_name": null,
|
| 6369 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 27.153846153846153,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 3530,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 6344 |
"mean_token_accuracy": 0.6476831644773483,
|
| 6345 |
"num_tokens": 12837657.0,
|
| 6346 |
"step": 3520
|
| 6347 |
+
},
|
| 6348 |
+
{
|
| 6349 |
+
"epoch": 27.115384615384617,
|
| 6350 |
+
"grad_norm": 0.0,
|
| 6351 |
+
"learning_rate": 3.238e-05,
|
| 6352 |
+
"loss": 1.7351,
|
| 6353 |
+
"mean_token_accuracy": 0.6554979294538498,
|
| 6354 |
+
"num_tokens": 12855426.0,
|
| 6355 |
+
"step": 3525
|
| 6356 |
+
},
|
| 6357 |
+
{
|
| 6358 |
+
"epoch": 27.153846153846153,
|
| 6359 |
+
"grad_norm": 0.0,
|
| 6360 |
+
"learning_rate": 3.2355000000000004e-05,
|
| 6361 |
+
"loss": 1.6828,
|
| 6362 |
+
"mean_token_accuracy": 0.661261597275734,
|
| 6363 |
+
"num_tokens": 12873752.0,
|
| 6364 |
+
"step": 3530
|
| 6365 |
}
|
| 6366 |
],
|
| 6367 |
"logging_steps": 5,
|
|
|
|
| 6381 |
"attributes": {}
|
| 6382 |
}
|
| 6383 |
},
|
| 6384 |
+
"total_flos": 6.522642564164813e+17,
|
| 6385 |
"train_batch_size": 4,
|
| 6386 |
"trial_name": null,
|
| 6387 |
"trial_params": null
|