Upload training info: trainer_state.json
Browse files- trainer_state.json +1153 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1153 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 80190,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.031175957101883027,
|
| 13 |
+
"grad_norm": 18.50511932373047,
|
| 14 |
+
"learning_rate": 4.968824042898117e-05,
|
| 15 |
+
"loss": 0.7295,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.062351914203766054,
|
| 20 |
+
"grad_norm": 13.994952201843262,
|
| 21 |
+
"learning_rate": 4.9376480857962345e-05,
|
| 22 |
+
"loss": 0.4767,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09352787130564909,
|
| 27 |
+
"grad_norm": 18.379146575927734,
|
| 28 |
+
"learning_rate": 4.906472128694351e-05,
|
| 29 |
+
"loss": 0.4072,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12470382840753211,
|
| 34 |
+
"grad_norm": 14.668228149414062,
|
| 35 |
+
"learning_rate": 4.875296171592468e-05,
|
| 36 |
+
"loss": 0.3413,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15587978550941514,
|
| 41 |
+
"grad_norm": 15.569215774536133,
|
| 42 |
+
"learning_rate": 4.844120214490585e-05,
|
| 43 |
+
"loss": 0.3295,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18705574261129818,
|
| 48 |
+
"grad_norm": 9.417423248291016,
|
| 49 |
+
"learning_rate": 4.8129442573887016e-05,
|
| 50 |
+
"loss": 0.3009,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21823169971318118,
|
| 55 |
+
"grad_norm": 14.446584701538086,
|
| 56 |
+
"learning_rate": 4.781768300286819e-05,
|
| 57 |
+
"loss": 0.2848,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24940765681506422,
|
| 62 |
+
"grad_norm": 13.457582473754883,
|
| 63 |
+
"learning_rate": 4.750592343184936e-05,
|
| 64 |
+
"loss": 0.2691,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.28058361391694725,
|
| 69 |
+
"grad_norm": 5.612010955810547,
|
| 70 |
+
"learning_rate": 4.719416386083053e-05,
|
| 71 |
+
"loss": 0.2663,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.3117595710188303,
|
| 76 |
+
"grad_norm": 16.31462860107422,
|
| 77 |
+
"learning_rate": 4.68824042898117e-05,
|
| 78 |
+
"loss": 0.2518,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3429355281207133,
|
| 83 |
+
"grad_norm": 7.700777053833008,
|
| 84 |
+
"learning_rate": 4.657064471879287e-05,
|
| 85 |
+
"loss": 0.2464,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.37411148522259635,
|
| 90 |
+
"grad_norm": 14.337154388427734,
|
| 91 |
+
"learning_rate": 4.6258885147774036e-05,
|
| 92 |
+
"loss": 0.2343,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.4052874423244794,
|
| 97 |
+
"grad_norm": 7.8251118659973145,
|
| 98 |
+
"learning_rate": 4.5947125576755204e-05,
|
| 99 |
+
"loss": 0.2309,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.43646339942636236,
|
| 104 |
+
"grad_norm": 11.534103393554688,
|
| 105 |
+
"learning_rate": 4.563536600573638e-05,
|
| 106 |
+
"loss": 0.232,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.4676393565282454,
|
| 111 |
+
"grad_norm": 5.473570823669434,
|
| 112 |
+
"learning_rate": 4.5323606434717546e-05,
|
| 113 |
+
"loss": 0.2224,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.49881531363012843,
|
| 118 |
+
"grad_norm": 10.417407035827637,
|
| 119 |
+
"learning_rate": 4.501184686369872e-05,
|
| 120 |
+
"loss": 0.2183,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5299912707320115,
|
| 125 |
+
"grad_norm": 8.27934455871582,
|
| 126 |
+
"learning_rate": 4.470008729267989e-05,
|
| 127 |
+
"loss": 0.212,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5611672278338945,
|
| 132 |
+
"grad_norm": 17.182754516601562,
|
| 133 |
+
"learning_rate": 4.438832772166106e-05,
|
| 134 |
+
"loss": 0.2101,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5923431849357775,
|
| 139 |
+
"grad_norm": 7.733582973480225,
|
| 140 |
+
"learning_rate": 4.407656815064223e-05,
|
| 141 |
+
"loss": 0.2002,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6235191420376606,
|
| 146 |
+
"grad_norm": 14.676047325134277,
|
| 147 |
+
"learning_rate": 4.376480857962339e-05,
|
| 148 |
+
"loss": 0.1986,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6546950991395436,
|
| 153 |
+
"grad_norm": 7.192004680633545,
|
| 154 |
+
"learning_rate": 4.345304900860457e-05,
|
| 155 |
+
"loss": 0.1996,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6858710562414266,
|
| 160 |
+
"grad_norm": 8.274066925048828,
|
| 161 |
+
"learning_rate": 4.3141289437585735e-05,
|
| 162 |
+
"loss": 0.1857,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.7170470133433097,
|
| 167 |
+
"grad_norm": 3.5944628715515137,
|
| 168 |
+
"learning_rate": 4.28295298665669e-05,
|
| 169 |
+
"loss": 0.1794,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7482229704451927,
|
| 174 |
+
"grad_norm": 11.68619441986084,
|
| 175 |
+
"learning_rate": 4.251777029554808e-05,
|
| 176 |
+
"loss": 0.182,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7793989275470757,
|
| 181 |
+
"grad_norm": 17.828542709350586,
|
| 182 |
+
"learning_rate": 4.2206010724529245e-05,
|
| 183 |
+
"loss": 0.1824,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.8105748846489588,
|
| 188 |
+
"grad_norm": 6.034427642822266,
|
| 189 |
+
"learning_rate": 4.189425115351042e-05,
|
| 190 |
+
"loss": 0.1794,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8417508417508418,
|
| 195 |
+
"grad_norm": 15.933755874633789,
|
| 196 |
+
"learning_rate": 4.158249158249159e-05,
|
| 197 |
+
"loss": 0.1845,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8729267988527247,
|
| 202 |
+
"grad_norm": 2.4596762657165527,
|
| 203 |
+
"learning_rate": 4.1270732011472755e-05,
|
| 204 |
+
"loss": 0.1745,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.9041027559546078,
|
| 209 |
+
"grad_norm": 14.37038803100586,
|
| 210 |
+
"learning_rate": 4.095897244045392e-05,
|
| 211 |
+
"loss": 0.1777,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9352787130564908,
|
| 216 |
+
"grad_norm": 8.180006980895996,
|
| 217 |
+
"learning_rate": 4.064721286943509e-05,
|
| 218 |
+
"loss": 0.1699,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9664546701583738,
|
| 223 |
+
"grad_norm": 13.317747116088867,
|
| 224 |
+
"learning_rate": 4.0335453298416265e-05,
|
| 225 |
+
"loss": 0.1673,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9976306272602569,
|
| 230 |
+
"grad_norm": 4.7334699630737305,
|
| 231 |
+
"learning_rate": 4.002369372739743e-05,
|
| 232 |
+
"loss": 0.1618,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.02880658436214,
|
| 237 |
+
"grad_norm": 0.9546562433242798,
|
| 238 |
+
"learning_rate": 3.971193415637861e-05,
|
| 239 |
+
"loss": 0.1347,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.059982541464023,
|
| 244 |
+
"grad_norm": 8.00651741027832,
|
| 245 |
+
"learning_rate": 3.9400174585359775e-05,
|
| 246 |
+
"loss": 0.1402,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.091158498565906,
|
| 251 |
+
"grad_norm": 6.071548938751221,
|
| 252 |
+
"learning_rate": 3.908841501434094e-05,
|
| 253 |
+
"loss": 0.1387,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.122334455667789,
|
| 258 |
+
"grad_norm": 7.217519760131836,
|
| 259 |
+
"learning_rate": 3.877665544332211e-05,
|
| 260 |
+
"loss": 0.1401,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.153510412769672,
|
| 265 |
+
"grad_norm": 4.944812774658203,
|
| 266 |
+
"learning_rate": 3.846489587230328e-05,
|
| 267 |
+
"loss": 0.1345,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.184686369871555,
|
| 272 |
+
"grad_norm": 8.488551139831543,
|
| 273 |
+
"learning_rate": 3.815313630128445e-05,
|
| 274 |
+
"loss": 0.1411,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.215862326973438,
|
| 279 |
+
"grad_norm": 13.055041313171387,
|
| 280 |
+
"learning_rate": 3.784137673026562e-05,
|
| 281 |
+
"loss": 0.1396,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2470382840753211,
|
| 286 |
+
"grad_norm": 11.56151294708252,
|
| 287 |
+
"learning_rate": 3.752961715924679e-05,
|
| 288 |
+
"loss": 0.1384,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2782142411772042,
|
| 293 |
+
"grad_norm": 6.223227500915527,
|
| 294 |
+
"learning_rate": 3.7217857588227963e-05,
|
| 295 |
+
"loss": 0.1271,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.3093901982790872,
|
| 300 |
+
"grad_norm": 7.8123459815979,
|
| 301 |
+
"learning_rate": 3.690609801720913e-05,
|
| 302 |
+
"loss": 0.1366,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.3405661553809702,
|
| 307 |
+
"grad_norm": 14.853376388549805,
|
| 308 |
+
"learning_rate": 3.65943384461903e-05,
|
| 309 |
+
"loss": 0.1309,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3717421124828533,
|
| 314 |
+
"grad_norm": 13.60897445678711,
|
| 315 |
+
"learning_rate": 3.628257887517147e-05,
|
| 316 |
+
"loss": 0.1227,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.4029180695847363,
|
| 321 |
+
"grad_norm": 8.778059959411621,
|
| 322 |
+
"learning_rate": 3.5970819304152635e-05,
|
| 323 |
+
"loss": 0.131,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.4340940266866193,
|
| 328 |
+
"grad_norm": 6.795323848724365,
|
| 329 |
+
"learning_rate": 3.565905973313381e-05,
|
| 330 |
+
"loss": 0.1262,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4652699837885024,
|
| 335 |
+
"grad_norm": 9.674951553344727,
|
| 336 |
+
"learning_rate": 3.534730016211498e-05,
|
| 337 |
+
"loss": 0.1296,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4964459408903854,
|
| 342 |
+
"grad_norm": 6.505063533782959,
|
| 343 |
+
"learning_rate": 3.503554059109615e-05,
|
| 344 |
+
"loss": 0.1289,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.5276218979922684,
|
| 349 |
+
"grad_norm": 3.013941764831543,
|
| 350 |
+
"learning_rate": 3.472378102007732e-05,
|
| 351 |
+
"loss": 0.1198,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5587978550941513,
|
| 356 |
+
"grad_norm": 2.129225254058838,
|
| 357 |
+
"learning_rate": 3.441202144905849e-05,
|
| 358 |
+
"loss": 0.1265,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5899738121960345,
|
| 363 |
+
"grad_norm": 16.27604866027832,
|
| 364 |
+
"learning_rate": 3.4100261878039655e-05,
|
| 365 |
+
"loss": 0.1209,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.6211497692979173,
|
| 370 |
+
"grad_norm": 7.551379680633545,
|
| 371 |
+
"learning_rate": 3.378850230702082e-05,
|
| 372 |
+
"loss": 0.1254,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.6523257263998006,
|
| 377 |
+
"grad_norm": 3.263906717300415,
|
| 378 |
+
"learning_rate": 3.3476742736002e-05,
|
| 379 |
+
"loss": 0.1184,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6835016835016834,
|
| 384 |
+
"grad_norm": 3.7056403160095215,
|
| 385 |
+
"learning_rate": 3.3164983164983165e-05,
|
| 386 |
+
"loss": 0.1131,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.7146776406035666,
|
| 391 |
+
"grad_norm": 12.94681453704834,
|
| 392 |
+
"learning_rate": 3.285322359396434e-05,
|
| 393 |
+
"loss": 0.116,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.7458535977054495,
|
| 398 |
+
"grad_norm": 12.127866744995117,
|
| 399 |
+
"learning_rate": 3.254146402294551e-05,
|
| 400 |
+
"loss": 0.118,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7770295548073327,
|
| 405 |
+
"grad_norm": 3.3519415855407715,
|
| 406 |
+
"learning_rate": 3.2229704451926675e-05,
|
| 407 |
+
"loss": 0.1149,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.8082055119092155,
|
| 412 |
+
"grad_norm": 3.702216625213623,
|
| 413 |
+
"learning_rate": 3.191794488090784e-05,
|
| 414 |
+
"loss": 0.118,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.8393814690110988,
|
| 419 |
+
"grad_norm": 2.697953462600708,
|
| 420 |
+
"learning_rate": 3.160618530988901e-05,
|
| 421 |
+
"loss": 0.1126,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8705574261129816,
|
| 426 |
+
"grad_norm": 3.441023826599121,
|
| 427 |
+
"learning_rate": 3.1294425738870185e-05,
|
| 428 |
+
"loss": 0.1142,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.9017333832148648,
|
| 433 |
+
"grad_norm": 5.061633586883545,
|
| 434 |
+
"learning_rate": 3.098266616785135e-05,
|
| 435 |
+
"loss": 0.1033,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.9329093403167477,
|
| 440 |
+
"grad_norm": 3.708242654800415,
|
| 441 |
+
"learning_rate": 3.067090659683252e-05,
|
| 442 |
+
"loss": 0.111,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.964085297418631,
|
| 447 |
+
"grad_norm": 5.391118049621582,
|
| 448 |
+
"learning_rate": 3.0359147025813696e-05,
|
| 449 |
+
"loss": 0.1052,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9952612545205137,
|
| 454 |
+
"grad_norm": 5.222867488861084,
|
| 455 |
+
"learning_rate": 3.0047387454794863e-05,
|
| 456 |
+
"loss": 0.1082,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 2.026437211622397,
|
| 461 |
+
"grad_norm": 4.524135112762451,
|
| 462 |
+
"learning_rate": 2.9735627883776035e-05,
|
| 463 |
+
"loss": 0.0938,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.05761316872428,
|
| 468 |
+
"grad_norm": 6.173551082611084,
|
| 469 |
+
"learning_rate": 2.9423868312757202e-05,
|
| 470 |
+
"loss": 0.0892,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.088789125826163,
|
| 475 |
+
"grad_norm": 11.031685829162598,
|
| 476 |
+
"learning_rate": 2.9112108741738377e-05,
|
| 477 |
+
"loss": 0.0886,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.119965082928046,
|
| 482 |
+
"grad_norm": 10.047669410705566,
|
| 483 |
+
"learning_rate": 2.880034917071954e-05,
|
| 484 |
+
"loss": 0.0871,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.151141040029929,
|
| 489 |
+
"grad_norm": 10.764500617980957,
|
| 490 |
+
"learning_rate": 2.848858959970071e-05,
|
| 491 |
+
"loss": 0.0886,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.182316997131812,
|
| 496 |
+
"grad_norm": 4.433384418487549,
|
| 497 |
+
"learning_rate": 2.8176830028681884e-05,
|
| 498 |
+
"loss": 0.0897,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.213492954233695,
|
| 503 |
+
"grad_norm": 3.753460645675659,
|
| 504 |
+
"learning_rate": 2.786507045766305e-05,
|
| 505 |
+
"loss": 0.0845,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.244668911335578,
|
| 510 |
+
"grad_norm": 13.063867568969727,
|
| 511 |
+
"learning_rate": 2.7553310886644223e-05,
|
| 512 |
+
"loss": 0.0909,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.275844868437461,
|
| 517 |
+
"grad_norm": 0.876360297203064,
|
| 518 |
+
"learning_rate": 2.724155131562539e-05,
|
| 519 |
+
"loss": 0.0821,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.307020825539344,
|
| 524 |
+
"grad_norm": 3.4214694499969482,
|
| 525 |
+
"learning_rate": 2.692979174460656e-05,
|
| 526 |
+
"loss": 0.0886,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.3381967826412273,
|
| 531 |
+
"grad_norm": 1.5247845649719238,
|
| 532 |
+
"learning_rate": 2.6618032173587733e-05,
|
| 533 |
+
"loss": 0.0911,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.36937273974311,
|
| 538 |
+
"grad_norm": 8.257843017578125,
|
| 539 |
+
"learning_rate": 2.6306272602568897e-05,
|
| 540 |
+
"loss": 0.0826,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.400548696844993,
|
| 545 |
+
"grad_norm": 4.504921913146973,
|
| 546 |
+
"learning_rate": 2.5994513031550072e-05,
|
| 547 |
+
"loss": 0.0831,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.431724653946876,
|
| 552 |
+
"grad_norm": 9.82140827178955,
|
| 553 |
+
"learning_rate": 2.568275346053124e-05,
|
| 554 |
+
"loss": 0.0837,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.462900611048759,
|
| 559 |
+
"grad_norm": 6.972949504852295,
|
| 560 |
+
"learning_rate": 2.5370993889512408e-05,
|
| 561 |
+
"loss": 0.0776,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4940765681506423,
|
| 566 |
+
"grad_norm": 1.5516741275787354,
|
| 567 |
+
"learning_rate": 2.505923431849358e-05,
|
| 568 |
+
"loss": 0.0802,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.525252525252525,
|
| 573 |
+
"grad_norm": 1.2083368301391602,
|
| 574 |
+
"learning_rate": 2.474747474747475e-05,
|
| 575 |
+
"loss": 0.0883,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.5564284823544083,
|
| 580 |
+
"grad_norm": 0.3468765318393707,
|
| 581 |
+
"learning_rate": 2.4435715176455918e-05,
|
| 582 |
+
"loss": 0.0787,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.587604439456291,
|
| 587 |
+
"grad_norm": 2.4665677547454834,
|
| 588 |
+
"learning_rate": 2.412395560543709e-05,
|
| 589 |
+
"loss": 0.0741,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.6187803965581744,
|
| 594 |
+
"grad_norm": 6.275604724884033,
|
| 595 |
+
"learning_rate": 2.3812196034418257e-05,
|
| 596 |
+
"loss": 0.0766,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.649956353660057,
|
| 601 |
+
"grad_norm": 3.2683846950531006,
|
| 602 |
+
"learning_rate": 2.3500436463399428e-05,
|
| 603 |
+
"loss": 0.0778,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.6811323107619405,
|
| 608 |
+
"grad_norm": 0.10167425125837326,
|
| 609 |
+
"learning_rate": 2.31886768923806e-05,
|
| 610 |
+
"loss": 0.0779,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.7123082678638233,
|
| 615 |
+
"grad_norm": 4.321603775024414,
|
| 616 |
+
"learning_rate": 2.2876917321361767e-05,
|
| 617 |
+
"loss": 0.0755,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.7434842249657065,
|
| 622 |
+
"grad_norm": 11.752392768859863,
|
| 623 |
+
"learning_rate": 2.2565157750342935e-05,
|
| 624 |
+
"loss": 0.0721,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.7746601820675894,
|
| 629 |
+
"grad_norm": 1.6715621948242188,
|
| 630 |
+
"learning_rate": 2.2253398179324106e-05,
|
| 631 |
+
"loss": 0.0759,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8058361391694726,
|
| 636 |
+
"grad_norm": 1.3633430004119873,
|
| 637 |
+
"learning_rate": 2.1941638608305277e-05,
|
| 638 |
+
"loss": 0.0754,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.8370120962713554,
|
| 643 |
+
"grad_norm": 3.60949969291687,
|
| 644 |
+
"learning_rate": 2.1629879037286445e-05,
|
| 645 |
+
"loss": 0.0808,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.8681880533732387,
|
| 650 |
+
"grad_norm": 11.720767974853516,
|
| 651 |
+
"learning_rate": 2.1318119466267616e-05,
|
| 652 |
+
"loss": 0.0702,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8993640104751215,
|
| 657 |
+
"grad_norm": 10.18453311920166,
|
| 658 |
+
"learning_rate": 2.1006359895248784e-05,
|
| 659 |
+
"loss": 0.0764,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.9305399675770047,
|
| 664 |
+
"grad_norm": 5.267646789550781,
|
| 665 |
+
"learning_rate": 2.0694600324229955e-05,
|
| 666 |
+
"loss": 0.0683,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.9617159246788876,
|
| 671 |
+
"grad_norm": 3.5827486515045166,
|
| 672 |
+
"learning_rate": 2.0382840753211123e-05,
|
| 673 |
+
"loss": 0.0745,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.992891881780771,
|
| 678 |
+
"grad_norm": 3.5235114097595215,
|
| 679 |
+
"learning_rate": 2.0071081182192294e-05,
|
| 680 |
+
"loss": 0.0755,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 3.0240678388826536,
|
| 685 |
+
"grad_norm": 5.0308003425598145,
|
| 686 |
+
"learning_rate": 1.9759321611173465e-05,
|
| 687 |
+
"loss": 0.0591,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 3.055243795984537,
|
| 692 |
+
"grad_norm": 1.5788720846176147,
|
| 693 |
+
"learning_rate": 1.9447562040154636e-05,
|
| 694 |
+
"loss": 0.0573,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0864197530864197,
|
| 699 |
+
"grad_norm": 7.884181022644043,
|
| 700 |
+
"learning_rate": 1.91358024691358e-05,
|
| 701 |
+
"loss": 0.0604,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.117595710188303,
|
| 706 |
+
"grad_norm": 11.146048545837402,
|
| 707 |
+
"learning_rate": 1.8824042898116972e-05,
|
| 708 |
+
"loss": 0.0566,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 3.1487716672901858,
|
| 713 |
+
"grad_norm": 3.091111421585083,
|
| 714 |
+
"learning_rate": 1.8512283327098143e-05,
|
| 715 |
+
"loss": 0.0567,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 3.179947624392069,
|
| 720 |
+
"grad_norm": 1.165995717048645,
|
| 721 |
+
"learning_rate": 1.8200523756079314e-05,
|
| 722 |
+
"loss": 0.059,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 3.211123581493952,
|
| 727 |
+
"grad_norm": 4.651956081390381,
|
| 728 |
+
"learning_rate": 1.7888764185060482e-05,
|
| 729 |
+
"loss": 0.0593,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 3.242299538595835,
|
| 734 |
+
"grad_norm": 4.40238618850708,
|
| 735 |
+
"learning_rate": 1.757700461404165e-05,
|
| 736 |
+
"loss": 0.0543,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 3.273475495697718,
|
| 741 |
+
"grad_norm": 1.9541442394256592,
|
| 742 |
+
"learning_rate": 1.726524504302282e-05,
|
| 743 |
+
"loss": 0.0607,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 3.304651452799601,
|
| 748 |
+
"grad_norm": 6.768977642059326,
|
| 749 |
+
"learning_rate": 1.6953485472003992e-05,
|
| 750 |
+
"loss": 0.0603,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 3.335827409901484,
|
| 755 |
+
"grad_norm": 5.167697906494141,
|
| 756 |
+
"learning_rate": 1.664172590098516e-05,
|
| 757 |
+
"loss": 0.0547,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 3.3670033670033668,
|
| 762 |
+
"grad_norm": 1.6765258312225342,
|
| 763 |
+
"learning_rate": 1.632996632996633e-05,
|
| 764 |
+
"loss": 0.0575,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 3.39817932410525,
|
| 769 |
+
"grad_norm": 1.890818476676941,
|
| 770 |
+
"learning_rate": 1.6018206758947502e-05,
|
| 771 |
+
"loss": 0.0562,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 3.4293552812071333,
|
| 776 |
+
"grad_norm": 3.50783109664917,
|
| 777 |
+
"learning_rate": 1.570644718792867e-05,
|
| 778 |
+
"loss": 0.0527,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 3.460531238309016,
|
| 783 |
+
"grad_norm": 6.326920986175537,
|
| 784 |
+
"learning_rate": 1.5394687616909838e-05,
|
| 785 |
+
"loss": 0.0561,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 3.491707195410899,
|
| 790 |
+
"grad_norm": 0.5801683664321899,
|
| 791 |
+
"learning_rate": 1.508292804589101e-05,
|
| 792 |
+
"loss": 0.0522,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 3.522883152512782,
|
| 797 |
+
"grad_norm": 0.1254473626613617,
|
| 798 |
+
"learning_rate": 1.477116847487218e-05,
|
| 799 |
+
"loss": 0.0553,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 3.5540591096146654,
|
| 804 |
+
"grad_norm": 11.128674507141113,
|
| 805 |
+
"learning_rate": 1.445940890385335e-05,
|
| 806 |
+
"loss": 0.0545,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 3.5852350667165482,
|
| 811 |
+
"grad_norm": 5.950380802154541,
|
| 812 |
+
"learning_rate": 1.414764933283452e-05,
|
| 813 |
+
"loss": 0.0513,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 3.616411023818431,
|
| 818 |
+
"grad_norm": 7.508421897888184,
|
| 819 |
+
"learning_rate": 1.3835889761815687e-05,
|
| 820 |
+
"loss": 0.0473,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 3.6475869809203143,
|
| 825 |
+
"grad_norm": 0.1135072112083435,
|
| 826 |
+
"learning_rate": 1.3524130190796858e-05,
|
| 827 |
+
"loss": 0.0496,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 3.678762938022197,
|
| 832 |
+
"grad_norm": 5.939940929412842,
|
| 833 |
+
"learning_rate": 1.3212370619778028e-05,
|
| 834 |
+
"loss": 0.048,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 3.7099388951240804,
|
| 839 |
+
"grad_norm": 0.04740321636199951,
|
| 840 |
+
"learning_rate": 1.2900611048759197e-05,
|
| 841 |
+
"loss": 0.0505,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 3.741114852225963,
|
| 846 |
+
"grad_norm": 6.120915412902832,
|
| 847 |
+
"learning_rate": 1.2588851477740369e-05,
|
| 848 |
+
"loss": 0.05,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 3.7722908093278464,
|
| 853 |
+
"grad_norm": 2.7435576915740967,
|
| 854 |
+
"learning_rate": 1.2277091906721536e-05,
|
| 855 |
+
"loss": 0.0512,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 3.8034667664297293,
|
| 860 |
+
"grad_norm": 5.130537986755371,
|
| 861 |
+
"learning_rate": 1.1965332335702708e-05,
|
| 862 |
+
"loss": 0.0487,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 3.8346427235316125,
|
| 867 |
+
"grad_norm": 0.5921691656112671,
|
| 868 |
+
"learning_rate": 1.1653572764683875e-05,
|
| 869 |
+
"loss": 0.049,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 3.8658186806334953,
|
| 874 |
+
"grad_norm": 0.21706895530223846,
|
| 875 |
+
"learning_rate": 1.1341813193665047e-05,
|
| 876 |
+
"loss": 0.0503,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 3.8969946377353786,
|
| 881 |
+
"grad_norm": 6.188902854919434,
|
| 882 |
+
"learning_rate": 1.1030053622646216e-05,
|
| 883 |
+
"loss": 0.0454,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 3.9281705948372614,
|
| 888 |
+
"grad_norm": 3.0390682220458984,
|
| 889 |
+
"learning_rate": 1.0718294051627386e-05,
|
| 890 |
+
"loss": 0.0449,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 3.9593465519391446,
|
| 895 |
+
"grad_norm": 2.7789669036865234,
|
| 896 |
+
"learning_rate": 1.0406534480608555e-05,
|
| 897 |
+
"loss": 0.0483,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 3.9905225090410275,
|
| 902 |
+
"grad_norm": 2.688272476196289,
|
| 903 |
+
"learning_rate": 1.0094774909589724e-05,
|
| 904 |
+
"loss": 0.0462,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 4.021698466142911,
|
| 909 |
+
"grad_norm": 4.848268032073975,
|
| 910 |
+
"learning_rate": 9.783015338570894e-06,
|
| 911 |
+
"loss": 0.0425,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 4.052874423244794,
|
| 916 |
+
"grad_norm": 3.2192070484161377,
|
| 917 |
+
"learning_rate": 9.471255767552065e-06,
|
| 918 |
+
"loss": 0.0403,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 4.084050380346676,
|
| 923 |
+
"grad_norm": 4.671278953552246,
|
| 924 |
+
"learning_rate": 9.159496196533233e-06,
|
| 925 |
+
"loss": 0.0376,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 4.11522633744856,
|
| 930 |
+
"grad_norm": 4.251095294952393,
|
| 931 |
+
"learning_rate": 8.847736625514404e-06,
|
| 932 |
+
"loss": 0.0369,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 4.146402294550443,
|
| 937 |
+
"grad_norm": 0.2051253616809845,
|
| 938 |
+
"learning_rate": 8.535977054495574e-06,
|
| 939 |
+
"loss": 0.0352,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 4.177578251652326,
|
| 944 |
+
"grad_norm": 2.640092134475708,
|
| 945 |
+
"learning_rate": 8.224217483476743e-06,
|
| 946 |
+
"loss": 0.0375,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 4.2087542087542085,
|
| 951 |
+
"grad_norm": 0.7651222944259644,
|
| 952 |
+
"learning_rate": 7.912457912457913e-06,
|
| 953 |
+
"loss": 0.039,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 4.239930165856092,
|
| 958 |
+
"grad_norm": 5.041400909423828,
|
| 959 |
+
"learning_rate": 7.600698341439083e-06,
|
| 960 |
+
"loss": 0.0376,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 4.271106122957975,
|
| 965 |
+
"grad_norm": 3.154355764389038,
|
| 966 |
+
"learning_rate": 7.288938770420252e-06,
|
| 967 |
+
"loss": 0.0382,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 4.302282080059858,
|
| 972 |
+
"grad_norm": 4.174047946929932,
|
| 973 |
+
"learning_rate": 6.977179199401422e-06,
|
| 974 |
+
"loss": 0.0334,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 4.333458037161741,
|
| 979 |
+
"grad_norm": 1.5814062356948853,
|
| 980 |
+
"learning_rate": 6.665419628382592e-06,
|
| 981 |
+
"loss": 0.0352,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 4.364633994263624,
|
| 986 |
+
"grad_norm": 0.0495087206363678,
|
| 987 |
+
"learning_rate": 6.353660057363761e-06,
|
| 988 |
+
"loss": 0.0352,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 4.395809951365507,
|
| 993 |
+
"grad_norm": 1.5100332498550415,
|
| 994 |
+
"learning_rate": 6.041900486344931e-06,
|
| 995 |
+
"loss": 0.0315,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 4.42698590846739,
|
| 1000 |
+
"grad_norm": 0.04012216255068779,
|
| 1001 |
+
"learning_rate": 5.730140915326101e-06,
|
| 1002 |
+
"loss": 0.0315,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 4.458161865569273,
|
| 1007 |
+
"grad_norm": 5.63645601272583,
|
| 1008 |
+
"learning_rate": 5.41838134430727e-06,
|
| 1009 |
+
"loss": 0.0375,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 4.489337822671156,
|
| 1014 |
+
"grad_norm": 2.0898067951202393,
|
| 1015 |
+
"learning_rate": 5.10662177328844e-06,
|
| 1016 |
+
"loss": 0.0363,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 4.520513779773039,
|
| 1021 |
+
"grad_norm": 6.37893009185791,
|
| 1022 |
+
"learning_rate": 4.79486220226961e-06,
|
| 1023 |
+
"loss": 0.0354,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 4.551689736874922,
|
| 1028 |
+
"grad_norm": 1.1932406425476074,
|
| 1029 |
+
"learning_rate": 4.48310263125078e-06,
|
| 1030 |
+
"loss": 0.0351,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 4.582865693976805,
|
| 1035 |
+
"grad_norm": 3.3668556213378906,
|
| 1036 |
+
"learning_rate": 4.171343060231949e-06,
|
| 1037 |
+
"loss": 0.033,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 4.614041651078688,
|
| 1042 |
+
"grad_norm": 5.648000717163086,
|
| 1043 |
+
"learning_rate": 3.859583489213119e-06,
|
| 1044 |
+
"loss": 0.0311,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 4.645217608180571,
|
| 1049 |
+
"grad_norm": 0.38028573989868164,
|
| 1050 |
+
"learning_rate": 3.547823918194289e-06,
|
| 1051 |
+
"loss": 0.0308,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 4.676393565282455,
|
| 1056 |
+
"grad_norm": 3.3189568519592285,
|
| 1057 |
+
"learning_rate": 3.2360643471754584e-06,
|
| 1058 |
+
"loss": 0.0335,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 4.707569522384337,
|
| 1063 |
+
"grad_norm": 2.5542211532592773,
|
| 1064 |
+
"learning_rate": 2.9243047761566283e-06,
|
| 1065 |
+
"loss": 0.0338,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 4.73874547948622,
|
| 1070 |
+
"grad_norm": 1.970572590827942,
|
| 1071 |
+
"learning_rate": 2.612545205137798e-06,
|
| 1072 |
+
"loss": 0.0322,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 4.7699214365881035,
|
| 1077 |
+
"grad_norm": 5.368270397186279,
|
| 1078 |
+
"learning_rate": 2.3007856341189673e-06,
|
| 1079 |
+
"loss": 0.031,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 4.801097393689986,
|
| 1084 |
+
"grad_norm": 3.2779486179351807,
|
| 1085 |
+
"learning_rate": 1.9890260631001372e-06,
|
| 1086 |
+
"loss": 0.0368,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 4.832273350791869,
|
| 1091 |
+
"grad_norm": 0.021892394870519638,
|
| 1092 |
+
"learning_rate": 1.677266492081307e-06,
|
| 1093 |
+
"loss": 0.0308,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 4.863449307893752,
|
| 1098 |
+
"grad_norm": 3.866298198699951,
|
| 1099 |
+
"learning_rate": 1.3655069210624766e-06,
|
| 1100 |
+
"loss": 0.0297,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 4.894625264995636,
|
| 1105 |
+
"grad_norm": 2.35339617729187,
|
| 1106 |
+
"learning_rate": 1.0537473500436463e-06,
|
| 1107 |
+
"loss": 0.0321,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 4.925801222097518,
|
| 1112 |
+
"grad_norm": 0.9487350583076477,
|
| 1113 |
+
"learning_rate": 7.41987779024816e-07,
|
| 1114 |
+
"loss": 0.0309,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 4.956977179199401,
|
| 1119 |
+
"grad_norm": 1.868185043334961,
|
| 1120 |
+
"learning_rate": 4.3022820800598584e-07,
|
| 1121 |
+
"loss": 0.0336,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 4.9881531363012845,
|
| 1126 |
+
"grad_norm": 0.15224005281925201,
|
| 1127 |
+
"learning_rate": 1.1846863698715552e-07,
|
| 1128 |
+
"loss": 0.0284,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
}
|
| 1131 |
+
],
|
| 1132 |
+
"logging_steps": 500,
|
| 1133 |
+
"max_steps": 80190,
|
| 1134 |
+
"num_input_tokens_seen": 0,
|
| 1135 |
+
"num_train_epochs": 5,
|
| 1136 |
+
"save_steps": 500,
|
| 1137 |
+
"stateful_callbacks": {
|
| 1138 |
+
"TrainerControl": {
|
| 1139 |
+
"args": {
|
| 1140 |
+
"should_epoch_stop": false,
|
| 1141 |
+
"should_evaluate": false,
|
| 1142 |
+
"should_log": false,
|
| 1143 |
+
"should_save": true,
|
| 1144 |
+
"should_training_stop": true
|
| 1145 |
+
},
|
| 1146 |
+
"attributes": {}
|
| 1147 |
+
}
|
| 1148 |
+
},
|
| 1149 |
+
"total_flos": 4.615573731242858e+17,
|
| 1150 |
+
"train_batch_size": 32,
|
| 1151 |
+
"trial_name": null,
|
| 1152 |
+
"trial_params": null
|
| 1153 |
+
}
|