Training in progress, step 7400, 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:e6f87e6120a7fe7b287ae15ebb9f31c9945509b4496c77d6a58afbe6710011e7
|
| 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 sha256:6fbd8c363f85326f0b0b6fb37a0108e9e398e7936f9f2a7f49df9aaf26f149df
|
| 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 sha256:7ad5f76345e0631018c86addfa807264f5a77319b70133d8bf1d2f2cb2d01443
|
| 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": 56.
|
| 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,
|
|
@@ -13310,6 +13310,24 @@
|
|
| 13310 |
"mean_token_accuracy": 0.6469128876924515,
|
| 13311 |
"num_tokens": 26950520.0,
|
| 13312 |
"step": 7390
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13313 |
}
|
| 13314 |
],
|
| 13315 |
"logging_steps": 5,
|
|
@@ -13329,7 +13347,7 @@
|
|
| 13329 |
"attributes": {}
|
| 13330 |
}
|
| 13331 |
},
|
| 13332 |
-
"total_flos": 1.
|
| 13333 |
"train_batch_size": 4,
|
| 13334 |
"trial_name": null,
|
| 13335 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 56.92307692307692,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 7400,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 13310 |
"mean_token_accuracy": 0.6469128876924515,
|
| 13311 |
"num_tokens": 26950520.0,
|
| 13312 |
"step": 7390
|
| 13313 |
+
},
|
| 13314 |
+
{
|
| 13315 |
+
"epoch": 56.88461538461539,
|
| 13316 |
+
"grad_norm": 0.0,
|
| 13317 |
+
"learning_rate": 1.303e-05,
|
| 13318 |
+
"loss": 1.724,
|
| 13319 |
+
"mean_token_accuracy": 0.6555041134357452,
|
| 13320 |
+
"num_tokens": 26969262.0,
|
| 13321 |
+
"step": 7395
|
| 13322 |
+
},
|
| 13323 |
+
{
|
| 13324 |
+
"epoch": 56.92307692307692,
|
| 13325 |
+
"grad_norm": 0.0,
|
| 13326 |
+
"learning_rate": 1.3005e-05,
|
| 13327 |
+
"loss": 1.7015,
|
| 13328 |
+
"mean_token_accuracy": 0.661999249458313,
|
| 13329 |
+
"num_tokens": 26987507.0,
|
| 13330 |
+
"step": 7400
|
| 13331 |
}
|
| 13332 |
],
|
| 13333 |
"logging_steps": 5,
|
|
|
|
| 13347 |
"attributes": {}
|
| 13348 |
}
|
| 13349 |
},
|
| 13350 |
+
"total_flos": 1.368326926467072e+18,
|
| 13351 |
"train_batch_size": 4,
|
| 13352 |
"trial_name": null,
|
| 13353 |
"trial_params": null
|