Upload trainer_state.json
Browse files- trainer_state.json +1205 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1205 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9982003599280143,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 832,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01199760047990402,
|
| 14 |
+
"grad_norm": 8.655085086138964,
|
| 15 |
+
"learning_rate": 2.9761904761904763e-06,
|
| 16 |
+
"loss": 1.9665,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02399520095980804,
|
| 21 |
+
"grad_norm": 4.415158437946272,
|
| 22 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 23 |
+
"loss": 1.6691,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.035992801439712056,
|
| 28 |
+
"grad_norm": 3.057605663016171,
|
| 29 |
+
"learning_rate": 8.92857142857143e-06,
|
| 30 |
+
"loss": 1.441,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04799040191961608,
|
| 35 |
+
"grad_norm": 3.042012061427229,
|
| 36 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 37 |
+
"loss": 1.3171,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.059988002399520096,
|
| 42 |
+
"grad_norm": 2.1436915691999014,
|
| 43 |
+
"learning_rate": 1.4880952380952381e-05,
|
| 44 |
+
"loss": 1.2588,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.07198560287942411,
|
| 49 |
+
"grad_norm": 2.03272405774148,
|
| 50 |
+
"learning_rate": 1.785714285714286e-05,
|
| 51 |
+
"loss": 1.2021,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.08398320335932813,
|
| 56 |
+
"grad_norm": 1.9317364452662562,
|
| 57 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 58 |
+
"loss": 1.1645,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.09598080383923216,
|
| 63 |
+
"grad_norm": 1.9450167856079439,
|
| 64 |
+
"learning_rate": 2.380952380952381e-05,
|
| 65 |
+
"loss": 1.1652,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.10797840431913618,
|
| 70 |
+
"grad_norm": 2.2175764125145943,
|
| 71 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 72 |
+
"loss": 1.1492,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.11997600479904019,
|
| 77 |
+
"grad_norm": 1.9900888357541513,
|
| 78 |
+
"learning_rate": 2.9761904761904762e-05,
|
| 79 |
+
"loss": 1.146,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.13197360527894422,
|
| 84 |
+
"grad_norm": 2.1707343241074715,
|
| 85 |
+
"learning_rate": 3.273809523809524e-05,
|
| 86 |
+
"loss": 1.1312,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.14397120575884823,
|
| 91 |
+
"grad_norm": 2.1008421926969874,
|
| 92 |
+
"learning_rate": 3.571428571428572e-05,
|
| 93 |
+
"loss": 1.1158,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.15596880623875226,
|
| 98 |
+
"grad_norm": 1.9739168298796461,
|
| 99 |
+
"learning_rate": 3.8690476190476195e-05,
|
| 100 |
+
"loss": 1.112,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.16796640671865626,
|
| 105 |
+
"grad_norm": 1.8793192108758627,
|
| 106 |
+
"learning_rate": 4.166666666666667e-05,
|
| 107 |
+
"loss": 1.1216,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.1799640071985603,
|
| 112 |
+
"grad_norm": 2.1680515905272393,
|
| 113 |
+
"learning_rate": 4.464285714285715e-05,
|
| 114 |
+
"loss": 1.1154,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.19196160767846432,
|
| 119 |
+
"grad_norm": 2.088302206757574,
|
| 120 |
+
"learning_rate": 4.761904761904762e-05,
|
| 121 |
+
"loss": 1.115,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.20395920815836832,
|
| 126 |
+
"grad_norm": 2.0040693696661096,
|
| 127 |
+
"learning_rate": 4.9999779501355384e-05,
|
| 128 |
+
"loss": 1.1166,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.21595680863827235,
|
| 133 |
+
"grad_norm": 1.8604233679285211,
|
| 134 |
+
"learning_rate": 4.9992062457191e-05,
|
| 135 |
+
"loss": 1.126,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.22795440911817635,
|
| 140 |
+
"grad_norm": 1.8293195388484245,
|
| 141 |
+
"learning_rate": 4.997332437005931e-05,
|
| 142 |
+
"loss": 1.1183,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.23995200959808038,
|
| 147 |
+
"grad_norm": 1.9227439225768022,
|
| 148 |
+
"learning_rate": 4.99435735031144e-05,
|
| 149 |
+
"loss": 1.1341,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.2519496100779844,
|
| 154 |
+
"grad_norm": 1.7353567528325675,
|
| 155 |
+
"learning_rate": 4.990282297594509e-05,
|
| 156 |
+
"loss": 1.1111,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.26394721055788845,
|
| 161 |
+
"grad_norm": 1.6246909472281814,
|
| 162 |
+
"learning_rate": 4.98510907587894e-05,
|
| 163 |
+
"loss": 1.1085,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.27594481103779245,
|
| 168 |
+
"grad_norm": 1.5550505481063543,
|
| 169 |
+
"learning_rate": 4.9788399664609985e-05,
|
| 170 |
+
"loss": 1.1081,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.28794241151769645,
|
| 175 |
+
"grad_norm": 1.6242317436366518,
|
| 176 |
+
"learning_rate": 4.97147773390341e-05,
|
| 177 |
+
"loss": 1.1187,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.29994001199760045,
|
| 182 |
+
"grad_norm": 1.6719313503244617,
|
| 183 |
+
"learning_rate": 4.963025624816232e-05,
|
| 184 |
+
"loss": 1.0934,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.3119376124775045,
|
| 189 |
+
"grad_norm": 1.7557488929934109,
|
| 190 |
+
"learning_rate": 4.953487366425163e-05,
|
| 191 |
+
"loss": 1.0939,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3239352129574085,
|
| 196 |
+
"grad_norm": 1.846886767486188,
|
| 197 |
+
"learning_rate": 4.942867164927899e-05,
|
| 198 |
+
"loss": 1.1016,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3359328134373125,
|
| 203 |
+
"grad_norm": 1.564323242123012,
|
| 204 |
+
"learning_rate": 4.931169703639282e-05,
|
| 205 |
+
"loss": 1.0755,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.3479304139172166,
|
| 210 |
+
"grad_norm": 1.5659567424784107,
|
| 211 |
+
"learning_rate": 4.918400140926042e-05,
|
| 212 |
+
"loss": 1.0772,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.3599280143971206,
|
| 217 |
+
"grad_norm": 1.6772397525925744,
|
| 218 |
+
"learning_rate": 4.9045641079320484e-05,
|
| 219 |
+
"loss": 1.0975,
|
| 220 |
+
"step": 150
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.3719256148770246,
|
| 224 |
+
"grad_norm": 1.4057626611756922,
|
| 225 |
+
"learning_rate": 4.889667706095084e-05,
|
| 226 |
+
"loss": 1.0894,
|
| 227 |
+
"step": 155
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.38392321535692864,
|
| 231 |
+
"grad_norm": 1.3508523989809198,
|
| 232 |
+
"learning_rate": 4.873717504456219e-05,
|
| 233 |
+
"loss": 1.0899,
|
| 234 |
+
"step": 160
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.39592081583683264,
|
| 238 |
+
"grad_norm": 1.4599727770410549,
|
| 239 |
+
"learning_rate": 4.8567205367629835e-05,
|
| 240 |
+
"loss": 1.0954,
|
| 241 |
+
"step": 165
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.40791841631673664,
|
| 245 |
+
"grad_norm": 1.557943385459695,
|
| 246 |
+
"learning_rate": 4.8386842983676164e-05,
|
| 247 |
+
"loss": 1.1022,
|
| 248 |
+
"step": 170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.41991601679664065,
|
| 252 |
+
"grad_norm": 1.4392521405523333,
|
| 253 |
+
"learning_rate": 4.8196167429217474e-05,
|
| 254 |
+
"loss": 1.0858,
|
| 255 |
+
"step": 175
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.4319136172765447,
|
| 259 |
+
"grad_norm": 1.4039232690942924,
|
| 260 |
+
"learning_rate": 4.799526278868987e-05,
|
| 261 |
+
"loss": 1.0776,
|
| 262 |
+
"step": 180
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4439112177564487,
|
| 266 |
+
"grad_norm": 1.3958637024927205,
|
| 267 |
+
"learning_rate": 4.778421765736951e-05,
|
| 268 |
+
"loss": 1.0656,
|
| 269 |
+
"step": 185
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.4559088182363527,
|
| 273 |
+
"grad_norm": 1.3165094843488907,
|
| 274 |
+
"learning_rate": 4.7563125102303766e-05,
|
| 275 |
+
"loss": 1.0701,
|
| 276 |
+
"step": 190
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.46790641871625677,
|
| 280 |
+
"grad_norm": 1.4462822477245483,
|
| 281 |
+
"learning_rate": 4.7332082621270326e-05,
|
| 282 |
+
"loss": 1.078,
|
| 283 |
+
"step": 195
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.47990401919616077,
|
| 287 |
+
"grad_norm": 1.4177688615580946,
|
| 288 |
+
"learning_rate": 4.709119209978242e-05,
|
| 289 |
+
"loss": 1.0748,
|
| 290 |
+
"step": 200
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.49190161967606477,
|
| 294 |
+
"grad_norm": 1.2487244087937184,
|
| 295 |
+
"learning_rate": 4.684055976615924e-05,
|
| 296 |
+
"loss": 1.0728,
|
| 297 |
+
"step": 205
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5038992201559688,
|
| 301 |
+
"grad_norm": 1.2843589104522166,
|
| 302 |
+
"learning_rate": 4.6580296144681157e-05,
|
| 303 |
+
"loss": 1.0741,
|
| 304 |
+
"step": 210
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5158968206358728,
|
| 308 |
+
"grad_norm": 1.3423526627782842,
|
| 309 |
+
"learning_rate": 4.631051600685051e-05,
|
| 310 |
+
"loss": 1.0545,
|
| 311 |
+
"step": 215
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.5278944211157769,
|
| 315 |
+
"grad_norm": 1.3987872492900897,
|
| 316 |
+
"learning_rate": 4.6031338320779534e-05,
|
| 317 |
+
"loss": 1.0466,
|
| 318 |
+
"step": 220
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.5398920215956808,
|
| 322 |
+
"grad_norm": 1.1228691632454497,
|
| 323 |
+
"learning_rate": 4.57428861987275e-05,
|
| 324 |
+
"loss": 1.0757,
|
| 325 |
+
"step": 225
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.5518896220755849,
|
| 329 |
+
"grad_norm": 1.4386106052561902,
|
| 330 |
+
"learning_rate": 4.544528684281056e-05,
|
| 331 |
+
"loss": 1.0659,
|
| 332 |
+
"step": 230
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.563887222555489,
|
| 336 |
+
"grad_norm": 1.54532036456569,
|
| 337 |
+
"learning_rate": 4.513867148890788e-05,
|
| 338 |
+
"loss": 1.0731,
|
| 339 |
+
"step": 235
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.5758848230353929,
|
| 343 |
+
"grad_norm": 1.1946309045723265,
|
| 344 |
+
"learning_rate": 4.482317534878901e-05,
|
| 345 |
+
"loss": 1.0724,
|
| 346 |
+
"step": 240
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.587882423515297,
|
| 350 |
+
"grad_norm": 1.2995921290959758,
|
| 351 |
+
"learning_rate": 4.449893755048799e-05,
|
| 352 |
+
"loss": 1.0538,
|
| 353 |
+
"step": 245
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.5998800239952009,
|
| 357 |
+
"grad_norm": 1.1629232540517822,
|
| 358 |
+
"learning_rate": 4.416610107695042e-05,
|
| 359 |
+
"loss": 1.0653,
|
| 360 |
+
"step": 250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.611877624475105,
|
| 364 |
+
"grad_norm": 1.141699716294497,
|
| 365 |
+
"learning_rate": 4.3824812702980595e-05,
|
| 366 |
+
"loss": 1.0593,
|
| 367 |
+
"step": 255
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.623875224955009,
|
| 371 |
+
"grad_norm": 1.1625584379097018,
|
| 372 |
+
"learning_rate": 4.347522293051648e-05,
|
| 373 |
+
"loss": 1.0728,
|
| 374 |
+
"step": 260
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.635872825434913,
|
| 378 |
+
"grad_norm": 1.2139933398328908,
|
| 379 |
+
"learning_rate": 4.3117485922261136e-05,
|
| 380 |
+
"loss": 1.0683,
|
| 381 |
+
"step": 265
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.647870425914817,
|
| 385 |
+
"grad_norm": 1.1676365588651592,
|
| 386 |
+
"learning_rate": 4.275175943369975e-05,
|
| 387 |
+
"loss": 1.0493,
|
| 388 |
+
"step": 270
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.6598680263947211,
|
| 392 |
+
"grad_norm": 1.2056917804998961,
|
| 393 |
+
"learning_rate": 4.2378204743532377e-05,
|
| 394 |
+
"loss": 1.0779,
|
| 395 |
+
"step": 275
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.671865626874625,
|
| 399 |
+
"grad_norm": 1.356550764377353,
|
| 400 |
+
"learning_rate": 4.199698658255298e-05,
|
| 401 |
+
"loss": 1.0507,
|
| 402 |
+
"step": 280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.6838632273545291,
|
| 406 |
+
"grad_norm": 1.1721573272728156,
|
| 407 |
+
"learning_rate": 4.160827306100611e-05,
|
| 408 |
+
"loss": 1.0541,
|
| 409 |
+
"step": 285
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.6958608278344331,
|
| 413 |
+
"grad_norm": 1.1802890937435413,
|
| 414 |
+
"learning_rate": 4.121223559445343e-05,
|
| 415 |
+
"loss": 1.0507,
|
| 416 |
+
"step": 290
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7078584283143371,
|
| 420 |
+
"grad_norm": 1.3591231013602916,
|
| 421 |
+
"learning_rate": 4.0809048828182534e-05,
|
| 422 |
+
"loss": 1.0529,
|
| 423 |
+
"step": 295
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.7198560287942412,
|
| 427 |
+
"grad_norm": 1.230075489545995,
|
| 428 |
+
"learning_rate": 4.039889056019159e-05,
|
| 429 |
+
"loss": 1.0611,
|
| 430 |
+
"step": 300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.7318536292741452,
|
| 434 |
+
"grad_norm": 1.3962403495748514,
|
| 435 |
+
"learning_rate": 3.9981941662783674e-05,
|
| 436 |
+
"loss": 1.0477,
|
| 437 |
+
"step": 305
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.7438512297540492,
|
| 441 |
+
"grad_norm": 1.2104058914840572,
|
| 442 |
+
"learning_rate": 3.955838600280535e-05,
|
| 443 |
+
"loss": 1.0425,
|
| 444 |
+
"step": 310
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.7558488302339532,
|
| 448 |
+
"grad_norm": 1.243572445349006,
|
| 449 |
+
"learning_rate": 3.91284103605648e-05,
|
| 450 |
+
"loss": 1.053,
|
| 451 |
+
"step": 315
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.7678464307138573,
|
| 455 |
+
"grad_norm": 1.2315818992929564,
|
| 456 |
+
"learning_rate": 3.869220434746509e-05,
|
| 457 |
+
"loss": 1.0469,
|
| 458 |
+
"step": 320
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.7798440311937612,
|
| 462 |
+
"grad_norm": 1.282062911918751,
|
| 463 |
+
"learning_rate": 3.8249960322389e-05,
|
| 464 |
+
"loss": 1.0507,
|
| 465 |
+
"step": 325
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.7918416316736653,
|
| 469 |
+
"grad_norm": 1.2926792700193703,
|
| 470 |
+
"learning_rate": 3.780187330687231e-05,
|
| 471 |
+
"loss": 1.0286,
|
| 472 |
+
"step": 330
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8038392321535693,
|
| 476 |
+
"grad_norm": 1.1789915377278588,
|
| 477 |
+
"learning_rate": 3.734814089910283e-05,
|
| 478 |
+
"loss": 1.0275,
|
| 479 |
+
"step": 335
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.8158368326334733,
|
| 483 |
+
"grad_norm": 1.2214772745932698,
|
| 484 |
+
"learning_rate": 3.6888963186783224e-05,
|
| 485 |
+
"loss": 1.0152,
|
| 486 |
+
"step": 340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.8278344331133773,
|
| 490 |
+
"grad_norm": 1.0572719303807887,
|
| 491 |
+
"learning_rate": 3.6424542658895944e-05,
|
| 492 |
+
"loss": 1.0329,
|
| 493 |
+
"step": 345
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.8398320335932813,
|
| 497 |
+
"grad_norm": 1.057795278450336,
|
| 498 |
+
"learning_rate": 3.5955084116409385e-05,
|
| 499 |
+
"loss": 1.0358,
|
| 500 |
+
"step": 350
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.8518296340731853,
|
| 504 |
+
"grad_norm": 1.1933118891898369,
|
| 505 |
+
"learning_rate": 3.5480794581964304e-05,
|
| 506 |
+
"loss": 1.0257,
|
| 507 |
+
"step": 355
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.8638272345530894,
|
| 511 |
+
"grad_norm": 1.1930511530685082,
|
| 512 |
+
"learning_rate": 3.5001883208580665e-05,
|
| 513 |
+
"loss": 1.0462,
|
| 514 |
+
"step": 360
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.8758248350329934,
|
| 518 |
+
"grad_norm": 1.206184339739026,
|
| 519 |
+
"learning_rate": 3.451856118742498e-05,
|
| 520 |
+
"loss": 1.0457,
|
| 521 |
+
"step": 365
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.8878224355128974,
|
| 525 |
+
"grad_norm": 1.2488722797945375,
|
| 526 |
+
"learning_rate": 3.403104165467883e-05,
|
| 527 |
+
"loss": 1.0273,
|
| 528 |
+
"step": 370
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.8998200359928015,
|
| 532 |
+
"grad_norm": 1.1879602242889826,
|
| 533 |
+
"learning_rate": 3.353953959754973e-05,
|
| 534 |
+
"loss": 1.0063,
|
| 535 |
+
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.9118176364727054,
|
| 539 |
+
"grad_norm": 1.1859307890368762,
|
| 540 |
+
"learning_rate": 3.30442717594657e-05,
|
| 541 |
+
"loss": 1.0155,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.9238152369526095,
|
| 546 |
+
"grad_norm": 1.1938213692010637,
|
| 547 |
+
"learning_rate": 3.2545456544495365e-05,
|
| 548 |
+
"loss": 1.025,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.9358128374325135,
|
| 553 |
+
"grad_norm": 1.1843748330164985,
|
| 554 |
+
"learning_rate": 3.2043313921035743e-05,
|
| 555 |
+
"loss": 1.0328,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.9478104379124175,
|
| 560 |
+
"grad_norm": 1.117527742694121,
|
| 561 |
+
"learning_rate": 3.1538065324810206e-05,
|
| 562 |
+
"loss": 1.0248,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.9598080383923215,
|
| 567 |
+
"grad_norm": 1.1115869963103877,
|
| 568 |
+
"learning_rate": 3.1029933561219375e-05,
|
| 569 |
+
"loss": 1.0027,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.9718056388722256,
|
| 574 |
+
"grad_norm": 1.0498821054649277,
|
| 575 |
+
"learning_rate": 3.0519142707088026e-05,
|
| 576 |
+
"loss": 1.0116,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.9838032393521295,
|
| 581 |
+
"grad_norm": 1.118470935254374,
|
| 582 |
+
"learning_rate": 3.000591801185124e-05,
|
| 583 |
+
"loss": 1.0147,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.9958008398320336,
|
| 588 |
+
"grad_norm": 1.1283471897635804,
|
| 589 |
+
"learning_rate": 2.9490485798223623e-05,
|
| 590 |
+
"loss": 1.0312,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.0095980803839233,
|
| 595 |
+
"grad_norm": 1.0933704505218012,
|
| 596 |
+
"learning_rate": 2.8973073362394998e-05,
|
| 597 |
+
"loss": 1.0189,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.0215956808638271,
|
| 602 |
+
"grad_norm": 1.2757088787270183,
|
| 603 |
+
"learning_rate": 2.8453908873797058e-05,
|
| 604 |
+
"loss": 0.8115,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.0335932813437312,
|
| 609 |
+
"grad_norm": 1.2014338555644504,
|
| 610 |
+
"learning_rate": 2.7933221274484723e-05,
|
| 611 |
+
"loss": 0.8041,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.0455908818236352,
|
| 616 |
+
"grad_norm": 1.136417187679828,
|
| 617 |
+
"learning_rate": 2.7411240178176927e-05,
|
| 618 |
+
"loss": 0.8078,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.0575884823035393,
|
| 623 |
+
"grad_norm": 1.1276372766792626,
|
| 624 |
+
"learning_rate": 2.6888195769001146e-05,
|
| 625 |
+
"loss": 0.8179,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.0695860827834434,
|
| 630 |
+
"grad_norm": 1.0072604306741504,
|
| 631 |
+
"learning_rate": 2.63643186999864e-05,
|
| 632 |
+
"loss": 0.8072,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.0815836832633474,
|
| 637 |
+
"grad_norm": 1.089321595722244,
|
| 638 |
+
"learning_rate": 2.5839839991349506e-05,
|
| 639 |
+
"loss": 0.8099,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.0935812837432513,
|
| 644 |
+
"grad_norm": 1.0056285329535712,
|
| 645 |
+
"learning_rate": 2.5314990928619337e-05,
|
| 646 |
+
"loss": 0.7985,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.1055788842231553,
|
| 651 |
+
"grad_norm": 0.9752484724026733,
|
| 652 |
+
"learning_rate": 2.479000296064417e-05,
|
| 653 |
+
"loss": 0.8224,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.1175764847030594,
|
| 658 |
+
"grad_norm": 1.0643013774775827,
|
| 659 |
+
"learning_rate": 2.4265107597526946e-05,
|
| 660 |
+
"loss": 0.8156,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.1295740851829634,
|
| 665 |
+
"grad_norm": 1.1300899339981707,
|
| 666 |
+
"learning_rate": 2.374053630853358e-05,
|
| 667 |
+
"loss": 0.8113,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.1415716856628675,
|
| 672 |
+
"grad_norm": 1.107379832173596,
|
| 673 |
+
"learning_rate": 2.3216520420019195e-05,
|
| 674 |
+
"loss": 0.8078,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.1535692861427713,
|
| 679 |
+
"grad_norm": 1.1381056597190673,
|
| 680 |
+
"learning_rate": 2.2693291013417453e-05,
|
| 681 |
+
"loss": 0.7979,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.1655668866226754,
|
| 686 |
+
"grad_norm": 1.0344901370146822,
|
| 687 |
+
"learning_rate": 2.2171078823337863e-05,
|
| 688 |
+
"loss": 0.8114,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.1775644871025794,
|
| 693 |
+
"grad_norm": 1.1433623054658555,
|
| 694 |
+
"learning_rate": 2.165011413581605e-05,
|
| 695 |
+
"loss": 0.8068,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.1895620875824835,
|
| 700 |
+
"grad_norm": 1.0574705191297809,
|
| 701 |
+
"learning_rate": 2.1130626686761762e-05,
|
| 702 |
+
"loss": 0.7931,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.2015596880623876,
|
| 707 |
+
"grad_norm": 1.0955960697193892,
|
| 708 |
+
"learning_rate": 2.0612845560649603e-05,
|
| 709 |
+
"loss": 0.7961,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.2135572885422916,
|
| 714 |
+
"grad_norm": 1.1142604155745692,
|
| 715 |
+
"learning_rate": 2.0096999089496913e-05,
|
| 716 |
+
"loss": 0.7949,
|
| 717 |
+
"step": 505
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.2255548890221957,
|
| 721 |
+
"grad_norm": 1.081207113394286,
|
| 722 |
+
"learning_rate": 1.958331475217357e-05,
|
| 723 |
+
"loss": 0.8081,
|
| 724 |
+
"step": 510
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.2375524895020995,
|
| 728 |
+
"grad_norm": 1.087424227244276,
|
| 729 |
+
"learning_rate": 1.9072019074087876e-05,
|
| 730 |
+
"loss": 0.7944,
|
| 731 |
+
"step": 515
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.2495500899820036,
|
| 735 |
+
"grad_norm": 1.1746075874674693,
|
| 736 |
+
"learning_rate": 1.856333752729311e-05,
|
| 737 |
+
"loss": 0.7982,
|
| 738 |
+
"step": 520
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.2615476904619076,
|
| 742 |
+
"grad_norm": 1.1261850868143939,
|
| 743 |
+
"learning_rate": 1.8057494431058365e-05,
|
| 744 |
+
"loss": 0.8013,
|
| 745 |
+
"step": 525
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.2735452909418117,
|
| 749 |
+
"grad_norm": 1.0858958449694596,
|
| 750 |
+
"learning_rate": 1.7554712852947913e-05,
|
| 751 |
+
"loss": 0.8031,
|
| 752 |
+
"step": 530
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.2855428914217157,
|
| 756 |
+
"grad_norm": 1.1039599626344443,
|
| 757 |
+
"learning_rate": 1.705521451045246e-05,
|
| 758 |
+
"loss": 0.7924,
|
| 759 |
+
"step": 535
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.2975404919016196,
|
| 763 |
+
"grad_norm": 1.0379577155650672,
|
| 764 |
+
"learning_rate": 1.6559219673215784e-05,
|
| 765 |
+
"loss": 0.784,
|
| 766 |
+
"step": 540
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.3095380923815236,
|
| 770 |
+
"grad_norm": 1.0180533259182083,
|
| 771 |
+
"learning_rate": 1.6066947065899847e-05,
|
| 772 |
+
"loss": 0.795,
|
| 773 |
+
"step": 545
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.3215356928614277,
|
| 777 |
+
"grad_norm": 1.1174277968751472,
|
| 778 |
+
"learning_rate": 1.5578613771731213e-05,
|
| 779 |
+
"loss": 0.7903,
|
| 780 |
+
"step": 550
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.3335332933413317,
|
| 784 |
+
"grad_norm": 1.113625762435688,
|
| 785 |
+
"learning_rate": 1.509443513677134e-05,
|
| 786 |
+
"loss": 0.7841,
|
| 787 |
+
"step": 555
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.3455308938212358,
|
| 791 |
+
"grad_norm": 1.046198809224864,
|
| 792 |
+
"learning_rate": 1.4614624674952842e-05,
|
| 793 |
+
"loss": 0.7932,
|
| 794 |
+
"step": 560
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.3575284943011399,
|
| 798 |
+
"grad_norm": 1.0511943871992961,
|
| 799 |
+
"learning_rate": 1.4139393973923798e-05,
|
| 800 |
+
"loss": 0.7815,
|
| 801 |
+
"step": 565
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.369526094781044,
|
| 805 |
+
"grad_norm": 1.012108314446409,
|
| 806 |
+
"learning_rate": 1.3668952601741441e-05,
|
| 807 |
+
"loss": 0.7899,
|
| 808 |
+
"step": 570
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.3815236952609478,
|
| 812 |
+
"grad_norm": 1.0590168334777237,
|
| 813 |
+
"learning_rate": 1.320350801445649e-05,
|
| 814 |
+
"loss": 0.7745,
|
| 815 |
+
"step": 575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.3935212957408518,
|
| 819 |
+
"grad_norm": 1.0208126844360186,
|
| 820 |
+
"learning_rate": 1.2743265464628786e-05,
|
| 821 |
+
"loss": 0.7761,
|
| 822 |
+
"step": 580
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.4055188962207559,
|
| 826 |
+
"grad_norm": 1.1048330533675996,
|
| 827 |
+
"learning_rate": 1.2288427910814699e-05,
|
| 828 |
+
"loss": 0.7549,
|
| 829 |
+
"step": 585
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.41751649670066,
|
| 833 |
+
"grad_norm": 1.087557213168186,
|
| 834 |
+
"learning_rate": 1.1839195928066102e-05,
|
| 835 |
+
"loss": 0.7887,
|
| 836 |
+
"step": 590
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.4295140971805638,
|
| 840 |
+
"grad_norm": 1.066790026980483,
|
| 841 |
+
"learning_rate": 1.1395767619480451e-05,
|
| 842 |
+
"loss": 0.7891,
|
| 843 |
+
"step": 595
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.4415116976604678,
|
| 847 |
+
"grad_norm": 1.0511422556482182,
|
| 848 |
+
"learning_rate": 1.0958338528840893e-05,
|
| 849 |
+
"loss": 0.7828,
|
| 850 |
+
"step": 600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.4535092981403719,
|
| 854 |
+
"grad_norm": 1.0811572565881435,
|
| 855 |
+
"learning_rate": 1.052710155438506e-05,
|
| 856 |
+
"loss": 0.7985,
|
| 857 |
+
"step": 605
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.465506898620276,
|
| 861 |
+
"grad_norm": 0.9838694729298545,
|
| 862 |
+
"learning_rate": 1.0102246863740496e-05,
|
| 863 |
+
"loss": 0.7712,
|
| 864 |
+
"step": 610
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.47750449910018,
|
| 868 |
+
"grad_norm": 1.039397195061213,
|
| 869 |
+
"learning_rate": 9.683961810064176e-06,
|
| 870 |
+
"loss": 0.7813,
|
| 871 |
+
"step": 615
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.489502099580084,
|
| 875 |
+
"grad_norm": 1.0652079719056124,
|
| 876 |
+
"learning_rate": 9.272430849423174e-06,
|
| 877 |
+
"loss": 0.774,
|
| 878 |
+
"step": 620
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.5014997000599881,
|
| 882 |
+
"grad_norm": 1.107963886239366,
|
| 883 |
+
"learning_rate": 8.867835459452925e-06,
|
| 884 |
+
"loss": 0.7756,
|
| 885 |
+
"step": 625
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.5134973005398922,
|
| 889 |
+
"grad_norm": 1.0470978361528913,
|
| 890 |
+
"learning_rate": 8.470354059328919e-06,
|
| 891 |
+
"loss": 0.7683,
|
| 892 |
+
"step": 630
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.525494901019796,
|
| 896 |
+
"grad_norm": 1.2007634054581608,
|
| 897 |
+
"learning_rate": 8.080161931087094e-06,
|
| 898 |
+
"loss": 0.7827,
|
| 899 |
+
"step": 635
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.5374925014997,
|
| 903 |
+
"grad_norm": 1.1547346563119187,
|
| 904 |
+
"learning_rate": 7.697431142327632e-06,
|
| 905 |
+
"loss": 0.7787,
|
| 906 |
+
"step": 640
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.5494901019796041,
|
| 910 |
+
"grad_norm": 1.17217513079497,
|
| 911 |
+
"learning_rate": 7.3223304703363135e-06,
|
| 912 |
+
"loss": 0.7858,
|
| 913 |
+
"step": 645
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.561487702459508,
|
| 917 |
+
"grad_norm": 1.0566429470152394,
|
| 918 |
+
"learning_rate": 6.955025327656839e-06,
|
| 919 |
+
"loss": 0.766,
|
| 920 |
+
"step": 650
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.573485302939412,
|
| 924 |
+
"grad_norm": 1.022660107263399,
|
| 925 |
+
"learning_rate": 6.5956776891468925e-06,
|
| 926 |
+
"loss": 0.7694,
|
| 927 |
+
"step": 655
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.585482903419316,
|
| 931 |
+
"grad_norm": 1.0475499247182583,
|
| 932 |
+
"learning_rate": 6.244446020550182e-06,
|
| 933 |
+
"loss": 0.7627,
|
| 934 |
+
"step": 660
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.5974805038992201,
|
| 938 |
+
"grad_norm": 1.0898061259359026,
|
| 939 |
+
"learning_rate": 5.901485208615948e-06,
|
| 940 |
+
"loss": 0.7672,
|
| 941 |
+
"step": 665
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.6094781043791242,
|
| 945 |
+
"grad_norm": 1.0594055484172262,
|
| 946 |
+
"learning_rate": 5.5669464927967655e-06,
|
| 947 |
+
"loss": 0.7783,
|
| 948 |
+
"step": 670
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.6214757048590283,
|
| 952 |
+
"grad_norm": 1.000104590500381,
|
| 953 |
+
"learning_rate": 5.240977398554673e-06,
|
| 954 |
+
"loss": 0.7578,
|
| 955 |
+
"step": 675
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.6334733053389323,
|
| 959 |
+
"grad_norm": 1.010755641973307,
|
| 960 |
+
"learning_rate": 4.9237216723051485e-06,
|
| 961 |
+
"loss": 0.7681,
|
| 962 |
+
"step": 680
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.6454709058188364,
|
| 966 |
+
"grad_norm": 1.0438538534480322,
|
| 967 |
+
"learning_rate": 4.615319218027561e-06,
|
| 968 |
+
"loss": 0.773,
|
| 969 |
+
"step": 685
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.6574685062987402,
|
| 973 |
+
"grad_norm": 1.00200099090404,
|
| 974 |
+
"learning_rate": 4.315906035570094e-06,
|
| 975 |
+
"loss": 0.7541,
|
| 976 |
+
"step": 690
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.6694661067786443,
|
| 980 |
+
"grad_norm": 0.9967511007903198,
|
| 981 |
+
"learning_rate": 4.0256141606762836e-06,
|
| 982 |
+
"loss": 0.7744,
|
| 983 |
+
"step": 695
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.6814637072585483,
|
| 987 |
+
"grad_norm": 0.954046235240287,
|
| 988 |
+
"learning_rate": 3.7445716067596503e-06,
|
| 989 |
+
"loss": 0.7472,
|
| 990 |
+
"step": 700
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.6934613077384522,
|
| 994 |
+
"grad_norm": 1.013820429314171,
|
| 995 |
+
"learning_rate": 3.4729023084521417e-06,
|
| 996 |
+
"loss": 0.7499,
|
| 997 |
+
"step": 705
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.7054589082183562,
|
| 1001 |
+
"grad_norm": 1.0142864719521136,
|
| 1002 |
+
"learning_rate": 3.2107260669512336e-06,
|
| 1003 |
+
"loss": 0.7539,
|
| 1004 |
+
"step": 710
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.7174565086982603,
|
| 1008 |
+
"grad_norm": 1.036622595232617,
|
| 1009 |
+
"learning_rate": 2.9581584971897697e-06,
|
| 1010 |
+
"loss": 0.7731,
|
| 1011 |
+
"step": 715
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.7294541091781643,
|
| 1015 |
+
"grad_norm": 1.02617492688862,
|
| 1016 |
+
"learning_rate": 2.7153109768518925e-06,
|
| 1017 |
+
"loss": 0.7495,
|
| 1018 |
+
"step": 720
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.7414517096580684,
|
| 1022 |
+
"grad_norm": 0.9783666612285059,
|
| 1023 |
+
"learning_rate": 2.4822905972575167e-06,
|
| 1024 |
+
"loss": 0.7649,
|
| 1025 |
+
"step": 725
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.7534493101379725,
|
| 1029 |
+
"grad_norm": 0.9915953729109124,
|
| 1030 |
+
"learning_rate": 2.2592001161370392e-06,
|
| 1031 |
+
"loss": 0.7662,
|
| 1032 |
+
"step": 730
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.7654469106178765,
|
| 1036 |
+
"grad_norm": 0.9904152103445725,
|
| 1037 |
+
"learning_rate": 2.0461379123170284e-06,
|
| 1038 |
+
"loss": 0.7681,
|
| 1039 |
+
"step": 735
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.7774445110977806,
|
| 1043 |
+
"grad_norm": 1.0176103785435524,
|
| 1044 |
+
"learning_rate": 1.8431979423369604e-06,
|
| 1045 |
+
"loss": 0.7606,
|
| 1046 |
+
"step": 740
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.7894421115776846,
|
| 1050 |
+
"grad_norm": 1.0288460382306641,
|
| 1051 |
+
"learning_rate": 1.650469699016116e-06,
|
| 1052 |
+
"loss": 0.7704,
|
| 1053 |
+
"step": 745
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 1.8014397120575885,
|
| 1057 |
+
"grad_norm": 0.9904963634021251,
|
| 1058 |
+
"learning_rate": 1.4680381719888807e-06,
|
| 1059 |
+
"loss": 0.7502,
|
| 1060 |
+
"step": 750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.8134373125374925,
|
| 1064 |
+
"grad_norm": 1.0256818925204132,
|
| 1065 |
+
"learning_rate": 1.2959838102258536e-06,
|
| 1066 |
+
"loss": 0.7713,
|
| 1067 |
+
"step": 755
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 1.8254349130173964,
|
| 1071 |
+
"grad_norm": 0.9849452100623455,
|
| 1072 |
+
"learning_rate": 1.134382486557342e-06,
|
| 1073 |
+
"loss": 0.758,
|
| 1074 |
+
"step": 760
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 1.8374325134973004,
|
| 1078 |
+
"grad_norm": 0.9397595503622964,
|
| 1079 |
+
"learning_rate": 9.833054642148066e-07,
|
| 1080 |
+
"loss": 0.7521,
|
| 1081 |
+
"step": 765
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 1.8494301139772045,
|
| 1085 |
+
"grad_norm": 1.0890895891545742,
|
| 1086 |
+
"learning_rate": 8.428193654051036e-07,
|
| 1087 |
+
"loss": 0.7509,
|
| 1088 |
+
"step": 770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 1.8614277144571085,
|
| 1092 |
+
"grad_norm": 0.9884112088685092,
|
| 1093 |
+
"learning_rate": 7.129861419312822e-07,
|
| 1094 |
+
"loss": 0.7655,
|
| 1095 |
+
"step": 775
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.8734253149370126,
|
| 1099 |
+
"grad_norm": 0.9719952905492264,
|
| 1100 |
+
"learning_rate": 5.938630478729917e-07,
|
| 1101 |
+
"loss": 0.7498,
|
| 1102 |
+
"step": 780
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.8854229154169166,
|
| 1106 |
+
"grad_norm": 1.0012032904437846,
|
| 1107 |
+
"learning_rate": 4.855026143384733e-07,
|
| 1108 |
+
"loss": 0.7536,
|
| 1109 |
+
"step": 785
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 1.8974205158968207,
|
| 1113 |
+
"grad_norm": 0.9926612758167743,
|
| 1114 |
+
"learning_rate": 3.8795262629928996e-07,
|
| 1115 |
+
"loss": 0.7668,
|
| 1116 |
+
"step": 790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.9094181163767248,
|
| 1120 |
+
"grad_norm": 0.9848391831057468,
|
| 1121 |
+
"learning_rate": 3.0125610151804374e-07,
|
| 1122 |
+
"loss": 0.7638,
|
| 1123 |
+
"step": 795
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 1.9214157168566288,
|
| 1127 |
+
"grad_norm": 1.0112841443251817,
|
| 1128 |
+
"learning_rate": 2.2545127157831413e-07,
|
| 1129 |
+
"loss": 0.755,
|
| 1130 |
+
"step": 800
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 1.9334133173365327,
|
| 1134 |
+
"grad_norm": 0.9951541506451825,
|
| 1135 |
+
"learning_rate": 1.605715650252415e-07,
|
| 1136 |
+
"loss": 0.7509,
|
| 1137 |
+
"step": 805
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 1.9454109178164367,
|
| 1141 |
+
"grad_norm": 1.0148344278049142,
|
| 1142 |
+
"learning_rate": 1.0664559262413831e-07,
|
| 1143 |
+
"loss": 0.7576,
|
| 1144 |
+
"step": 810
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 1.9574085182963408,
|
| 1148 |
+
"grad_norm": 1.0628673000773126,
|
| 1149 |
+
"learning_rate": 6.369713474366212e-08,
|
| 1150 |
+
"loss": 0.7668,
|
| 1151 |
+
"step": 815
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.9694061187762446,
|
| 1155 |
+
"grad_norm": 0.9218706861308771,
|
| 1156 |
+
"learning_rate": 3.1745130869123566e-08,
|
| 1157 |
+
"loss": 0.7593,
|
| 1158 |
+
"step": 820
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 1.9814037192561487,
|
| 1162 |
+
"grad_norm": 1.0123478252373648,
|
| 1163 |
+
"learning_rate": 1.08036712505033e-08,
|
| 1164 |
+
"loss": 0.7563,
|
| 1165 |
+
"step": 825
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.9934013197360527,
|
| 1169 |
+
"grad_norm": 1.0032109470714232,
|
| 1170 |
+
"learning_rate": 8.819906889168117e-10,
|
| 1171 |
+
"loss": 0.7589,
|
| 1172 |
+
"step": 830
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.9982003599280143,
|
| 1176 |
+
"step": 832,
|
| 1177 |
+
"total_flos": 619672101191680.0,
|
| 1178 |
+
"train_loss": 0.9412782498850272,
|
| 1179 |
+
"train_runtime": 43950.8497,
|
| 1180 |
+
"train_samples_per_second": 1.214,
|
| 1181 |
+
"train_steps_per_second": 0.019
|
| 1182 |
+
}
|
| 1183 |
+
],
|
| 1184 |
+
"logging_steps": 5,
|
| 1185 |
+
"max_steps": 832,
|
| 1186 |
+
"num_input_tokens_seen": 0,
|
| 1187 |
+
"num_train_epochs": 2,
|
| 1188 |
+
"save_steps": 300,
|
| 1189 |
+
"stateful_callbacks": {
|
| 1190 |
+
"TrainerControl": {
|
| 1191 |
+
"args": {
|
| 1192 |
+
"should_epoch_stop": false,
|
| 1193 |
+
"should_evaluate": false,
|
| 1194 |
+
"should_log": false,
|
| 1195 |
+
"should_save": true,
|
| 1196 |
+
"should_training_stop": true
|
| 1197 |
+
},
|
| 1198 |
+
"attributes": {}
|
| 1199 |
+
}
|
| 1200 |
+
},
|
| 1201 |
+
"total_flos": 619672101191680.0,
|
| 1202 |
+
"train_batch_size": 1,
|
| 1203 |
+
"trial_name": null,
|
| 1204 |
+
"trial_params": null
|
| 1205 |
+
}
|