Training in progress, step 5470, 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 sha256:49e87031479f844a8dde097f59efd115e12b54b78296813db7031c1eee68d86b
|
| 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:c2df5dd10f52805b5e1f2681a381e9a2d5dfdb270e16d25780f817964f14ba1c
|
| 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 sha256:95ebd518fb8dc5f22decdd733d6528f578f7377e3260204cc055c07ea933f43f
|
| 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": 42.
|
| 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,
|
|
@@ -9836,6 +9836,24 @@
|
|
| 9836 |
"mean_token_accuracy": 0.6474265098571778,
|
| 9837 |
"num_tokens": 19912410.0,
|
| 9838 |
"step": 5460
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9839 |
}
|
| 9840 |
],
|
| 9841 |
"logging_steps": 5,
|
|
@@ -9855,7 +9873,7 @@
|
|
| 9855 |
"attributes": {}
|
| 9856 |
}
|
| 9857 |
},
|
| 9858 |
-
"total_flos": 1.
|
| 9859 |
"train_batch_size": 4,
|
| 9860 |
"trial_name": null,
|
| 9861 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 42.07692307692308,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 5470,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 9836 |
"mean_token_accuracy": 0.6474265098571778,
|
| 9837 |
"num_tokens": 19912410.0,
|
| 9838 |
"step": 5460
|
| 9839 |
+
},
|
| 9840 |
+
{
|
| 9841 |
+
"epoch": 42.03846153846154,
|
| 9842 |
+
"grad_norm": 0.0,
|
| 9843 |
+
"learning_rate": 2.268e-05,
|
| 9844 |
+
"loss": 1.7478,
|
| 9845 |
+
"mean_token_accuracy": 0.646801871061325,
|
| 9846 |
+
"num_tokens": 19930455.0,
|
| 9847 |
+
"step": 5465
|
| 9848 |
+
},
|
| 9849 |
+
{
|
| 9850 |
+
"epoch": 42.07692307692308,
|
| 9851 |
+
"grad_norm": 0.0,
|
| 9852 |
+
"learning_rate": 2.2655000000000002e-05,
|
| 9853 |
+
"loss": 1.8042,
|
| 9854 |
+
"mean_token_accuracy": 0.6378622591495514,
|
| 9855 |
+
"num_tokens": 19948140.0,
|
| 9856 |
+
"step": 5470
|
| 9857 |
}
|
| 9858 |
],
|
| 9859 |
"logging_steps": 5,
|
|
|
|
| 9873 |
"attributes": {}
|
| 9874 |
}
|
| 9875 |
},
|
| 9876 |
+
"total_flos": 1.0108471108062413e+18,
|
| 9877 |
"train_batch_size": 4,
|
| 9878 |
"trial_name": null,
|
| 9879 |
"trial_params": null
|