Training in progress, step 4750, 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 sha256:9cd22abf52873489fe69f0ac6d4c06c63b0ff128cbab908a7d80797f00eaf8f8
|
| 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 sha256:8c115d6d679b6c5d34e74e3c4d9b86cbaab4d08aa5b541578bacc4c9aff3a9d4
|
| 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:eedc2d3291dbaeb8d92cf355f6681cfa957e1d760c673c59c829b261e8b2c894
|
| 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": 36.
|
| 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,
|
|
@@ -8540,6 +8540,24 @@
|
|
| 8540 |
"mean_token_accuracy": 0.6479008972644806,
|
| 8541 |
"num_tokens": 17285956.0,
|
| 8542 |
"step": 4740
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8543 |
}
|
| 8544 |
],
|
| 8545 |
"logging_steps": 5,
|
|
@@ -8559,7 +8577,7 @@
|
|
| 8559 |
"attributes": {}
|
| 8560 |
}
|
| 8561 |
},
|
| 8562 |
-
"total_flos": 8.
|
| 8563 |
"train_batch_size": 4,
|
| 8564 |
"trial_name": null,
|
| 8565 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 36.53846153846154,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 4750,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 8540 |
"mean_token_accuracy": 0.6479008972644806,
|
| 8541 |
"num_tokens": 17285956.0,
|
| 8542 |
"step": 4740
|
| 8543 |
+
},
|
| 8544 |
+
{
|
| 8545 |
+
"epoch": 36.5,
|
| 8546 |
+
"grad_norm": 0.0,
|
| 8547 |
+
"learning_rate": 2.628e-05,
|
| 8548 |
+
"loss": 1.7722,
|
| 8549 |
+
"mean_token_accuracy": 0.6436032384634018,
|
| 8550 |
+
"num_tokens": 17303770.0,
|
| 8551 |
+
"step": 4745
|
| 8552 |
+
},
|
| 8553 |
+
{
|
| 8554 |
+
"epoch": 36.53846153846154,
|
| 8555 |
+
"grad_norm": 0.0,
|
| 8556 |
+
"learning_rate": 2.6255000000000002e-05,
|
| 8557 |
+
"loss": 1.7904,
|
| 8558 |
+
"mean_token_accuracy": 0.6443703442811965,
|
| 8559 |
+
"num_tokens": 17321120.0,
|
| 8560 |
+
"step": 4750
|
| 8561 |
}
|
| 8562 |
],
|
| 8563 |
"logging_steps": 5,
|
|
|
|
| 8577 |
"attributes": {}
|
| 8578 |
}
|
| 8579 |
},
|
| 8580 |
+
"total_flos": 8.778171787119821e+17,
|
| 8581 |
"train_batch_size": 4,
|
| 8582 |
"trial_name": null,
|
| 8583 |
"trial_params": null
|