Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1215 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1215 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 5000,
|
| 6 |
+
"global_step": 9822,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"learning_rate": 1.9898187741804116e-05,
|
| 14 |
+
"loss": 0.7281,
|
| 15 |
+
"step": 50
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.03,
|
| 19 |
+
"learning_rate": 1.9796375483608227e-05,
|
| 20 |
+
"loss": 0.6362,
|
| 21 |
+
"step": 100
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.05,
|
| 25 |
+
"learning_rate": 1.9694563225412342e-05,
|
| 26 |
+
"loss": 0.5684,
|
| 27 |
+
"step": 150
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.06,
|
| 31 |
+
"learning_rate": 1.9592750967216453e-05,
|
| 32 |
+
"loss": 0.5082,
|
| 33 |
+
"step": 200
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.08,
|
| 37 |
+
"learning_rate": 1.9490938709020567e-05,
|
| 38 |
+
"loss": 0.4732,
|
| 39 |
+
"step": 250
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.09,
|
| 43 |
+
"learning_rate": 1.938912645082468e-05,
|
| 44 |
+
"loss": 0.4337,
|
| 45 |
+
"step": 300
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11,
|
| 49 |
+
"learning_rate": 1.9287314192628793e-05,
|
| 50 |
+
"loss": 0.437,
|
| 51 |
+
"step": 350
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"learning_rate": 1.9185501934432908e-05,
|
| 56 |
+
"loss": 0.4676,
|
| 57 |
+
"step": 400
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.14,
|
| 61 |
+
"learning_rate": 1.9083689676237022e-05,
|
| 62 |
+
"loss": 0.3977,
|
| 63 |
+
"step": 450
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.15,
|
| 67 |
+
"learning_rate": 1.8981877418041133e-05,
|
| 68 |
+
"loss": 0.4522,
|
| 69 |
+
"step": 500
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.17,
|
| 73 |
+
"learning_rate": 1.8880065159845248e-05,
|
| 74 |
+
"loss": 0.431,
|
| 75 |
+
"step": 550
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.18,
|
| 79 |
+
"learning_rate": 1.8778252901649362e-05,
|
| 80 |
+
"loss": 0.4559,
|
| 81 |
+
"step": 600
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.2,
|
| 85 |
+
"learning_rate": 1.8676440643453473e-05,
|
| 86 |
+
"loss": 0.4181,
|
| 87 |
+
"step": 650
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.21,
|
| 91 |
+
"learning_rate": 1.8574628385257588e-05,
|
| 92 |
+
"loss": 0.3736,
|
| 93 |
+
"step": 700
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"learning_rate": 1.84728161270617e-05,
|
| 98 |
+
"loss": 0.4084,
|
| 99 |
+
"step": 750
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.24,
|
| 103 |
+
"learning_rate": 1.8371003868865813e-05,
|
| 104 |
+
"loss": 0.3837,
|
| 105 |
+
"step": 800
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.26,
|
| 109 |
+
"learning_rate": 1.8269191610669924e-05,
|
| 110 |
+
"loss": 0.3677,
|
| 111 |
+
"step": 850
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.27,
|
| 115 |
+
"learning_rate": 1.816737935247404e-05,
|
| 116 |
+
"loss": 0.3895,
|
| 117 |
+
"step": 900
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.29,
|
| 121 |
+
"learning_rate": 1.8065567094278153e-05,
|
| 122 |
+
"loss": 0.3257,
|
| 123 |
+
"step": 950
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.31,
|
| 127 |
+
"learning_rate": 1.7963754836082265e-05,
|
| 128 |
+
"loss": 0.4083,
|
| 129 |
+
"step": 1000
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.32,
|
| 133 |
+
"learning_rate": 1.786194257788638e-05,
|
| 134 |
+
"loss": 0.3841,
|
| 135 |
+
"step": 1050
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34,
|
| 139 |
+
"learning_rate": 1.776013031969049e-05,
|
| 140 |
+
"loss": 0.3435,
|
| 141 |
+
"step": 1100
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.35,
|
| 145 |
+
"learning_rate": 1.7658318061494605e-05,
|
| 146 |
+
"loss": 0.3665,
|
| 147 |
+
"step": 1150
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.37,
|
| 151 |
+
"learning_rate": 1.755650580329872e-05,
|
| 152 |
+
"loss": 0.3304,
|
| 153 |
+
"step": 1200
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.38,
|
| 157 |
+
"learning_rate": 1.7454693545102834e-05,
|
| 158 |
+
"loss": 0.3626,
|
| 159 |
+
"step": 1250
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.4,
|
| 163 |
+
"learning_rate": 1.7352881286906945e-05,
|
| 164 |
+
"loss": 0.3624,
|
| 165 |
+
"step": 1300
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.41,
|
| 169 |
+
"learning_rate": 1.725106902871106e-05,
|
| 170 |
+
"loss": 0.3736,
|
| 171 |
+
"step": 1350
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.43,
|
| 175 |
+
"learning_rate": 1.7149256770515174e-05,
|
| 176 |
+
"loss": 0.3487,
|
| 177 |
+
"step": 1400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.44,
|
| 181 |
+
"learning_rate": 1.7047444512319285e-05,
|
| 182 |
+
"loss": 0.3948,
|
| 183 |
+
"step": 1450
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.46,
|
| 187 |
+
"learning_rate": 1.69456322541234e-05,
|
| 188 |
+
"loss": 0.3533,
|
| 189 |
+
"step": 1500
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.47,
|
| 193 |
+
"learning_rate": 1.684381999592751e-05,
|
| 194 |
+
"loss": 0.3633,
|
| 195 |
+
"step": 1550
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.49,
|
| 199 |
+
"learning_rate": 1.6742007737731625e-05,
|
| 200 |
+
"loss": 0.3657,
|
| 201 |
+
"step": 1600
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.5,
|
| 205 |
+
"learning_rate": 1.6640195479535736e-05,
|
| 206 |
+
"loss": 0.3514,
|
| 207 |
+
"step": 1650
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.52,
|
| 211 |
+
"learning_rate": 1.653838322133985e-05,
|
| 212 |
+
"loss": 0.3521,
|
| 213 |
+
"step": 1700
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.53,
|
| 217 |
+
"learning_rate": 1.6436570963143962e-05,
|
| 218 |
+
"loss": 0.3461,
|
| 219 |
+
"step": 1750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.55,
|
| 223 |
+
"learning_rate": 1.6334758704948076e-05,
|
| 224 |
+
"loss": 0.3433,
|
| 225 |
+
"step": 1800
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.57,
|
| 229 |
+
"learning_rate": 1.623294644675219e-05,
|
| 230 |
+
"loss": 0.3531,
|
| 231 |
+
"step": 1850
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.58,
|
| 235 |
+
"learning_rate": 1.6131134188556302e-05,
|
| 236 |
+
"loss": 0.3323,
|
| 237 |
+
"step": 1900
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.6,
|
| 241 |
+
"learning_rate": 1.6029321930360416e-05,
|
| 242 |
+
"loss": 0.3638,
|
| 243 |
+
"step": 1950
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.61,
|
| 247 |
+
"learning_rate": 1.592750967216453e-05,
|
| 248 |
+
"loss": 0.3214,
|
| 249 |
+
"step": 2000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.63,
|
| 253 |
+
"learning_rate": 1.5825697413968642e-05,
|
| 254 |
+
"loss": 0.3596,
|
| 255 |
+
"step": 2050
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.64,
|
| 259 |
+
"learning_rate": 1.5723885155772757e-05,
|
| 260 |
+
"loss": 0.3631,
|
| 261 |
+
"step": 2100
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.66,
|
| 265 |
+
"learning_rate": 1.562207289757687e-05,
|
| 266 |
+
"loss": 0.3482,
|
| 267 |
+
"step": 2150
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.67,
|
| 271 |
+
"learning_rate": 1.5520260639380982e-05,
|
| 272 |
+
"loss": 0.3049,
|
| 273 |
+
"step": 2200
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.69,
|
| 277 |
+
"learning_rate": 1.5418448381185097e-05,
|
| 278 |
+
"loss": 0.326,
|
| 279 |
+
"step": 2250
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.7,
|
| 283 |
+
"learning_rate": 1.531663612298921e-05,
|
| 284 |
+
"loss": 0.3219,
|
| 285 |
+
"step": 2300
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.72,
|
| 289 |
+
"learning_rate": 1.5214823864793322e-05,
|
| 290 |
+
"loss": 0.3637,
|
| 291 |
+
"step": 2350
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.73,
|
| 295 |
+
"learning_rate": 1.5113011606597437e-05,
|
| 296 |
+
"loss": 0.3123,
|
| 297 |
+
"step": 2400
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.75,
|
| 301 |
+
"learning_rate": 1.5011199348401548e-05,
|
| 302 |
+
"loss": 0.3368,
|
| 303 |
+
"step": 2450
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.76,
|
| 307 |
+
"learning_rate": 1.4909387090205662e-05,
|
| 308 |
+
"loss": 0.3322,
|
| 309 |
+
"step": 2500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.78,
|
| 313 |
+
"learning_rate": 1.4807574832009775e-05,
|
| 314 |
+
"loss": 0.3237,
|
| 315 |
+
"step": 2550
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.79,
|
| 319 |
+
"learning_rate": 1.4705762573813888e-05,
|
| 320 |
+
"loss": 0.3336,
|
| 321 |
+
"step": 2600
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.81,
|
| 325 |
+
"learning_rate": 1.4603950315618001e-05,
|
| 326 |
+
"loss": 0.3075,
|
| 327 |
+
"step": 2650
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.82,
|
| 331 |
+
"learning_rate": 1.4502138057422115e-05,
|
| 332 |
+
"loss": 0.3259,
|
| 333 |
+
"step": 2700
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.84,
|
| 337 |
+
"learning_rate": 1.440032579922623e-05,
|
| 338 |
+
"loss": 0.3239,
|
| 339 |
+
"step": 2750
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.86,
|
| 343 |
+
"learning_rate": 1.4298513541030341e-05,
|
| 344 |
+
"loss": 0.3087,
|
| 345 |
+
"step": 2800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.87,
|
| 349 |
+
"learning_rate": 1.4196701282834456e-05,
|
| 350 |
+
"loss": 0.3192,
|
| 351 |
+
"step": 2850
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.89,
|
| 355 |
+
"learning_rate": 1.4094889024638567e-05,
|
| 356 |
+
"loss": 0.2907,
|
| 357 |
+
"step": 2900
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.9,
|
| 361 |
+
"learning_rate": 1.3993076766442681e-05,
|
| 362 |
+
"loss": 0.3213,
|
| 363 |
+
"step": 2950
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.92,
|
| 367 |
+
"learning_rate": 1.3891264508246794e-05,
|
| 368 |
+
"loss": 0.2869,
|
| 369 |
+
"step": 3000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.93,
|
| 373 |
+
"learning_rate": 1.3789452250050907e-05,
|
| 374 |
+
"loss": 0.3254,
|
| 375 |
+
"step": 3050
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.95,
|
| 379 |
+
"learning_rate": 1.368763999185502e-05,
|
| 380 |
+
"loss": 0.3051,
|
| 381 |
+
"step": 3100
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.96,
|
| 385 |
+
"learning_rate": 1.3585827733659134e-05,
|
| 386 |
+
"loss": 0.294,
|
| 387 |
+
"step": 3150
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.98,
|
| 391 |
+
"learning_rate": 1.3484015475463249e-05,
|
| 392 |
+
"loss": 0.3388,
|
| 393 |
+
"step": 3200
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.99,
|
| 397 |
+
"learning_rate": 1.338220321726736e-05,
|
| 398 |
+
"loss": 0.2742,
|
| 399 |
+
"step": 3250
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.01,
|
| 403 |
+
"learning_rate": 1.3280390959071474e-05,
|
| 404 |
+
"loss": 0.2752,
|
| 405 |
+
"step": 3300
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.02,
|
| 409 |
+
"learning_rate": 1.3178578700875585e-05,
|
| 410 |
+
"loss": 0.249,
|
| 411 |
+
"step": 3350
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.04,
|
| 415 |
+
"learning_rate": 1.30767664426797e-05,
|
| 416 |
+
"loss": 0.2498,
|
| 417 |
+
"step": 3400
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.05,
|
| 421 |
+
"learning_rate": 1.2974954184483813e-05,
|
| 422 |
+
"loss": 0.2535,
|
| 423 |
+
"step": 3450
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.07,
|
| 427 |
+
"learning_rate": 1.2873141926287927e-05,
|
| 428 |
+
"loss": 0.2131,
|
| 429 |
+
"step": 3500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.08,
|
| 433 |
+
"learning_rate": 1.2771329668092038e-05,
|
| 434 |
+
"loss": 0.2743,
|
| 435 |
+
"step": 3550
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.1,
|
| 439 |
+
"learning_rate": 1.2669517409896153e-05,
|
| 440 |
+
"loss": 0.2753,
|
| 441 |
+
"step": 3600
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.11,
|
| 445 |
+
"learning_rate": 1.2567705151700267e-05,
|
| 446 |
+
"loss": 0.2511,
|
| 447 |
+
"step": 3650
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 1.13,
|
| 451 |
+
"learning_rate": 1.2465892893504378e-05,
|
| 452 |
+
"loss": 0.2377,
|
| 453 |
+
"step": 3700
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 1.15,
|
| 457 |
+
"learning_rate": 1.2364080635308493e-05,
|
| 458 |
+
"loss": 0.2531,
|
| 459 |
+
"step": 3750
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.16,
|
| 463 |
+
"learning_rate": 1.2262268377112606e-05,
|
| 464 |
+
"loss": 0.2313,
|
| 465 |
+
"step": 3800
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.18,
|
| 469 |
+
"learning_rate": 1.2160456118916718e-05,
|
| 470 |
+
"loss": 0.2539,
|
| 471 |
+
"step": 3850
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.19,
|
| 475 |
+
"learning_rate": 1.2058643860720831e-05,
|
| 476 |
+
"loss": 0.2352,
|
| 477 |
+
"step": 3900
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 1.21,
|
| 481 |
+
"learning_rate": 1.1956831602524946e-05,
|
| 482 |
+
"loss": 0.2543,
|
| 483 |
+
"step": 3950
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 1.22,
|
| 487 |
+
"learning_rate": 1.1855019344329057e-05,
|
| 488 |
+
"loss": 0.243,
|
| 489 |
+
"step": 4000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.24,
|
| 493 |
+
"learning_rate": 1.1753207086133171e-05,
|
| 494 |
+
"loss": 0.2253,
|
| 495 |
+
"step": 4050
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 1.25,
|
| 499 |
+
"learning_rate": 1.1651394827937284e-05,
|
| 500 |
+
"loss": 0.2428,
|
| 501 |
+
"step": 4100
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.27,
|
| 505 |
+
"learning_rate": 1.1549582569741397e-05,
|
| 506 |
+
"loss": 0.2551,
|
| 507 |
+
"step": 4150
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.28,
|
| 511 |
+
"learning_rate": 1.1447770311545512e-05,
|
| 512 |
+
"loss": 0.2582,
|
| 513 |
+
"step": 4200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3,
|
| 517 |
+
"learning_rate": 1.1345958053349624e-05,
|
| 518 |
+
"loss": 0.248,
|
| 519 |
+
"step": 4250
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.31,
|
| 523 |
+
"learning_rate": 1.1244145795153739e-05,
|
| 524 |
+
"loss": 0.2346,
|
| 525 |
+
"step": 4300
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.33,
|
| 529 |
+
"learning_rate": 1.114233353695785e-05,
|
| 530 |
+
"loss": 0.2472,
|
| 531 |
+
"step": 4350
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 1.34,
|
| 535 |
+
"learning_rate": 1.1040521278761964e-05,
|
| 536 |
+
"loss": 0.2212,
|
| 537 |
+
"step": 4400
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.36,
|
| 541 |
+
"learning_rate": 1.0938709020566076e-05,
|
| 542 |
+
"loss": 0.243,
|
| 543 |
+
"step": 4450
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.37,
|
| 547 |
+
"learning_rate": 1.083689676237019e-05,
|
| 548 |
+
"loss": 0.255,
|
| 549 |
+
"step": 4500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.39,
|
| 553 |
+
"learning_rate": 1.0735084504174303e-05,
|
| 554 |
+
"loss": 0.2608,
|
| 555 |
+
"step": 4550
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.41,
|
| 559 |
+
"learning_rate": 1.0633272245978416e-05,
|
| 560 |
+
"loss": 0.2457,
|
| 561 |
+
"step": 4600
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 1.42,
|
| 565 |
+
"learning_rate": 1.053145998778253e-05,
|
| 566 |
+
"loss": 0.2499,
|
| 567 |
+
"step": 4650
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 1.44,
|
| 571 |
+
"learning_rate": 1.0429647729586643e-05,
|
| 572 |
+
"loss": 0.2274,
|
| 573 |
+
"step": 4700
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 1.45,
|
| 577 |
+
"learning_rate": 1.0327835471390758e-05,
|
| 578 |
+
"loss": 0.2018,
|
| 579 |
+
"step": 4750
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 1.47,
|
| 583 |
+
"learning_rate": 1.0226023213194869e-05,
|
| 584 |
+
"loss": 0.2122,
|
| 585 |
+
"step": 4800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.48,
|
| 589 |
+
"learning_rate": 1.0124210954998983e-05,
|
| 590 |
+
"loss": 0.2288,
|
| 591 |
+
"step": 4850
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5,
|
| 595 |
+
"learning_rate": 1.0022398696803094e-05,
|
| 596 |
+
"loss": 0.2497,
|
| 597 |
+
"step": 4900
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.51,
|
| 601 |
+
"learning_rate": 9.920586438607209e-06,
|
| 602 |
+
"loss": 0.255,
|
| 603 |
+
"step": 4950
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 1.53,
|
| 607 |
+
"learning_rate": 9.818774180411322e-06,
|
| 608 |
+
"loss": 0.2396,
|
| 609 |
+
"step": 5000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 1.53,
|
| 613 |
+
"eval_accuracy": 0.9035328574043566,
|
| 614 |
+
"eval_loss": 0.2439582496881485,
|
| 615 |
+
"eval_runtime": 34.2486,
|
| 616 |
+
"eval_samples_per_second": 159.51,
|
| 617 |
+
"eval_steps_per_second": 9.986,
|
| 618 |
+
"step": 5000
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.54,
|
| 622 |
+
"learning_rate": 9.716961922215436e-06,
|
| 623 |
+
"loss": 0.2742,
|
| 624 |
+
"step": 5050
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 1.56,
|
| 628 |
+
"learning_rate": 9.615149664019549e-06,
|
| 629 |
+
"loss": 0.2454,
|
| 630 |
+
"step": 5100
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 1.57,
|
| 634 |
+
"learning_rate": 9.513337405823662e-06,
|
| 635 |
+
"loss": 0.2243,
|
| 636 |
+
"step": 5150
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.59,
|
| 640 |
+
"learning_rate": 9.411525147627775e-06,
|
| 641 |
+
"loss": 0.2542,
|
| 642 |
+
"step": 5200
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 1.6,
|
| 646 |
+
"learning_rate": 9.309712889431889e-06,
|
| 647 |
+
"loss": 0.2566,
|
| 648 |
+
"step": 5250
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.62,
|
| 652 |
+
"learning_rate": 9.207900631236002e-06,
|
| 653 |
+
"loss": 0.2302,
|
| 654 |
+
"step": 5300
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.63,
|
| 658 |
+
"learning_rate": 9.106088373040115e-06,
|
| 659 |
+
"loss": 0.2159,
|
| 660 |
+
"step": 5350
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.65,
|
| 664 |
+
"learning_rate": 9.004276114844227e-06,
|
| 665 |
+
"loss": 0.2364,
|
| 666 |
+
"step": 5400
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 1.66,
|
| 670 |
+
"learning_rate": 8.902463856648342e-06,
|
| 671 |
+
"loss": 0.2192,
|
| 672 |
+
"step": 5450
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 1.68,
|
| 676 |
+
"learning_rate": 8.800651598452455e-06,
|
| 677 |
+
"loss": 0.2352,
|
| 678 |
+
"step": 5500
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.7,
|
| 682 |
+
"learning_rate": 8.698839340256568e-06,
|
| 683 |
+
"loss": 0.2424,
|
| 684 |
+
"step": 5550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 1.71,
|
| 688 |
+
"learning_rate": 8.59702708206068e-06,
|
| 689 |
+
"loss": 0.2394,
|
| 690 |
+
"step": 5600
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.73,
|
| 694 |
+
"learning_rate": 8.495214823864793e-06,
|
| 695 |
+
"loss": 0.261,
|
| 696 |
+
"step": 5650
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.74,
|
| 700 |
+
"learning_rate": 8.393402565668908e-06,
|
| 701 |
+
"loss": 0.21,
|
| 702 |
+
"step": 5700
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.76,
|
| 706 |
+
"learning_rate": 8.29159030747302e-06,
|
| 707 |
+
"loss": 0.2596,
|
| 708 |
+
"step": 5750
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.77,
|
| 712 |
+
"learning_rate": 8.189778049277133e-06,
|
| 713 |
+
"loss": 0.2356,
|
| 714 |
+
"step": 5800
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 1.79,
|
| 718 |
+
"learning_rate": 8.087965791081248e-06,
|
| 719 |
+
"loss": 0.249,
|
| 720 |
+
"step": 5850
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.8,
|
| 724 |
+
"learning_rate": 7.98615353288536e-06,
|
| 725 |
+
"loss": 0.2272,
|
| 726 |
+
"step": 5900
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.82,
|
| 730 |
+
"learning_rate": 7.884341274689473e-06,
|
| 731 |
+
"loss": 0.2296,
|
| 732 |
+
"step": 5950
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.83,
|
| 736 |
+
"learning_rate": 7.782529016493586e-06,
|
| 737 |
+
"loss": 0.2377,
|
| 738 |
+
"step": 6000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.85,
|
| 742 |
+
"learning_rate": 7.680716758297699e-06,
|
| 743 |
+
"loss": 0.2483,
|
| 744 |
+
"step": 6050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.86,
|
| 748 |
+
"learning_rate": 7.578904500101813e-06,
|
| 749 |
+
"loss": 0.2158,
|
| 750 |
+
"step": 6100
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 1.88,
|
| 754 |
+
"learning_rate": 7.4770922419059255e-06,
|
| 755 |
+
"loss": 0.2314,
|
| 756 |
+
"step": 6150
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 1.89,
|
| 760 |
+
"learning_rate": 7.37527998371004e-06,
|
| 761 |
+
"loss": 0.2894,
|
| 762 |
+
"step": 6200
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1.91,
|
| 766 |
+
"learning_rate": 7.273467725514153e-06,
|
| 767 |
+
"loss": 0.24,
|
| 768 |
+
"step": 6250
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.92,
|
| 772 |
+
"learning_rate": 7.171655467318266e-06,
|
| 773 |
+
"loss": 0.2256,
|
| 774 |
+
"step": 6300
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.94,
|
| 778 |
+
"learning_rate": 7.069843209122379e-06,
|
| 779 |
+
"loss": 0.2382,
|
| 780 |
+
"step": 6350
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.95,
|
| 784 |
+
"learning_rate": 6.968030950926492e-06,
|
| 785 |
+
"loss": 0.2296,
|
| 786 |
+
"step": 6400
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.97,
|
| 790 |
+
"learning_rate": 6.866218692730605e-06,
|
| 791 |
+
"loss": 0.2134,
|
| 792 |
+
"step": 6450
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 1.99,
|
| 796 |
+
"learning_rate": 6.764406434534719e-06,
|
| 797 |
+
"loss": 0.2443,
|
| 798 |
+
"step": 6500
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 2.0,
|
| 802 |
+
"learning_rate": 6.662594176338831e-06,
|
| 803 |
+
"loss": 0.2272,
|
| 804 |
+
"step": 6550
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.02,
|
| 808 |
+
"learning_rate": 6.560781918142944e-06,
|
| 809 |
+
"loss": 0.1363,
|
| 810 |
+
"step": 6600
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.03,
|
| 814 |
+
"learning_rate": 6.458969659947059e-06,
|
| 815 |
+
"loss": 0.1584,
|
| 816 |
+
"step": 6650
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 2.05,
|
| 820 |
+
"learning_rate": 6.3571574017511715e-06,
|
| 821 |
+
"loss": 0.1623,
|
| 822 |
+
"step": 6700
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.06,
|
| 826 |
+
"learning_rate": 6.255345143555285e-06,
|
| 827 |
+
"loss": 0.1813,
|
| 828 |
+
"step": 6750
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.08,
|
| 832 |
+
"learning_rate": 6.153532885359398e-06,
|
| 833 |
+
"loss": 0.1715,
|
| 834 |
+
"step": 6800
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 2.09,
|
| 838 |
+
"learning_rate": 6.051720627163511e-06,
|
| 839 |
+
"loss": 0.1529,
|
| 840 |
+
"step": 6850
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 2.11,
|
| 844 |
+
"learning_rate": 5.949908368967624e-06,
|
| 845 |
+
"loss": 0.1623,
|
| 846 |
+
"step": 6900
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.12,
|
| 850 |
+
"learning_rate": 5.848096110771737e-06,
|
| 851 |
+
"loss": 0.1595,
|
| 852 |
+
"step": 6950
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.14,
|
| 856 |
+
"learning_rate": 5.74628385257585e-06,
|
| 857 |
+
"loss": 0.1591,
|
| 858 |
+
"step": 7000
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 2.15,
|
| 862 |
+
"learning_rate": 5.644471594379963e-06,
|
| 863 |
+
"loss": 0.1363,
|
| 864 |
+
"step": 7050
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.17,
|
| 868 |
+
"learning_rate": 5.542659336184077e-06,
|
| 869 |
+
"loss": 0.1721,
|
| 870 |
+
"step": 7100
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.18,
|
| 874 |
+
"learning_rate": 5.44084707798819e-06,
|
| 875 |
+
"loss": 0.1601,
|
| 876 |
+
"step": 7150
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 2.2,
|
| 880 |
+
"learning_rate": 5.339034819792304e-06,
|
| 881 |
+
"loss": 0.1598,
|
| 882 |
+
"step": 7200
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 2.21,
|
| 886 |
+
"learning_rate": 5.237222561596417e-06,
|
| 887 |
+
"loss": 0.1455,
|
| 888 |
+
"step": 7250
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.23,
|
| 892 |
+
"learning_rate": 5.1354103034005295e-06,
|
| 893 |
+
"loss": 0.1465,
|
| 894 |
+
"step": 7300
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.24,
|
| 898 |
+
"learning_rate": 5.033598045204643e-06,
|
| 899 |
+
"loss": 0.1432,
|
| 900 |
+
"step": 7350
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 2.26,
|
| 904 |
+
"learning_rate": 4.931785787008756e-06,
|
| 905 |
+
"loss": 0.1617,
|
| 906 |
+
"step": 7400
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.28,
|
| 910 |
+
"learning_rate": 4.82997352881287e-06,
|
| 911 |
+
"loss": 0.1588,
|
| 912 |
+
"step": 7450
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.29,
|
| 916 |
+
"learning_rate": 4.728161270616982e-06,
|
| 917 |
+
"loss": 0.1799,
|
| 918 |
+
"step": 7500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 2.31,
|
| 922 |
+
"learning_rate": 4.626349012421096e-06,
|
| 923 |
+
"loss": 0.1657,
|
| 924 |
+
"step": 7550
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 2.32,
|
| 928 |
+
"learning_rate": 4.524536754225209e-06,
|
| 929 |
+
"loss": 0.1534,
|
| 930 |
+
"step": 7600
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.34,
|
| 934 |
+
"learning_rate": 4.4227244960293225e-06,
|
| 935 |
+
"loss": 0.1615,
|
| 936 |
+
"step": 7650
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 2.35,
|
| 940 |
+
"learning_rate": 4.320912237833435e-06,
|
| 941 |
+
"loss": 0.1695,
|
| 942 |
+
"step": 7700
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 2.37,
|
| 946 |
+
"learning_rate": 4.219099979637549e-06,
|
| 947 |
+
"loss": 0.146,
|
| 948 |
+
"step": 7750
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.38,
|
| 952 |
+
"learning_rate": 4.117287721441662e-06,
|
| 953 |
+
"loss": 0.1575,
|
| 954 |
+
"step": 7800
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 2.4,
|
| 958 |
+
"learning_rate": 4.015475463245775e-06,
|
| 959 |
+
"loss": 0.1711,
|
| 960 |
+
"step": 7850
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 2.41,
|
| 964 |
+
"learning_rate": 3.913663205049888e-06,
|
| 965 |
+
"loss": 0.1517,
|
| 966 |
+
"step": 7900
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 2.43,
|
| 970 |
+
"learning_rate": 3.8118509468540015e-06,
|
| 971 |
+
"loss": 0.146,
|
| 972 |
+
"step": 7950
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.44,
|
| 976 |
+
"learning_rate": 3.7100386886581147e-06,
|
| 977 |
+
"loss": 0.1675,
|
| 978 |
+
"step": 8000
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.46,
|
| 982 |
+
"learning_rate": 3.6082264304622276e-06,
|
| 983 |
+
"loss": 0.1878,
|
| 984 |
+
"step": 8050
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 2.47,
|
| 988 |
+
"learning_rate": 3.506414172266341e-06,
|
| 989 |
+
"loss": 0.1595,
|
| 990 |
+
"step": 8100
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.49,
|
| 994 |
+
"learning_rate": 3.4046019140704544e-06,
|
| 995 |
+
"loss": 0.1779,
|
| 996 |
+
"step": 8150
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 2.5,
|
| 1000 |
+
"learning_rate": 3.3027896558745677e-06,
|
| 1001 |
+
"loss": 0.1196,
|
| 1002 |
+
"step": 8200
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 2.52,
|
| 1006 |
+
"learning_rate": 3.2009773976786805e-06,
|
| 1007 |
+
"loss": 0.1632,
|
| 1008 |
+
"step": 8250
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 2.54,
|
| 1012 |
+
"learning_rate": 3.0991651394827937e-06,
|
| 1013 |
+
"loss": 0.1533,
|
| 1014 |
+
"step": 8300
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.55,
|
| 1018 |
+
"learning_rate": 2.9973528812869074e-06,
|
| 1019 |
+
"loss": 0.1701,
|
| 1020 |
+
"step": 8350
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.57,
|
| 1024 |
+
"learning_rate": 2.8955406230910206e-06,
|
| 1025 |
+
"loss": 0.1659,
|
| 1026 |
+
"step": 8400
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 2.58,
|
| 1030 |
+
"learning_rate": 2.7937283648951334e-06,
|
| 1031 |
+
"loss": 0.1335,
|
| 1032 |
+
"step": 8450
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.6,
|
| 1036 |
+
"learning_rate": 2.6919161066992467e-06,
|
| 1037 |
+
"loss": 0.1747,
|
| 1038 |
+
"step": 8500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 2.61,
|
| 1042 |
+
"learning_rate": 2.5901038485033603e-06,
|
| 1043 |
+
"loss": 0.175,
|
| 1044 |
+
"step": 8550
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 2.63,
|
| 1048 |
+
"learning_rate": 2.4882915903074735e-06,
|
| 1049 |
+
"loss": 0.1459,
|
| 1050 |
+
"step": 8600
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 2.64,
|
| 1054 |
+
"learning_rate": 2.3864793321115864e-06,
|
| 1055 |
+
"loss": 0.1491,
|
| 1056 |
+
"step": 8650
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.66,
|
| 1060 |
+
"learning_rate": 2.2846670739156996e-06,
|
| 1061 |
+
"loss": 0.1535,
|
| 1062 |
+
"step": 8700
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 2.67,
|
| 1066 |
+
"learning_rate": 2.182854815719813e-06,
|
| 1067 |
+
"loss": 0.1559,
|
| 1068 |
+
"step": 8750
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 2.69,
|
| 1072 |
+
"learning_rate": 2.081042557523926e-06,
|
| 1073 |
+
"loss": 0.1391,
|
| 1074 |
+
"step": 8800
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.7,
|
| 1078 |
+
"learning_rate": 1.9792302993280393e-06,
|
| 1079 |
+
"loss": 0.1601,
|
| 1080 |
+
"step": 8850
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 2.72,
|
| 1084 |
+
"learning_rate": 1.8774180411321525e-06,
|
| 1085 |
+
"loss": 0.1541,
|
| 1086 |
+
"step": 8900
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 2.73,
|
| 1090 |
+
"learning_rate": 1.7756057829362658e-06,
|
| 1091 |
+
"loss": 0.1495,
|
| 1092 |
+
"step": 8950
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 2.75,
|
| 1096 |
+
"learning_rate": 1.6737935247403788e-06,
|
| 1097 |
+
"loss": 0.1338,
|
| 1098 |
+
"step": 9000
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.76,
|
| 1102 |
+
"learning_rate": 1.571981266544492e-06,
|
| 1103 |
+
"loss": 0.1569,
|
| 1104 |
+
"step": 9050
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.78,
|
| 1108 |
+
"learning_rate": 1.4701690083486052e-06,
|
| 1109 |
+
"loss": 0.17,
|
| 1110 |
+
"step": 9100
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 2.79,
|
| 1114 |
+
"learning_rate": 1.3683567501527185e-06,
|
| 1115 |
+
"loss": 0.1694,
|
| 1116 |
+
"step": 9150
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.81,
|
| 1120 |
+
"learning_rate": 1.2665444919568317e-06,
|
| 1121 |
+
"loss": 0.1528,
|
| 1122 |
+
"step": 9200
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 2.83,
|
| 1126 |
+
"learning_rate": 1.164732233760945e-06,
|
| 1127 |
+
"loss": 0.1541,
|
| 1128 |
+
"step": 9250
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.84,
|
| 1132 |
+
"learning_rate": 1.0629199755650582e-06,
|
| 1133 |
+
"loss": 0.1657,
|
| 1134 |
+
"step": 9300
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 2.86,
|
| 1138 |
+
"learning_rate": 9.611077173691712e-07,
|
| 1139 |
+
"loss": 0.1646,
|
| 1140 |
+
"step": 9350
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.87,
|
| 1144 |
+
"learning_rate": 8.592954591732845e-07,
|
| 1145 |
+
"loss": 0.1597,
|
| 1146 |
+
"step": 9400
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.89,
|
| 1150 |
+
"learning_rate": 7.574832009773978e-07,
|
| 1151 |
+
"loss": 0.1408,
|
| 1152 |
+
"step": 9450
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 2.9,
|
| 1156 |
+
"learning_rate": 6.55670942781511e-07,
|
| 1157 |
+
"loss": 0.1676,
|
| 1158 |
+
"step": 9500
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.92,
|
| 1162 |
+
"learning_rate": 5.538586845856241e-07,
|
| 1163 |
+
"loss": 0.1364,
|
| 1164 |
+
"step": 9550
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 2.93,
|
| 1168 |
+
"learning_rate": 4.5204642638973736e-07,
|
| 1169 |
+
"loss": 0.1571,
|
| 1170 |
+
"step": 9600
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.95,
|
| 1174 |
+
"learning_rate": 3.502341681938506e-07,
|
| 1175 |
+
"loss": 0.1537,
|
| 1176 |
+
"step": 9650
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 2.96,
|
| 1180 |
+
"learning_rate": 2.484219099979638e-07,
|
| 1181 |
+
"loss": 0.1574,
|
| 1182 |
+
"step": 9700
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.98,
|
| 1186 |
+
"learning_rate": 1.4660965180207698e-07,
|
| 1187 |
+
"loss": 0.1345,
|
| 1188 |
+
"step": 9750
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 2.99,
|
| 1192 |
+
"learning_rate": 4.4797393606190183e-08,
|
| 1193 |
+
"loss": 0.1483,
|
| 1194 |
+
"step": 9800
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 3.0,
|
| 1198 |
+
"step": 9822,
|
| 1199 |
+
"total_flos": 7.333639468651878e+16,
|
| 1200 |
+
"train_loss": 0.25678691951885313,
|
| 1201 |
+
"train_runtime": 5962.3369,
|
| 1202 |
+
"train_samples_per_second": 52.702,
|
| 1203 |
+
"train_steps_per_second": 1.647
|
| 1204 |
+
}
|
| 1205 |
+
],
|
| 1206 |
+
"logging_steps": 50,
|
| 1207 |
+
"max_steps": 9822,
|
| 1208 |
+
"num_input_tokens_seen": 0,
|
| 1209 |
+
"num_train_epochs": 3,
|
| 1210 |
+
"save_steps": 5000,
|
| 1211 |
+
"total_flos": 7.333639468651878e+16,
|
| 1212 |
+
"train_batch_size": 32,
|
| 1213 |
+
"trial_name": null,
|
| 1214 |
+
"trial_params": null
|
| 1215 |
+
}
|