Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1267 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1267 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.988603988603988,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1751,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02849002849002849,
|
| 13 |
+
"grad_norm": 4.866657257080078,
|
| 14 |
+
"learning_rate": 2.272727272727273e-05,
|
| 15 |
+
"loss": 1.0817,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05698005698005698,
|
| 20 |
+
"grad_norm": 2.505408525466919,
|
| 21 |
+
"learning_rate": 4.545454545454546e-05,
|
| 22 |
+
"loss": 0.553,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.08547008547008547,
|
| 27 |
+
"grad_norm": 1.517915964126587,
|
| 28 |
+
"learning_rate": 6.818181818181818e-05,
|
| 29 |
+
"loss": 0.3016,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.11396011396011396,
|
| 34 |
+
"grad_norm": 1.582049012184143,
|
| 35 |
+
"learning_rate": 9.090909090909092e-05,
|
| 36 |
+
"loss": 0.2392,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.14245014245014245,
|
| 41 |
+
"grad_norm": 1.7641891241073608,
|
| 42 |
+
"learning_rate": 0.00011363636363636365,
|
| 43 |
+
"loss": 0.2087,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.17094017094017094,
|
| 48 |
+
"grad_norm": 1.5408800840377808,
|
| 49 |
+
"learning_rate": 0.00013636363636363637,
|
| 50 |
+
"loss": 0.2057,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.19943019943019943,
|
| 55 |
+
"grad_norm": 1.1511545181274414,
|
| 56 |
+
"learning_rate": 0.0001590909090909091,
|
| 57 |
+
"loss": 0.1738,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.22792022792022792,
|
| 62 |
+
"grad_norm": 1.0401157140731812,
|
| 63 |
+
"learning_rate": 0.00018181818181818183,
|
| 64 |
+
"loss": 0.1387,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2564102564102564,
|
| 69 |
+
"grad_norm": 0.794109582901001,
|
| 70 |
+
"learning_rate": 0.0001999992862522931,
|
| 71 |
+
"loss": 0.1362,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2849002849002849,
|
| 76 |
+
"grad_norm": 1.0705766677856445,
|
| 77 |
+
"learning_rate": 0.00019997430615234976,
|
| 78 |
+
"loss": 0.1311,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.31339031339031337,
|
| 83 |
+
"grad_norm": 0.7216269373893738,
|
| 84 |
+
"learning_rate": 0.0001999136488551224,
|
| 85 |
+
"loss": 0.1267,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3418803418803419,
|
| 90 |
+
"grad_norm": 0.9830422401428223,
|
| 91 |
+
"learning_rate": 0.00019981733600699645,
|
| 92 |
+
"loss": 0.1186,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.37037037037037035,
|
| 97 |
+
"grad_norm": 0.8986383676528931,
|
| 98 |
+
"learning_rate": 0.00019968540197852787,
|
| 99 |
+
"loss": 0.1082,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.39886039886039887,
|
| 104 |
+
"grad_norm": 0.9030706286430359,
|
| 105 |
+
"learning_rate": 0.00019951789385217757,
|
| 106 |
+
"loss": 0.1081,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.42735042735042733,
|
| 111 |
+
"grad_norm": 0.8084211349487305,
|
| 112 |
+
"learning_rate": 0.00019931487140550935,
|
| 113 |
+
"loss": 0.1159,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.45584045584045585,
|
| 118 |
+
"grad_norm": 0.4105013310909271,
|
| 119 |
+
"learning_rate": 0.00019907640708985766,
|
| 120 |
+
"loss": 0.1002,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4843304843304843,
|
| 125 |
+
"grad_norm": 0.6791099905967712,
|
| 126 |
+
"learning_rate": 0.0001988025860044721,
|
| 127 |
+
"loss": 0.0992,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5128205128205128,
|
| 132 |
+
"grad_norm": 0.611254870891571,
|
| 133 |
+
"learning_rate": 0.00019849350586614866,
|
| 134 |
+
"loss": 0.0741,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5413105413105413,
|
| 139 |
+
"grad_norm": 0.49187201261520386,
|
| 140 |
+
"learning_rate": 0.00019814927697435827,
|
| 141 |
+
"loss": 0.0913,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5698005698005698,
|
| 146 |
+
"grad_norm": 0.9148091077804565,
|
| 147 |
+
"learning_rate": 0.00019777002217188482,
|
| 148 |
+
"loss": 0.0858,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5982905982905983,
|
| 153 |
+
"grad_norm": 0.6754726767539978,
|
| 154 |
+
"learning_rate": 0.000197355876800987,
|
| 155 |
+
"loss": 0.0878,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6267806267806267,
|
| 160 |
+
"grad_norm": 0.5443580150604248,
|
| 161 |
+
"learning_rate": 0.00019690698865509966,
|
| 162 |
+
"loss": 0.0816,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6552706552706553,
|
| 167 |
+
"grad_norm": 0.5719950795173645,
|
| 168 |
+
"learning_rate": 0.00019642351792609165,
|
| 169 |
+
"loss": 0.0818,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.6837606837606838,
|
| 174 |
+
"grad_norm": 0.570486843585968,
|
| 175 |
+
"learning_rate": 0.00019590563714709918,
|
| 176 |
+
"loss": 0.089,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7122507122507122,
|
| 181 |
+
"grad_norm": 0.5645405054092407,
|
| 182 |
+
"learning_rate": 0.00019535353113095494,
|
| 183 |
+
"loss": 0.0748,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7407407407407407,
|
| 188 |
+
"grad_norm": 0.4007212817668915,
|
| 189 |
+
"learning_rate": 0.00019476739690423532,
|
| 190 |
+
"loss": 0.0735,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.7692307692307693,
|
| 195 |
+
"grad_norm": 0.5221670269966125,
|
| 196 |
+
"learning_rate": 0.00019414744363694845,
|
| 197 |
+
"loss": 0.0702,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7977207977207977,
|
| 202 |
+
"grad_norm": 0.4181097447872162,
|
| 203 |
+
"learning_rate": 0.00019349389256788943,
|
| 204 |
+
"loss": 0.0927,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8262108262108262,
|
| 209 |
+
"grad_norm": 0.42916539311408997,
|
| 210 |
+
"learning_rate": 0.0001928069769256879,
|
| 211 |
+
"loss": 0.0662,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.8547008547008547,
|
| 216 |
+
"grad_norm": 0.4586508572101593,
|
| 217 |
+
"learning_rate": 0.00019208694184557736,
|
| 218 |
+
"loss": 0.0757,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8831908831908832,
|
| 223 |
+
"grad_norm": 0.49600309133529663,
|
| 224 |
+
"learning_rate": 0.00019133404428191533,
|
| 225 |
+
"loss": 0.0641,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9116809116809117,
|
| 230 |
+
"grad_norm": 0.5553969740867615,
|
| 231 |
+
"learning_rate": 0.00019054855291648562,
|
| 232 |
+
"loss": 0.0733,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.9401709401709402,
|
| 237 |
+
"grad_norm": 0.508593738079071,
|
| 238 |
+
"learning_rate": 0.00018973074806261558,
|
| 239 |
+
"loss": 0.0747,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.9686609686609686,
|
| 244 |
+
"grad_norm": 0.5126791000366211,
|
| 245 |
+
"learning_rate": 0.00018888092156514255,
|
| 246 |
+
"loss": 0.0563,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9971509971509972,
|
| 251 |
+
"grad_norm": 0.6333485841751099,
|
| 252 |
+
"learning_rate": 0.00018799937669626484,
|
| 253 |
+
"loss": 0.059,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0256410256410255,
|
| 258 |
+
"grad_norm": 0.5621979236602783,
|
| 259 |
+
"learning_rate": 0.00018708642804731517,
|
| 260 |
+
"loss": 0.0533,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.0541310541310542,
|
| 265 |
+
"grad_norm": 0.25756293535232544,
|
| 266 |
+
"learning_rate": 0.0001861424014164941,
|
| 267 |
+
"loss": 0.0511,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.0826210826210827,
|
| 272 |
+
"grad_norm": 0.4778578281402588,
|
| 273 |
+
"learning_rate": 0.00018516763369260493,
|
| 274 |
+
"loss": 0.061,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1111111111111112,
|
| 279 |
+
"grad_norm": 0.4505862295627594,
|
| 280 |
+
"learning_rate": 0.00018416247273482988,
|
| 281 |
+
"loss": 0.0707,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.1396011396011396,
|
| 286 |
+
"grad_norm": 0.4163416624069214,
|
| 287 |
+
"learning_rate": 0.0001831272772485922,
|
| 288 |
+
"loss": 0.0514,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.168091168091168,
|
| 293 |
+
"grad_norm": 0.44356927275657654,
|
| 294 |
+
"learning_rate": 0.00018206241665754688,
|
| 295 |
+
"loss": 0.057,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.1965811965811965,
|
| 300 |
+
"grad_norm": 0.2967655658721924,
|
| 301 |
+
"learning_rate": 0.000180968270971747,
|
| 302 |
+
"loss": 0.0641,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.225071225071225,
|
| 307 |
+
"grad_norm": 0.47910967469215393,
|
| 308 |
+
"learning_rate": 0.00017984523065203188,
|
| 309 |
+
"loss": 0.0618,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.2535612535612537,
|
| 314 |
+
"grad_norm": 0.3292757570743561,
|
| 315 |
+
"learning_rate": 0.0001786936964706858,
|
| 316 |
+
"loss": 0.0517,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.282051282051282,
|
| 321 |
+
"grad_norm": 0.37650346755981445,
|
| 322 |
+
"learning_rate": 0.00017751407936841688,
|
| 323 |
+
"loss": 0.0589,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.3105413105413106,
|
| 328 |
+
"grad_norm": 0.4217515289783478,
|
| 329 |
+
"learning_rate": 0.00017630680030770735,
|
| 330 |
+
"loss": 0.0564,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.339031339031339,
|
| 335 |
+
"grad_norm": 0.3700660467147827,
|
| 336 |
+
"learning_rate": 0.00017507229012258732,
|
| 337 |
+
"loss": 0.0588,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.3675213675213675,
|
| 342 |
+
"grad_norm": 0.2905580401420593,
|
| 343 |
+
"learning_rate": 0.00017381098936488574,
|
| 344 |
+
"loss": 0.0464,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.396011396011396,
|
| 349 |
+
"grad_norm": 0.39054426550865173,
|
| 350 |
+
"learning_rate": 0.00017252334814701353,
|
| 351 |
+
"loss": 0.0578,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.4245014245014245,
|
| 356 |
+
"grad_norm": 0.3755801320075989,
|
| 357 |
+
"learning_rate": 0.00017120982598133456,
|
| 358 |
+
"loss": 0.0455,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.452991452991453,
|
| 363 |
+
"grad_norm": 0.3761601448059082,
|
| 364 |
+
"learning_rate": 0.0001698708916161829,
|
| 365 |
+
"loss": 0.0471,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.4814814814814814,
|
| 370 |
+
"grad_norm": 0.3036906123161316,
|
| 371 |
+
"learning_rate": 0.000168507022868583,
|
| 372 |
+
"loss": 0.0563,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.50997150997151,
|
| 377 |
+
"grad_norm": 0.39106476306915283,
|
| 378 |
+
"learning_rate": 0.0001671187064537345,
|
| 379 |
+
"loss": 0.0587,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.5384615384615383,
|
| 384 |
+
"grad_norm": 0.6104059219360352,
|
| 385 |
+
"learning_rate": 0.00016570643781132122,
|
| 386 |
+
"loss": 0.0612,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.566951566951567,
|
| 391 |
+
"grad_norm": 0.3543621003627777,
|
| 392 |
+
"learning_rate": 0.00016427072092870651,
|
| 393 |
+
"loss": 0.0507,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.5954415954415955,
|
| 398 |
+
"grad_norm": 0.4155671000480652,
|
| 399 |
+
"learning_rate": 0.0001628120681610789,
|
| 400 |
+
"loss": 0.0545,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.623931623931624,
|
| 405 |
+
"grad_norm": 0.318392276763916,
|
| 406 |
+
"learning_rate": 0.00016133100004861082,
|
| 407 |
+
"loss": 0.0563,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.6524216524216524,
|
| 412 |
+
"grad_norm": 0.34228989481925964,
|
| 413 |
+
"learning_rate": 0.00015982804513069664,
|
| 414 |
+
"loss": 0.0629,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.6809116809116809,
|
| 419 |
+
"grad_norm": 0.3822249472141266,
|
| 420 |
+
"learning_rate": 0.0001583037397573366,
|
| 421 |
+
"loss": 0.0473,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.7094017094017095,
|
| 426 |
+
"grad_norm": 0.3222091794013977,
|
| 427 |
+
"learning_rate": 0.00015675862789773243,
|
| 428 |
+
"loss": 0.0465,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.7378917378917378,
|
| 433 |
+
"grad_norm": 0.4308369755744934,
|
| 434 |
+
"learning_rate": 0.00015519326094616507,
|
| 435 |
+
"loss": 0.0538,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.7663817663817665,
|
| 440 |
+
"grad_norm": 0.33870652318000793,
|
| 441 |
+
"learning_rate": 0.00015360819752522164,
|
| 442 |
+
"loss": 0.0575,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.7948717948717947,
|
| 447 |
+
"grad_norm": 0.2933538258075714,
|
| 448 |
+
"learning_rate": 0.00015200400328644356,
|
| 449 |
+
"loss": 0.0411,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.8233618233618234,
|
| 454 |
+
"grad_norm": 0.39347216486930847,
|
| 455 |
+
"learning_rate": 0.000150381250708466,
|
| 456 |
+
"loss": 0.042,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.8518518518518519,
|
| 461 |
+
"grad_norm": 0.2737712562084198,
|
| 462 |
+
"learning_rate": 0.00014874051889272107,
|
| 463 |
+
"loss": 0.0497,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.8803418803418803,
|
| 468 |
+
"grad_norm": 0.23161457479000092,
|
| 469 |
+
"learning_rate": 0.0001470823933567776,
|
| 470 |
+
"loss": 0.044,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.9088319088319088,
|
| 475 |
+
"grad_norm": 0.25970926880836487,
|
| 476 |
+
"learning_rate": 0.00014540746582539108,
|
| 477 |
+
"loss": 0.04,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.9373219373219372,
|
| 482 |
+
"grad_norm": 0.33561354875564575,
|
| 483 |
+
"learning_rate": 0.00014371633401933872,
|
| 484 |
+
"loss": 0.0495,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.965811965811966,
|
| 489 |
+
"grad_norm": 0.2616577744483948,
|
| 490 |
+
"learning_rate": 0.0001420096014421146,
|
| 491 |
+
"loss": 0.0411,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.9943019943019942,
|
| 496 |
+
"grad_norm": 0.3040335774421692,
|
| 497 |
+
"learning_rate": 0.0001402878771645611,
|
| 498 |
+
"loss": 0.0418,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.022792022792023,
|
| 503 |
+
"grad_norm": 0.4659029245376587,
|
| 504 |
+
"learning_rate": 0.00013855177560751376,
|
| 505 |
+
"loss": 0.0465,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.051282051282051,
|
| 510 |
+
"grad_norm": 0.36660319566726685,
|
| 511 |
+
"learning_rate": 0.00013680191632253682,
|
| 512 |
+
"loss": 0.0441,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.07977207977208,
|
| 517 |
+
"grad_norm": 0.3704872131347656,
|
| 518 |
+
"learning_rate": 0.00013503892377082763,
|
| 519 |
+
"loss": 0.0578,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.1082621082621085,
|
| 524 |
+
"grad_norm": 0.26848968863487244,
|
| 525 |
+
"learning_rate": 0.00013326342710036934,
|
| 526 |
+
"loss": 0.0493,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.1367521367521367,
|
| 531 |
+
"grad_norm": 0.23576417565345764,
|
| 532 |
+
"learning_rate": 0.00013147605992141066,
|
| 533 |
+
"loss": 0.0389,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.1652421652421654,
|
| 538 |
+
"grad_norm": 0.31747928261756897,
|
| 539 |
+
"learning_rate": 0.00012967746008035348,
|
| 540 |
+
"loss": 0.0439,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.1937321937321936,
|
| 545 |
+
"grad_norm": 0.41468170285224915,
|
| 546 |
+
"learning_rate": 0.00012786826943212874,
|
| 547 |
+
"loss": 0.0432,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.2222222222222223,
|
| 552 |
+
"grad_norm": 0.27693742513656616,
|
| 553 |
+
"learning_rate": 0.00012604913361114148,
|
| 554 |
+
"loss": 0.0433,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.2507122507122506,
|
| 559 |
+
"grad_norm": 0.31801000237464905,
|
| 560 |
+
"learning_rate": 0.00012422070180086772,
|
| 561 |
+
"loss": 0.0312,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.2792022792022792,
|
| 566 |
+
"grad_norm": 0.5060178637504578,
|
| 567 |
+
"learning_rate": 0.0001223836265021841,
|
| 568 |
+
"loss": 0.0474,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.3076923076923075,
|
| 573 |
+
"grad_norm": 0.4824029803276062,
|
| 574 |
+
"learning_rate": 0.0001205385633005144,
|
| 575 |
+
"loss": 0.0354,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.336182336182336,
|
| 580 |
+
"grad_norm": 0.36395755410194397,
|
| 581 |
+
"learning_rate": 0.00011868617063187462,
|
| 582 |
+
"loss": 0.0485,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.364672364672365,
|
| 587 |
+
"grad_norm": 0.43524518609046936,
|
| 588 |
+
"learning_rate": 0.0001168271095479012,
|
| 589 |
+
"loss": 0.0433,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.393162393162393,
|
| 594 |
+
"grad_norm": 0.24890539050102234,
|
| 595 |
+
"learning_rate": 0.0001149620434799457,
|
| 596 |
+
"loss": 0.0396,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.421652421652422,
|
| 601 |
+
"grad_norm": 0.2868204712867737,
|
| 602 |
+
"learning_rate": 0.00011309163800232039,
|
| 603 |
+
"loss": 0.0428,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.45014245014245,
|
| 608 |
+
"grad_norm": 0.30886110663414,
|
| 609 |
+
"learning_rate": 0.00011121656059477874,
|
| 610 |
+
"loss": 0.048,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.4786324786324787,
|
| 615 |
+
"grad_norm": 0.33962446451187134,
|
| 616 |
+
"learning_rate": 0.00010933748040431626,
|
| 617 |
+
"loss": 0.038,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.5071225071225074,
|
| 622 |
+
"grad_norm": 0.414177268743515,
|
| 623 |
+
"learning_rate": 0.00010745506800637624,
|
| 624 |
+
"loss": 0.0376,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.5356125356125356,
|
| 629 |
+
"grad_norm": 0.4189615547657013,
|
| 630 |
+
"learning_rate": 0.00010556999516554583,
|
| 631 |
+
"loss": 0.0406,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.564102564102564,
|
| 636 |
+
"grad_norm": 0.32983464002609253,
|
| 637 |
+
"learning_rate": 0.00010368293459582743,
|
| 638 |
+
"loss": 0.0387,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.5925925925925926,
|
| 643 |
+
"grad_norm": 0.37069281935691833,
|
| 644 |
+
"learning_rate": 0.00010179455972057182,
|
| 645 |
+
"loss": 0.0382,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.6210826210826212,
|
| 650 |
+
"grad_norm": 0.28479892015457153,
|
| 651 |
+
"learning_rate": 9.990554443215748e-05,
|
| 652 |
+
"loss": 0.0443,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.6495726495726495,
|
| 657 |
+
"grad_norm": 0.31271642446517944,
|
| 658 |
+
"learning_rate": 9.80165628515033e-05,
|
| 659 |
+
"loss": 0.0479,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.678062678062678,
|
| 664 |
+
"grad_norm": 0.17170262336730957,
|
| 665 |
+
"learning_rate": 9.612828908749888e-05,
|
| 666 |
+
"loss": 0.035,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.7065527065527064,
|
| 671 |
+
"grad_norm": 0.33237120509147644,
|
| 672 |
+
"learning_rate": 9.424139699643992e-05,
|
| 673 |
+
"loss": 0.033,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.735042735042735,
|
| 678 |
+
"grad_norm": 0.2521449625492096,
|
| 679 |
+
"learning_rate": 9.235655994155326e-05,
|
| 680 |
+
"loss": 0.0399,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.763532763532764,
|
| 685 |
+
"grad_norm": 0.20456038415431976,
|
| 686 |
+
"learning_rate": 9.047445055269784e-05,
|
| 687 |
+
"loss": 0.0299,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.792022792022792,
|
| 692 |
+
"grad_norm": 0.2347419261932373,
|
| 693 |
+
"learning_rate": 8.859574048632786e-05,
|
| 694 |
+
"loss": 0.0344,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.8205128205128203,
|
| 699 |
+
"grad_norm": 0.2831404507160187,
|
| 700 |
+
"learning_rate": 8.672110018580282e-05,
|
| 701 |
+
"loss": 0.0346,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.849002849002849,
|
| 706 |
+
"grad_norm": 0.2277534157037735,
|
| 707 |
+
"learning_rate": 8.485119864213058e-05,
|
| 708 |
+
"loss": 0.0399,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.8774928774928776,
|
| 713 |
+
"grad_norm": 0.27118125557899475,
|
| 714 |
+
"learning_rate": 8.298670315522894e-05,
|
| 715 |
+
"loss": 0.0455,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.905982905982906,
|
| 720 |
+
"grad_norm": 0.21337424218654633,
|
| 721 |
+
"learning_rate": 8.112827909579045e-05,
|
| 722 |
+
"loss": 0.025,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.9344729344729346,
|
| 727 |
+
"grad_norm": 0.28749632835388184,
|
| 728 |
+
"learning_rate": 7.927658966783576e-05,
|
| 729 |
+
"loss": 0.0398,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.962962962962963,
|
| 734 |
+
"grad_norm": 0.26412132382392883,
|
| 735 |
+
"learning_rate": 7.743229567204036e-05,
|
| 736 |
+
"loss": 0.029,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.9914529914529915,
|
| 741 |
+
"grad_norm": 0.2315894365310669,
|
| 742 |
+
"learning_rate": 7.559605526991859e-05,
|
| 743 |
+
"loss": 0.0305,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.0199430199430197,
|
| 748 |
+
"grad_norm": 0.22662468254566193,
|
| 749 |
+
"learning_rate": 7.376852374894989e-05,
|
| 750 |
+
"loss": 0.0405,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.0484330484330484,
|
| 755 |
+
"grad_norm": 0.19257226586341858,
|
| 756 |
+
"learning_rate": 7.195035328873048e-05,
|
| 757 |
+
"loss": 0.0266,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.076923076923077,
|
| 762 |
+
"grad_norm": 0.3498987555503845,
|
| 763 |
+
"learning_rate": 7.014219272823407e-05,
|
| 764 |
+
"loss": 0.0335,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.1054131054131053,
|
| 769 |
+
"grad_norm": 0.20060497522354126,
|
| 770 |
+
"learning_rate": 6.834468733426498e-05,
|
| 771 |
+
"loss": 0.0273,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.133903133903134,
|
| 776 |
+
"grad_norm": 0.23581984639167786,
|
| 777 |
+
"learning_rate": 6.65584785711856e-05,
|
| 778 |
+
"loss": 0.0393,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.1623931623931623,
|
| 783 |
+
"grad_norm": 0.35779276490211487,
|
| 784 |
+
"learning_rate": 6.478420387200135e-05,
|
| 785 |
+
"loss": 0.0281,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.190883190883191,
|
| 790 |
+
"grad_norm": 0.46333932876586914,
|
| 791 |
+
"learning_rate": 6.302249641088384e-05,
|
| 792 |
+
"loss": 0.0319,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.219373219373219,
|
| 797 |
+
"grad_norm": 0.2620326280593872,
|
| 798 |
+
"learning_rate": 6.127398487721398e-05,
|
| 799 |
+
"loss": 0.0386,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.247863247863248,
|
| 804 |
+
"grad_norm": 0.28111764788627625,
|
| 805 |
+
"learning_rate": 5.953929325122578e-05,
|
| 806 |
+
"loss": 0.0248,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.2763532763532766,
|
| 811 |
+
"grad_norm": 0.2124904990196228,
|
| 812 |
+
"learning_rate": 5.781904058133016e-05,
|
| 813 |
+
"loss": 0.0359,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.304843304843305,
|
| 818 |
+
"grad_norm": 0.15361037850379944,
|
| 819 |
+
"learning_rate": 5.611384076319944e-05,
|
| 820 |
+
"loss": 0.0235,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.3333333333333335,
|
| 825 |
+
"grad_norm": 0.2945825159549713,
|
| 826 |
+
"learning_rate": 5.442430232069e-05,
|
| 827 |
+
"loss": 0.0316,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.3618233618233617,
|
| 832 |
+
"grad_norm": 0.2039274275302887,
|
| 833 |
+
"learning_rate": 5.2751028188682315e-05,
|
| 834 |
+
"loss": 0.0301,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.3903133903133904,
|
| 839 |
+
"grad_norm": 0.18716038763523102,
|
| 840 |
+
"learning_rate": 5.109461549791541e-05,
|
| 841 |
+
"loss": 0.0335,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.4188034188034186,
|
| 846 |
+
"grad_norm": 0.25252339243888855,
|
| 847 |
+
"learning_rate": 4.9455655361892663e-05,
|
| 848 |
+
"loss": 0.0295,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.4472934472934473,
|
| 853 |
+
"grad_norm": 0.3167603313922882,
|
| 854 |
+
"learning_rate": 4.783473266593471e-05,
|
| 855 |
+
"loss": 0.0365,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.4757834757834756,
|
| 860 |
+
"grad_norm": 0.3018173575401306,
|
| 861 |
+
"learning_rate": 4.62324258584553e-05,
|
| 862 |
+
"loss": 0.0361,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.5042735042735043,
|
| 867 |
+
"grad_norm": 0.3451387882232666,
|
| 868 |
+
"learning_rate": 4.464930674453393e-05,
|
| 869 |
+
"loss": 0.0321,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.532763532763533,
|
| 874 |
+
"grad_norm": 0.22405466437339783,
|
| 875 |
+
"learning_rate": 4.308594028185944e-05,
|
| 876 |
+
"loss": 0.0238,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.561253561253561,
|
| 881 |
+
"grad_norm": 0.37276986241340637,
|
| 882 |
+
"learning_rate": 4.154288437911732e-05,
|
| 883 |
+
"loss": 0.0341,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.58974358974359,
|
| 888 |
+
"grad_norm": 0.3159347474575043,
|
| 889 |
+
"learning_rate": 4.0020689696891944e-05,
|
| 890 |
+
"loss": 0.0227,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.618233618233618,
|
| 895 |
+
"grad_norm": 0.2659304738044739,
|
| 896 |
+
"learning_rate": 3.8519899451156325e-05,
|
| 897 |
+
"loss": 0.0239,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.646723646723647,
|
| 902 |
+
"grad_norm": 0.3740275502204895,
|
| 903 |
+
"learning_rate": 3.704104921941768e-05,
|
| 904 |
+
"loss": 0.0309,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 3.6752136752136755,
|
| 909 |
+
"grad_norm": 0.29278329014778137,
|
| 910 |
+
"learning_rate": 3.558466674958947e-05,
|
| 911 |
+
"loss": 0.0313,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 3.7037037037037037,
|
| 916 |
+
"grad_norm": 0.30194053053855896,
|
| 917 |
+
"learning_rate": 3.415127177165729e-05,
|
| 918 |
+
"loss": 0.0396,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 3.732193732193732,
|
| 923 |
+
"grad_norm": 0.1718418151140213,
|
| 924 |
+
"learning_rate": 3.274137581220614e-05,
|
| 925 |
+
"loss": 0.0225,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 3.7606837606837606,
|
| 930 |
+
"grad_norm": 0.22094787657260895,
|
| 931 |
+
"learning_rate": 3.135548201187514e-05,
|
| 932 |
+
"loss": 0.0225,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 3.7891737891737893,
|
| 937 |
+
"grad_norm": 0.19169935584068298,
|
| 938 |
+
"learning_rate": 2.9994084945805047e-05,
|
| 939 |
+
"loss": 0.0341,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 3.8176638176638176,
|
| 944 |
+
"grad_norm": 0.33710628747940063,
|
| 945 |
+
"learning_rate": 2.865767044714205e-05,
|
| 946 |
+
"loss": 0.0378,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.8461538461538463,
|
| 951 |
+
"grad_norm": 0.22694231569766998,
|
| 952 |
+
"learning_rate": 2.7346715433661985e-05,
|
| 953 |
+
"loss": 0.0299,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.8746438746438745,
|
| 958 |
+
"grad_norm": 0.4005008339881897,
|
| 959 |
+
"learning_rate": 2.6061687737575378e-05,
|
| 960 |
+
"loss": 0.029,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 3.903133903133903,
|
| 965 |
+
"grad_norm": 0.25156232714653015,
|
| 966 |
+
"learning_rate": 2.4803045938575287e-05,
|
| 967 |
+
"loss": 0.0242,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 3.931623931623932,
|
| 972 |
+
"grad_norm": 0.30339252948760986,
|
| 973 |
+
"learning_rate": 2.3571239200186844e-05,
|
| 974 |
+
"loss": 0.03,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 3.96011396011396,
|
| 979 |
+
"grad_norm": 0.2436976432800293,
|
| 980 |
+
"learning_rate": 2.236670710947707e-05,
|
| 981 |
+
"loss": 0.0292,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 3.9886039886039883,
|
| 986 |
+
"grad_norm": 0.2527185082435608,
|
| 987 |
+
"learning_rate": 2.1189879520182267e-05,
|
| 988 |
+
"loss": 0.0269,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.017094017094017,
|
| 993 |
+
"grad_norm": 0.32539618015289307,
|
| 994 |
+
"learning_rate": 2.0041176399308924e-05,
|
| 995 |
+
"loss": 0.0463,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.045584045584046,
|
| 1000 |
+
"grad_norm": 0.28222835063934326,
|
| 1001 |
+
"learning_rate": 1.892100767726247e-05,
|
| 1002 |
+
"loss": 0.0275,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.074074074074074,
|
| 1007 |
+
"grad_norm": 0.30946633219718933,
|
| 1008 |
+
"learning_rate": 1.7829773101558365e-05,
|
| 1009 |
+
"loss": 0.0261,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.102564102564102,
|
| 1014 |
+
"grad_norm": 0.1978222280740738,
|
| 1015 |
+
"learning_rate": 1.6767862094166498e-05,
|
| 1016 |
+
"loss": 0.0226,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.131054131054131,
|
| 1021 |
+
"grad_norm": 0.14639084041118622,
|
| 1022 |
+
"learning_rate": 1.573565361254069e-05,
|
| 1023 |
+
"loss": 0.0238,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.15954415954416,
|
| 1028 |
+
"grad_norm": 0.25420257449150085,
|
| 1029 |
+
"learning_rate": 1.4733516014382586e-05,
|
| 1030 |
+
"loss": 0.0298,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.188034188034188,
|
| 1035 |
+
"grad_norm": 0.281435489654541,
|
| 1036 |
+
"learning_rate": 1.3761806926188148e-05,
|
| 1037 |
+
"loss": 0.0197,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.216524216524217,
|
| 1042 |
+
"grad_norm": 0.26053982973098755,
|
| 1043 |
+
"learning_rate": 1.2820873115623722e-05,
|
| 1044 |
+
"loss": 0.0229,
|
| 1045 |
+
"step": 1480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.245014245014245,
|
| 1049 |
+
"grad_norm": 0.21949049830436707,
|
| 1050 |
+
"learning_rate": 1.1911050367777388e-05,
|
| 1051 |
+
"loss": 0.0205,
|
| 1052 |
+
"step": 1490
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.273504273504273,
|
| 1056 |
+
"grad_norm": 0.3151997923851013,
|
| 1057 |
+
"learning_rate": 1.1032663365329255e-05,
|
| 1058 |
+
"loss": 0.0325,
|
| 1059 |
+
"step": 1500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.301994301994302,
|
| 1063 |
+
"grad_norm": 0.2235259711742401,
|
| 1064 |
+
"learning_rate": 1.0186025572684189e-05,
|
| 1065 |
+
"loss": 0.0208,
|
| 1066 |
+
"step": 1510
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.330484330484331,
|
| 1070 |
+
"grad_norm": 0.21038542687892914,
|
| 1071 |
+
"learning_rate": 9.37143912410765e-06,
|
| 1072 |
+
"loss": 0.0216,
|
| 1073 |
+
"step": 1520
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.358974358974359,
|
| 1077 |
+
"grad_norm": 0.21731996536254883,
|
| 1078 |
+
"learning_rate": 8.589194715905036e-06,
|
| 1079 |
+
"loss": 0.0283,
|
| 1080 |
+
"step": 1530
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.387464387464387,
|
| 1084 |
+
"grad_norm": 0.20443207025527954,
|
| 1085 |
+
"learning_rate": 7.839571502682652e-06,
|
| 1086 |
+
"loss": 0.0194,
|
| 1087 |
+
"step": 1540
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.415954415954416,
|
| 1091 |
+
"grad_norm": 0.32589828968048096,
|
| 1092 |
+
"learning_rate": 7.122836997727611e-06,
|
| 1093 |
+
"loss": 0.0313,
|
| 1094 |
+
"step": 1550
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.444444444444445,
|
| 1098 |
+
"grad_norm": 0.19662369787693024,
|
| 1099 |
+
"learning_rate": 6.439246977542091e-06,
|
| 1100 |
+
"loss": 0.0218,
|
| 1101 |
+
"step": 1560
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.472934472934473,
|
| 1105 |
+
"grad_norm": 0.24466551840305328,
|
| 1106 |
+
"learning_rate": 5.789045390565951e-06,
|
| 1107 |
+
"loss": 0.0196,
|
| 1108 |
+
"step": 1570
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.501424501424501,
|
| 1112 |
+
"grad_norm": 0.20958594977855682,
|
| 1113 |
+
"learning_rate": 5.1724642701203916e-06,
|
| 1114 |
+
"loss": 0.0215,
|
| 1115 |
+
"step": 1580
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.52991452991453,
|
| 1119 |
+
"grad_norm": 0.28344252705574036,
|
| 1120 |
+
"learning_rate": 4.589723651603739e-06,
|
| 1121 |
+
"loss": 0.0233,
|
| 1122 |
+
"step": 1590
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.5584045584045585,
|
| 1126 |
+
"grad_norm": 0.15124832093715668,
|
| 1127 |
+
"learning_rate": 4.0410314939687725e-06,
|
| 1128 |
+
"loss": 0.0225,
|
| 1129 |
+
"step": 1600
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 4.586894586894587,
|
| 1133 |
+
"grad_norm": 0.2302241027355194,
|
| 1134 |
+
"learning_rate": 3.5265836055097258e-06,
|
| 1135 |
+
"loss": 0.0239,
|
| 1136 |
+
"step": 1610
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 4.615384615384615,
|
| 1140 |
+
"grad_norm": 0.3110940456390381,
|
| 1141 |
+
"learning_rate": 3.0465635739854725e-06,
|
| 1142 |
+
"loss": 0.0357,
|
| 1143 |
+
"step": 1620
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 4.643874643874644,
|
| 1147 |
+
"grad_norm": 0.19868485629558563,
|
| 1148 |
+
"learning_rate": 2.6011427011036317e-06,
|
| 1149 |
+
"loss": 0.0296,
|
| 1150 |
+
"step": 1630
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 4.672364672364672,
|
| 1154 |
+
"grad_norm": 0.180753692984581,
|
| 1155 |
+
"learning_rate": 2.1904799413893116e-06,
|
| 1156 |
+
"loss": 0.0271,
|
| 1157 |
+
"step": 1640
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 4.700854700854701,
|
| 1161 |
+
"grad_norm": 0.24652840197086334,
|
| 1162 |
+
"learning_rate": 1.814721845459977e-06,
|
| 1163 |
+
"loss": 0.0192,
|
| 1164 |
+
"step": 1650
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 4.72934472934473,
|
| 1168 |
+
"grad_norm": 0.27937212586402893,
|
| 1169 |
+
"learning_rate": 1.4740025077268128e-06,
|
| 1170 |
+
"loss": 0.0265,
|
| 1171 |
+
"step": 1660
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 4.7578347578347575,
|
| 1175 |
+
"grad_norm": 0.2583824694156647,
|
| 1176 |
+
"learning_rate": 1.1684435185413733e-06,
|
| 1177 |
+
"loss": 0.0274,
|
| 1178 |
+
"step": 1670
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 4.786324786324786,
|
| 1182 |
+
"grad_norm": 0.2571396231651306,
|
| 1183 |
+
"learning_rate": 8.981539208043788e-07,
|
| 1184 |
+
"loss": 0.0248,
|
| 1185 |
+
"step": 1680
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 4.814814814814815,
|
| 1189 |
+
"grad_norm": 0.26664578914642334,
|
| 1190 |
+
"learning_rate": 6.632301710522205e-07,
|
| 1191 |
+
"loss": 0.022,
|
| 1192 |
+
"step": 1690
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 4.843304843304844,
|
| 1196 |
+
"grad_norm": 0.22444820404052734,
|
| 1197 |
+
"learning_rate": 4.63756105035229e-07,
|
| 1198 |
+
"loss": 0.0357,
|
| 1199 |
+
"step": 1700
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 4.871794871794872,
|
| 1203 |
+
"grad_norm": 0.21229873597621918,
|
| 1204 |
+
"learning_rate": 2.9980290779960676e-07,
|
| 1205 |
+
"loss": 0.0236,
|
| 1206 |
+
"step": 1710
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 4.9002849002849,
|
| 1210 |
+
"grad_norm": 0.2009340524673462,
|
| 1211 |
+
"learning_rate": 1.71429088284214e-07,
|
| 1212 |
+
"loss": 0.0201,
|
| 1213 |
+
"step": 1720
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 4.928774928774929,
|
| 1217 |
+
"grad_norm": 0.2139723300933838,
|
| 1218 |
+
"learning_rate": 7.868045844073857e-08,
|
| 1219 |
+
"loss": 0.0255,
|
| 1220 |
+
"step": 1730
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 4.957264957264957,
|
| 1224 |
+
"grad_norm": 0.10056009143590927,
|
| 1225 |
+
"learning_rate": 2.1590116885139567e-08,
|
| 1226 |
+
"loss": 0.0238,
|
| 1227 |
+
"step": 1740
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 4.985754985754986,
|
| 1231 |
+
"grad_norm": 0.21970975399017334,
|
| 1232 |
+
"learning_rate": 1.7843708592923947e-10,
|
| 1233 |
+
"loss": 0.0277,
|
| 1234 |
+
"step": 1750
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 4.988603988603988,
|
| 1238 |
+
"step": 1751,
|
| 1239 |
+
"total_flos": 3.081693778301952e+16,
|
| 1240 |
+
"train_loss": 0.060192157869157555,
|
| 1241 |
+
"train_runtime": 593.7414,
|
| 1242 |
+
"train_samples_per_second": 47.186,
|
| 1243 |
+
"train_steps_per_second": 2.949
|
| 1244 |
+
}
|
| 1245 |
+
],
|
| 1246 |
+
"logging_steps": 10,
|
| 1247 |
+
"max_steps": 1751,
|
| 1248 |
+
"num_input_tokens_seen": 0,
|
| 1249 |
+
"num_train_epochs": 5,
|
| 1250 |
+
"save_steps": 500,
|
| 1251 |
+
"stateful_callbacks": {
|
| 1252 |
+
"TrainerControl": {
|
| 1253 |
+
"args": {
|
| 1254 |
+
"should_epoch_stop": false,
|
| 1255 |
+
"should_evaluate": false,
|
| 1256 |
+
"should_log": false,
|
| 1257 |
+
"should_save": true,
|
| 1258 |
+
"should_training_stop": true
|
| 1259 |
+
},
|
| 1260 |
+
"attributes": {}
|
| 1261 |
+
}
|
| 1262 |
+
},
|
| 1263 |
+
"total_flos": 3.081693778301952e+16,
|
| 1264 |
+
"train_batch_size": 16,
|
| 1265 |
+
"trial_name": null,
|
| 1266 |
+
"trial_params": null
|
| 1267 |
+
}
|