Update model
Browse files- trainer_state.json +1149 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1149 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 124,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008064516129032258,
|
| 13 |
+
"grad_norm": 82.1250991821289,
|
| 14 |
+
"learning_rate": 4.999197688241076e-07,
|
| 15 |
+
"loss": 3.3188,
|
| 16 |
+
"mean_token_accuracy": 0.4619899392127991,
|
| 17 |
+
"num_tokens": 3582.0,
|
| 18 |
+
"step": 1
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.016129032258064516,
|
| 22 |
+
"grad_norm": 85.84162139892578,
|
| 23 |
+
"learning_rate": 4.996791267927632e-07,
|
| 24 |
+
"loss": 3.1019,
|
| 25 |
+
"mean_token_accuracy": 0.48747536540031433,
|
| 26 |
+
"num_tokens": 7139.0,
|
| 27 |
+
"step": 2
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.024193548387096774,
|
| 31 |
+
"grad_norm": 77.11351776123047,
|
| 32 |
+
"learning_rate": 4.992782283619118e-07,
|
| 33 |
+
"loss": 3.0994,
|
| 34 |
+
"mean_token_accuracy": 0.46798279881477356,
|
| 35 |
+
"num_tokens": 11094.0,
|
| 36 |
+
"step": 3
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.03225806451612903,
|
| 40 |
+
"grad_norm": 74.11935424804688,
|
| 41 |
+
"learning_rate": 4.987173308479737e-07,
|
| 42 |
+
"loss": 3.0893,
|
| 43 |
+
"mean_token_accuracy": 0.4673512279987335,
|
| 44 |
+
"num_tokens": 14896.0,
|
| 45 |
+
"step": 4
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.04032258064516129,
|
| 49 |
+
"grad_norm": 58.4622917175293,
|
| 50 |
+
"learning_rate": 4.979967942626857e-07,
|
| 51 |
+
"loss": 2.9759,
|
| 52 |
+
"mean_token_accuracy": 0.4749932289123535,
|
| 53 |
+
"num_tokens": 18599.0,
|
| 54 |
+
"step": 5
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.04838709677419355,
|
| 58 |
+
"grad_norm": 71.54608154296875,
|
| 59 |
+
"learning_rate": 4.971170810820278e-07,
|
| 60 |
+
"loss": 2.8627,
|
| 61 |
+
"mean_token_accuracy": 0.489130437374115,
|
| 62 |
+
"num_tokens": 22099.0,
|
| 63 |
+
"step": 6
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.056451612903225805,
|
| 67 |
+
"grad_norm": 67.31790924072266,
|
| 68 |
+
"learning_rate": 4.960787559493836e-07,
|
| 69 |
+
"loss": 2.9878,
|
| 70 |
+
"mean_token_accuracy": 0.49747729301452637,
|
| 71 |
+
"num_tokens": 25076.0,
|
| 72 |
+
"step": 7
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06451612903225806,
|
| 76 |
+
"grad_norm": 51.87714767456055,
|
| 77 |
+
"learning_rate": 4.948824853131236e-07,
|
| 78 |
+
"loss": 2.8431,
|
| 79 |
+
"mean_token_accuracy": 0.4804079532623291,
|
| 80 |
+
"num_tokens": 28806.0,
|
| 81 |
+
"step": 8
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.07258064516129033,
|
| 85 |
+
"grad_norm": 50.492496490478516,
|
| 86 |
+
"learning_rate": 4.935290369988468e-07,
|
| 87 |
+
"loss": 2.7265,
|
| 88 |
+
"mean_token_accuracy": 0.4897351861000061,
|
| 89 |
+
"num_tokens": 32171.0,
|
| 90 |
+
"step": 9
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.08064516129032258,
|
| 94 |
+
"grad_norm": 50.19187927246094,
|
| 95 |
+
"learning_rate": 4.920192797165511e-07,
|
| 96 |
+
"loss": 2.5757,
|
| 97 |
+
"mean_token_accuracy": 0.5013927817344666,
|
| 98 |
+
"num_tokens": 35765.0,
|
| 99 |
+
"step": 10
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.08870967741935484,
|
| 103 |
+
"grad_norm": 53.92109298706055,
|
| 104 |
+
"learning_rate": 4.903541825030531e-07,
|
| 105 |
+
"loss": 2.5328,
|
| 106 |
+
"mean_token_accuracy": 0.5096573233604431,
|
| 107 |
+
"num_tokens": 38979.0,
|
| 108 |
+
"step": 11
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.0967741935483871,
|
| 112 |
+
"grad_norm": 131.06643676757812,
|
| 113 |
+
"learning_rate": 4.885348141000122e-07,
|
| 114 |
+
"loss": 2.4167,
|
| 115 |
+
"mean_token_accuracy": 0.507402777671814,
|
| 116 |
+
"num_tokens": 42968.0,
|
| 117 |
+
"step": 12
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.10483870967741936,
|
| 121 |
+
"grad_norm": 42.45437240600586,
|
| 122 |
+
"learning_rate": 4.865623422679592e-07,
|
| 123 |
+
"loss": 2.3694,
|
| 124 |
+
"mean_token_accuracy": 0.5128635168075562,
|
| 125 |
+
"num_tokens": 46548.0,
|
| 126 |
+
"step": 13
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.11290322580645161,
|
| 130 |
+
"grad_norm": 40.65642547607422,
|
| 131 |
+
"learning_rate": 4.844380330367701e-07,
|
| 132 |
+
"loss": 2.2659,
|
| 133 |
+
"mean_token_accuracy": 0.5277366042137146,
|
| 134 |
+
"num_tokens": 49923.0,
|
| 135 |
+
"step": 14
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.12096774193548387,
|
| 139 |
+
"grad_norm": 40.563663482666016,
|
| 140 |
+
"learning_rate": 4.821632498930656e-07,
|
| 141 |
+
"loss": 2.325,
|
| 142 |
+
"mean_token_accuracy": 0.5243242979049683,
|
| 143 |
+
"num_tokens": 53257.0,
|
| 144 |
+
"step": 15
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.12903225806451613,
|
| 148 |
+
"grad_norm": 29.71602439880371,
|
| 149 |
+
"learning_rate": 4.797394529050577e-07,
|
| 150 |
+
"loss": 2.2635,
|
| 151 |
+
"mean_token_accuracy": 0.5110940933227539,
|
| 152 |
+
"num_tokens": 57182.0,
|
| 153 |
+
"step": 16
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.13709677419354838,
|
| 157 |
+
"grad_norm": 36.85178756713867,
|
| 158 |
+
"learning_rate": 4.771681977854062e-07,
|
| 159 |
+
"loss": 2.2947,
|
| 160 |
+
"mean_token_accuracy": 0.49719932675361633,
|
| 161 |
+
"num_tokens": 60221.0,
|
| 162 |
+
"step": 17
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.14516129032258066,
|
| 166 |
+
"grad_norm": 40.669857025146484,
|
| 167 |
+
"learning_rate": 4.744511348926854e-07,
|
| 168 |
+
"loss": 2.1618,
|
| 169 |
+
"mean_token_accuracy": 0.5076530575752258,
|
| 170 |
+
"num_tokens": 63753.0,
|
| 171 |
+
"step": 18
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1532258064516129,
|
| 175 |
+
"grad_norm": 31.07614517211914,
|
| 176 |
+
"learning_rate": 4.7159000817210204e-07,
|
| 177 |
+
"loss": 2.2338,
|
| 178 |
+
"mean_token_accuracy": 0.5,
|
| 179 |
+
"num_tokens": 67491.0,
|
| 180 |
+
"step": 19
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.16129032258064516,
|
| 184 |
+
"grad_norm": 39.705753326416016,
|
| 185 |
+
"learning_rate": 4.685866540361455e-07,
|
| 186 |
+
"loss": 2.1938,
|
| 187 |
+
"mean_token_accuracy": 0.5107913613319397,
|
| 188 |
+
"num_tokens": 70831.0,
|
| 189 |
+
"step": 20
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.1693548387096774,
|
| 193 |
+
"grad_norm": 40.17426681518555,
|
| 194 |
+
"learning_rate": 4.654430001858874e-07,
|
| 195 |
+
"loss": 2.1139,
|
| 196 |
+
"mean_token_accuracy": 0.5218411087989807,
|
| 197 |
+
"num_tokens": 74246.0,
|
| 198 |
+
"step": 21
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.1774193548387097,
|
| 202 |
+
"grad_norm": 34.13058853149414,
|
| 203 |
+
"learning_rate": 4.6216106437368775e-07,
|
| 204 |
+
"loss": 2.0303,
|
| 205 |
+
"mean_token_accuracy": 0.5338891744613647,
|
| 206 |
+
"num_tokens": 77968.0,
|
| 207 |
+
"step": 22
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.18548387096774194,
|
| 211 |
+
"grad_norm": 25.600557327270508,
|
| 212 |
+
"learning_rate": 4.5874295310810185e-07,
|
| 213 |
+
"loss": 2.0356,
|
| 214 |
+
"mean_token_accuracy": 0.515510618686676,
|
| 215 |
+
"num_tokens": 80841.0,
|
| 216 |
+
"step": 23
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 0.1935483870967742,
|
| 220 |
+
"grad_norm": 14.962992668151855,
|
| 221 |
+
"learning_rate": 4.551908603018191e-07,
|
| 222 |
+
"loss": 1.9918,
|
| 223 |
+
"mean_token_accuracy": 0.5420023202896118,
|
| 224 |
+
"num_tokens": 84321.0,
|
| 225 |
+
"step": 24
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.20161290322580644,
|
| 229 |
+
"grad_norm": 10.336100578308105,
|
| 230 |
+
"learning_rate": 4.5150706586350127e-07,
|
| 231 |
+
"loss": 2.0338,
|
| 232 |
+
"mean_token_accuracy": 0.5391054153442383,
|
| 233 |
+
"num_tokens": 88148.0,
|
| 234 |
+
"step": 25
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.20967741935483872,
|
| 238 |
+
"grad_norm": 11.232644081115723,
|
| 239 |
+
"learning_rate": 4.476939342344246e-07,
|
| 240 |
+
"loss": 2.0344,
|
| 241 |
+
"mean_token_accuracy": 0.5330302119255066,
|
| 242 |
+
"num_tokens": 91891.0,
|
| 243 |
+
"step": 26
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.21774193548387097,
|
| 247 |
+
"grad_norm": 8.730783462524414,
|
| 248 |
+
"learning_rate": 4.437539128708647e-07,
|
| 249 |
+
"loss": 2.0291,
|
| 250 |
+
"mean_token_accuracy": 0.5340464115142822,
|
| 251 |
+
"num_tokens": 95728.0,
|
| 252 |
+
"step": 27
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.22580645161290322,
|
| 256 |
+
"grad_norm": 9.901666641235352,
|
| 257 |
+
"learning_rate": 4.396895306731977e-07,
|
| 258 |
+
"loss": 1.9593,
|
| 259 |
+
"mean_token_accuracy": 0.5452925562858582,
|
| 260 |
+
"num_tokens": 99099.0,
|
| 261 |
+
"step": 28
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.23387096774193547,
|
| 265 |
+
"grad_norm": 8.69468879699707,
|
| 266 |
+
"learning_rate": 4.355033963627277e-07,
|
| 267 |
+
"loss": 1.9023,
|
| 268 |
+
"mean_token_accuracy": 0.5690703988075256,
|
| 269 |
+
"num_tokens": 102556.0,
|
| 270 |
+
"step": 29
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 0.24193548387096775,
|
| 274 |
+
"grad_norm": 8.984487533569336,
|
| 275 |
+
"learning_rate": 4.3119819680727996e-07,
|
| 276 |
+
"loss": 1.938,
|
| 277 |
+
"mean_token_accuracy": 0.5465425252914429,
|
| 278 |
+
"num_tokens": 106320.0,
|
| 279 |
+
"step": 30
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.25,
|
| 283 |
+
"grad_norm": 9.061673164367676,
|
| 284 |
+
"learning_rate": 4.2677669529663686e-07,
|
| 285 |
+
"loss": 1.9321,
|
| 286 |
+
"mean_token_accuracy": 0.5477131605148315,
|
| 287 |
+
"num_tokens": 109866.0,
|
| 288 |
+
"step": 31
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.25806451612903225,
|
| 292 |
+
"grad_norm": 8.572089195251465,
|
| 293 |
+
"learning_rate": 4.2224172976892166e-07,
|
| 294 |
+
"loss": 1.9352,
|
| 295 |
+
"mean_token_accuracy": 0.5485040545463562,
|
| 296 |
+
"num_tokens": 113179.0,
|
| 297 |
+
"step": 32
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.2661290322580645,
|
| 301 |
+
"grad_norm": 8.716224670410156,
|
| 302 |
+
"learning_rate": 4.175962109890696e-07,
|
| 303 |
+
"loss": 1.8432,
|
| 304 |
+
"mean_token_accuracy": 0.550000011920929,
|
| 305 |
+
"num_tokens": 116683.0,
|
| 306 |
+
"step": 33
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.27419354838709675,
|
| 310 |
+
"grad_norm": 11.871490478515625,
|
| 311 |
+
"learning_rate": 4.128431206805556e-07,
|
| 312 |
+
"loss": 1.8008,
|
| 313 |
+
"mean_token_accuracy": 0.5759973526000977,
|
| 314 |
+
"num_tokens": 119720.0,
|
| 315 |
+
"step": 34
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.28225806451612906,
|
| 319 |
+
"grad_norm": 9.3665132522583,
|
| 320 |
+
"learning_rate": 4.0798550961157595e-07,
|
| 321 |
+
"loss": 1.8901,
|
| 322 |
+
"mean_token_accuracy": 0.552654504776001,
|
| 323 |
+
"num_tokens": 123171.0,
|
| 324 |
+
"step": 35
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.2903225806451613,
|
| 328 |
+
"grad_norm": 9.704719543457031,
|
| 329 |
+
"learning_rate": 4.030264956369157e-07,
|
| 330 |
+
"loss": 1.8121,
|
| 331 |
+
"mean_token_accuracy": 0.5599347352981567,
|
| 332 |
+
"num_tokens": 126854.0,
|
| 333 |
+
"step": 36
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.29838709677419356,
|
| 337 |
+
"grad_norm": 10.195917129516602,
|
| 338 |
+
"learning_rate": 3.9796926169675424e-07,
|
| 339 |
+
"loss": 1.885,
|
| 340 |
+
"mean_token_accuracy": 0.5533230304718018,
|
| 341 |
+
"num_tokens": 130740.0,
|
| 342 |
+
"step": 37
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3064516129032258,
|
| 346 |
+
"grad_norm": 9.518072128295898,
|
| 347 |
+
"learning_rate": 3.9281705377369805e-07,
|
| 348 |
+
"loss": 1.9059,
|
| 349 |
+
"mean_token_accuracy": 0.5531600713729858,
|
| 350 |
+
"num_tokens": 134130.0,
|
| 351 |
+
"step": 38
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.31451612903225806,
|
| 355 |
+
"grad_norm": 7.841033458709717,
|
| 356 |
+
"learning_rate": 3.875731788093478e-07,
|
| 357 |
+
"loss": 1.8547,
|
| 358 |
+
"mean_token_accuracy": 0.5586956739425659,
|
| 359 |
+
"num_tokens": 137814.0,
|
| 360 |
+
"step": 39
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.3225806451612903,
|
| 364 |
+
"grad_norm": 10.610803604125977,
|
| 365 |
+
"learning_rate": 3.822410025817406e-07,
|
| 366 |
+
"loss": 1.8147,
|
| 367 |
+
"mean_token_accuracy": 0.5637563467025757,
|
| 368 |
+
"num_tokens": 141151.0,
|
| 369 |
+
"step": 40
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.33064516129032256,
|
| 373 |
+
"grad_norm": 10.157751083374023,
|
| 374 |
+
"learning_rate": 3.768239475450268e-07,
|
| 375 |
+
"loss": 1.8701,
|
| 376 |
+
"mean_token_accuracy": 0.549227237701416,
|
| 377 |
+
"num_tokens": 144649.0,
|
| 378 |
+
"step": 41
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.3387096774193548,
|
| 382 |
+
"grad_norm": 8.627851486206055,
|
| 383 |
+
"learning_rate": 3.713254906327703e-07,
|
| 384 |
+
"loss": 1.8508,
|
| 385 |
+
"mean_token_accuracy": 0.557908833026886,
|
| 386 |
+
"num_tokens": 148383.0,
|
| 387 |
+
"step": 42
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.3467741935483871,
|
| 391 |
+
"grad_norm": 7.615920543670654,
|
| 392 |
+
"learning_rate": 3.657491610262802e-07,
|
| 393 |
+
"loss": 1.8584,
|
| 394 |
+
"mean_token_accuracy": 0.5659451484680176,
|
| 395 |
+
"num_tokens": 151852.0,
|
| 396 |
+
"step": 43
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 0.3548387096774194,
|
| 400 |
+
"grad_norm": 8.200079917907715,
|
| 401 |
+
"learning_rate": 3.6009853788940856e-07,
|
| 402 |
+
"loss": 1.7812,
|
| 403 |
+
"mean_token_accuracy": 0.5586913228034973,
|
| 404 |
+
"num_tokens": 155860.0,
|
| 405 |
+
"step": 44
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.3629032258064516,
|
| 409 |
+
"grad_norm": 7.837497234344482,
|
| 410 |
+
"learning_rate": 3.543772480712658e-07,
|
| 411 |
+
"loss": 1.7635,
|
| 412 |
+
"mean_token_accuracy": 0.5686478018760681,
|
| 413 |
+
"num_tokens": 159717.0,
|
| 414 |
+
"step": 45
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.3709677419354839,
|
| 418 |
+
"grad_norm": 8.181253433227539,
|
| 419 |
+
"learning_rate": 3.4858896377832965e-07,
|
| 420 |
+
"loss": 1.7004,
|
| 421 |
+
"mean_token_accuracy": 0.5794797539710999,
|
| 422 |
+
"num_tokens": 163181.0,
|
| 423 |
+
"step": 46
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3790322580645161,
|
| 427 |
+
"grad_norm": 6.781255722045898,
|
| 428 |
+
"learning_rate": 3.42737400217442e-07,
|
| 429 |
+
"loss": 1.8779,
|
| 430 |
+
"mean_token_accuracy": 0.5520778298377991,
|
| 431 |
+
"num_tokens": 166987.0,
|
| 432 |
+
"step": 47
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.3870967741935484,
|
| 436 |
+
"grad_norm": 7.666635036468506,
|
| 437 |
+
"learning_rate": 3.36826313211205e-07,
|
| 438 |
+
"loss": 1.8343,
|
| 439 |
+
"mean_token_accuracy": 0.5685471296310425,
|
| 440 |
+
"num_tokens": 170164.0,
|
| 441 |
+
"step": 48
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.3951612903225806,
|
| 445 |
+
"grad_norm": 8.650765419006348,
|
| 446 |
+
"learning_rate": 3.308594967873095e-07,
|
| 447 |
+
"loss": 1.7022,
|
| 448 |
+
"mean_token_accuracy": 0.5841107368469238,
|
| 449 |
+
"num_tokens": 173491.0,
|
| 450 |
+
"step": 49
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4032258064516129,
|
| 454 |
+
"grad_norm": 7.792832851409912,
|
| 455 |
+
"learning_rate": 3.2484078074333956e-07,
|
| 456 |
+
"loss": 1.7076,
|
| 457 |
+
"mean_token_accuracy": 0.5851721167564392,
|
| 458 |
+
"num_tokens": 176894.0,
|
| 459 |
+
"step": 50
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.4112903225806452,
|
| 463 |
+
"grad_norm": 9.63578987121582,
|
| 464 |
+
"learning_rate": 3.1877402818861946e-07,
|
| 465 |
+
"loss": 1.7369,
|
| 466 |
+
"mean_token_accuracy": 0.576729416847229,
|
| 467 |
+
"num_tokens": 180671.0,
|
| 468 |
+
"step": 51
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 0.41935483870967744,
|
| 472 |
+
"grad_norm": 7.068120002746582,
|
| 473 |
+
"learning_rate": 3.126631330646801e-07,
|
| 474 |
+
"loss": 1.7723,
|
| 475 |
+
"mean_token_accuracy": 0.5763046741485596,
|
| 476 |
+
"num_tokens": 184220.0,
|
| 477 |
+
"step": 52
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.4274193548387097,
|
| 481 |
+
"grad_norm": 7.844783782958984,
|
| 482 |
+
"learning_rate": 3.065120176459337e-07,
|
| 483 |
+
"loss": 1.7888,
|
| 484 |
+
"mean_token_accuracy": 0.5663133859634399,
|
| 485 |
+
"num_tokens": 187715.0,
|
| 486 |
+
"step": 53
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.43548387096774194,
|
| 490 |
+
"grad_norm": 6.963573932647705,
|
| 491 |
+
"learning_rate": 3.00324630022165e-07,
|
| 492 |
+
"loss": 1.6962,
|
| 493 |
+
"mean_token_accuracy": 0.5830934047698975,
|
| 494 |
+
"num_tokens": 191540.0,
|
| 495 |
+
"step": 54
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.4435483870967742,
|
| 499 |
+
"grad_norm": 7.507209777832031,
|
| 500 |
+
"learning_rate": 2.9410494156445216e-07,
|
| 501 |
+
"loss": 1.7422,
|
| 502 |
+
"mean_token_accuracy": 0.5805253982543945,
|
| 503 |
+
"num_tokens": 195046.0,
|
| 504 |
+
"step": 55
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.45161290322580644,
|
| 508 |
+
"grad_norm": 7.959713459014893,
|
| 509 |
+
"learning_rate": 2.8785694437614416e-07,
|
| 510 |
+
"loss": 1.7368,
|
| 511 |
+
"mean_token_accuracy": 0.5760632753372192,
|
| 512 |
+
"num_tokens": 198718.0,
|
| 513 |
+
"step": 56
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.4596774193548387,
|
| 517 |
+
"grad_norm": 7.953831195831299,
|
| 518 |
+
"learning_rate": 2.8158464873053234e-07,
|
| 519 |
+
"loss": 1.7707,
|
| 520 |
+
"mean_token_accuracy": 0.5705274939537048,
|
| 521 |
+
"num_tokens": 202210.0,
|
| 522 |
+
"step": 57
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.46774193548387094,
|
| 526 |
+
"grad_norm": 8.072004318237305,
|
| 527 |
+
"learning_rate": 2.7529208049685804e-07,
|
| 528 |
+
"loss": 1.7197,
|
| 529 |
+
"mean_token_accuracy": 0.5858024954795837,
|
| 530 |
+
"num_tokens": 205454.0,
|
| 531 |
+
"step": 58
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.47580645161290325,
|
| 535 |
+
"grad_norm": 8.691699028015137,
|
| 536 |
+
"learning_rate": 2.6898327855631154e-07,
|
| 537 |
+
"loss": 1.807,
|
| 538 |
+
"mean_token_accuracy": 0.560819149017334,
|
| 539 |
+
"num_tokens": 209511.0,
|
| 540 |
+
"step": 59
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.4838709677419355,
|
| 544 |
+
"grad_norm": 6.996568202972412,
|
| 545 |
+
"learning_rate": 2.626622922096782e-07,
|
| 546 |
+
"loss": 1.7697,
|
| 547 |
+
"mean_token_accuracy": 0.5675398111343384,
|
| 548 |
+
"num_tokens": 213409.0,
|
| 549 |
+
"step": 60
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.49193548387096775,
|
| 553 |
+
"grad_norm": 8.156956672668457,
|
| 554 |
+
"learning_rate": 2.5633317857829693e-07,
|
| 555 |
+
"loss": 1.6853,
|
| 556 |
+
"mean_token_accuracy": 0.5745296478271484,
|
| 557 |
+
"num_tokens": 216868.0,
|
| 558 |
+
"step": 61
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.5,
|
| 562 |
+
"grad_norm": 8.733380317687988,
|
| 563 |
+
"learning_rate": 2.5e-07,
|
| 564 |
+
"loss": 1.6566,
|
| 565 |
+
"mean_token_accuracy": 0.5956632494926453,
|
| 566 |
+
"num_tokens": 220008.0,
|
| 567 |
+
"step": 62
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 0.5080645161290323,
|
| 571 |
+
"grad_norm": 6.773242473602295,
|
| 572 |
+
"learning_rate": 2.4366682142170305e-07,
|
| 573 |
+
"loss": 1.7601,
|
| 574 |
+
"mean_token_accuracy": 0.5683262944221497,
|
| 575 |
+
"num_tokens": 223788.0,
|
| 576 |
+
"step": 63
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5161290322580645,
|
| 580 |
+
"grad_norm": 6.594358444213867,
|
| 581 |
+
"learning_rate": 2.3733770779032184e-07,
|
| 582 |
+
"loss": 1.754,
|
| 583 |
+
"mean_token_accuracy": 0.5651482939720154,
|
| 584 |
+
"num_tokens": 227568.0,
|
| 585 |
+
"step": 64
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.5241935483870968,
|
| 589 |
+
"grad_norm": 7.253048419952393,
|
| 590 |
+
"learning_rate": 2.3101672144368846e-07,
|
| 591 |
+
"loss": 1.7411,
|
| 592 |
+
"mean_token_accuracy": 0.5777088403701782,
|
| 593 |
+
"num_tokens": 231439.0,
|
| 594 |
+
"step": 65
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.532258064516129,
|
| 598 |
+
"grad_norm": 6.758819580078125,
|
| 599 |
+
"learning_rate": 2.2470791950314196e-07,
|
| 600 |
+
"loss": 1.6841,
|
| 601 |
+
"mean_token_accuracy": 0.583798885345459,
|
| 602 |
+
"num_tokens": 235023.0,
|
| 603 |
+
"step": 66
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.5403225806451613,
|
| 607 |
+
"grad_norm": 7.620910167694092,
|
| 608 |
+
"learning_rate": 2.1841535126946775e-07,
|
| 609 |
+
"loss": 1.7435,
|
| 610 |
+
"mean_token_accuracy": 0.5806363224983215,
|
| 611 |
+
"num_tokens": 238673.0,
|
| 612 |
+
"step": 67
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.5483870967741935,
|
| 616 |
+
"grad_norm": 7.808475017547607,
|
| 617 |
+
"learning_rate": 2.121430556238559e-07,
|
| 618 |
+
"loss": 1.7878,
|
| 619 |
+
"mean_token_accuracy": 0.5660945773124695,
|
| 620 |
+
"num_tokens": 241998.0,
|
| 621 |
+
"step": 68
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5564516129032258,
|
| 625 |
+
"grad_norm": 10.025287628173828,
|
| 626 |
+
"learning_rate": 2.0589505843554795e-07,
|
| 627 |
+
"loss": 1.8058,
|
| 628 |
+
"mean_token_accuracy": 0.566918134689331,
|
| 629 |
+
"num_tokens": 245581.0,
|
| 630 |
+
"step": 69
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.5645161290322581,
|
| 634 |
+
"grad_norm": 7.763560771942139,
|
| 635 |
+
"learning_rate": 1.9967536997783493e-07,
|
| 636 |
+
"loss": 1.6668,
|
| 637 |
+
"mean_token_accuracy": 0.5830458402633667,
|
| 638 |
+
"num_tokens": 249643.0,
|
| 639 |
+
"step": 70
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.5725806451612904,
|
| 643 |
+
"grad_norm": 7.218297958374023,
|
| 644 |
+
"learning_rate": 1.9348798235406626e-07,
|
| 645 |
+
"loss": 1.6796,
|
| 646 |
+
"mean_token_accuracy": 0.5924479365348816,
|
| 647 |
+
"num_tokens": 253487.0,
|
| 648 |
+
"step": 71
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.5806451612903226,
|
| 652 |
+
"grad_norm": 7.534745693206787,
|
| 653 |
+
"learning_rate": 1.8733686693531982e-07,
|
| 654 |
+
"loss": 1.6667,
|
| 655 |
+
"mean_token_accuracy": 0.5781828761100769,
|
| 656 |
+
"num_tokens": 257002.0,
|
| 657 |
+
"step": 72
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.5887096774193549,
|
| 661 |
+
"grad_norm": 9.321051597595215,
|
| 662 |
+
"learning_rate": 1.8122597181138047e-07,
|
| 663 |
+
"loss": 1.6971,
|
| 664 |
+
"mean_token_accuracy": 0.5793435573577881,
|
| 665 |
+
"num_tokens": 260327.0,
|
| 666 |
+
"step": 73
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.5967741935483871,
|
| 670 |
+
"grad_norm": 7.620558261871338,
|
| 671 |
+
"learning_rate": 1.751592192566605e-07,
|
| 672 |
+
"loss": 1.8091,
|
| 673 |
+
"mean_token_accuracy": 0.5647743940353394,
|
| 674 |
+
"num_tokens": 263766.0,
|
| 675 |
+
"step": 74
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.6048387096774194,
|
| 679 |
+
"grad_norm": 6.960766315460205,
|
| 680 |
+
"learning_rate": 1.6914050321269047e-07,
|
| 681 |
+
"loss": 1.7232,
|
| 682 |
+
"mean_token_accuracy": 0.5714285969734192,
|
| 683 |
+
"num_tokens": 267431.0,
|
| 684 |
+
"step": 75
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.6129032258064516,
|
| 688 |
+
"grad_norm": 8.637341499328613,
|
| 689 |
+
"learning_rate": 1.6317368678879496e-07,
|
| 690 |
+
"loss": 1.8332,
|
| 691 |
+
"mean_token_accuracy": 0.5513805747032166,
|
| 692 |
+
"num_tokens": 270948.0,
|
| 693 |
+
"step": 76
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 0.6209677419354839,
|
| 697 |
+
"grad_norm": 6.999222278594971,
|
| 698 |
+
"learning_rate": 1.5726259978255807e-07,
|
| 699 |
+
"loss": 1.737,
|
| 700 |
+
"mean_token_accuracy": 0.578661322593689,
|
| 701 |
+
"num_tokens": 274448.0,
|
| 702 |
+
"step": 77
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.6290322580645161,
|
| 706 |
+
"grad_norm": 6.716799736022949,
|
| 707 |
+
"learning_rate": 1.514110362216704e-07,
|
| 708 |
+
"loss": 1.5999,
|
| 709 |
+
"mean_token_accuracy": 0.6048541069030762,
|
| 710 |
+
"num_tokens": 278119.0,
|
| 711 |
+
"step": 78
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.6370967741935484,
|
| 715 |
+
"grad_norm": 7.0507659912109375,
|
| 716 |
+
"learning_rate": 1.4562275192873428e-07,
|
| 717 |
+
"loss": 1.7363,
|
| 718 |
+
"mean_token_accuracy": 0.5828737020492554,
|
| 719 |
+
"num_tokens": 281568.0,
|
| 720 |
+
"step": 79
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6451612903225806,
|
| 724 |
+
"grad_norm": 8.561986923217773,
|
| 725 |
+
"learning_rate": 1.3990146211059139e-07,
|
| 726 |
+
"loss": 1.6661,
|
| 727 |
+
"mean_token_accuracy": 0.591465175151825,
|
| 728 |
+
"num_tokens": 285087.0,
|
| 729 |
+
"step": 80
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.6532258064516129,
|
| 733 |
+
"grad_norm": 7.29276704788208,
|
| 734 |
+
"learning_rate": 1.342508389737198e-07,
|
| 735 |
+
"loss": 1.6752,
|
| 736 |
+
"mean_token_accuracy": 0.5835070013999939,
|
| 737 |
+
"num_tokens": 288450.0,
|
| 738 |
+
"step": 81
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6612903225806451,
|
| 742 |
+
"grad_norm": 7.706023216247559,
|
| 743 |
+
"learning_rate": 1.2867450936722978e-07,
|
| 744 |
+
"loss": 1.7451,
|
| 745 |
+
"mean_token_accuracy": 0.5716311931610107,
|
| 746 |
+
"num_tokens": 291979.0,
|
| 747 |
+
"step": 82
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.6693548387096774,
|
| 751 |
+
"grad_norm": 7.056520462036133,
|
| 752 |
+
"learning_rate": 1.2317605245497323e-07,
|
| 753 |
+
"loss": 1.7154,
|
| 754 |
+
"mean_token_accuracy": 0.5758506059646606,
|
| 755 |
+
"num_tokens": 295892.0,
|
| 756 |
+
"step": 83
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.6774193548387096,
|
| 760 |
+
"grad_norm": 7.503317356109619,
|
| 761 |
+
"learning_rate": 1.1775899741825945e-07,
|
| 762 |
+
"loss": 1.6899,
|
| 763 |
+
"mean_token_accuracy": 0.5809495449066162,
|
| 764 |
+
"num_tokens": 299287.0,
|
| 765 |
+
"step": 84
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.6854838709677419,
|
| 769 |
+
"grad_norm": 6.785901069641113,
|
| 770 |
+
"learning_rate": 1.1242682119065216e-07,
|
| 771 |
+
"loss": 1.7004,
|
| 772 |
+
"mean_token_accuracy": 0.5904300808906555,
|
| 773 |
+
"num_tokens": 302802.0,
|
| 774 |
+
"step": 85
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.6935483870967742,
|
| 778 |
+
"grad_norm": 8.526169776916504,
|
| 779 |
+
"learning_rate": 1.0718294622630186e-07,
|
| 780 |
+
"loss": 1.764,
|
| 781 |
+
"mean_token_accuracy": 0.5716612339019775,
|
| 782 |
+
"num_tokens": 305876.0,
|
| 783 |
+
"step": 86
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.7016129032258065,
|
| 787 |
+
"grad_norm": 8.141705513000488,
|
| 788 |
+
"learning_rate": 1.0203073830324565e-07,
|
| 789 |
+
"loss": 1.6696,
|
| 790 |
+
"mean_token_accuracy": 0.5959654450416565,
|
| 791 |
+
"num_tokens": 309350.0,
|
| 792 |
+
"step": 87
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.7096774193548387,
|
| 796 |
+
"grad_norm": 7.417829513549805,
|
| 797 |
+
"learning_rate": 9.697350436308427e-08,
|
| 798 |
+
"loss": 1.6542,
|
| 799 |
+
"mean_token_accuracy": 0.5915982127189636,
|
| 800 |
+
"num_tokens": 312520.0,
|
| 801 |
+
"step": 88
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.717741935483871,
|
| 805 |
+
"grad_norm": 7.390657424926758,
|
| 806 |
+
"learning_rate": 9.201449038842401e-08,
|
| 807 |
+
"loss": 1.7805,
|
| 808 |
+
"mean_token_accuracy": 0.5598756074905396,
|
| 809 |
+
"num_tokens": 316382.0,
|
| 810 |
+
"step": 89
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.7258064516129032,
|
| 814 |
+
"grad_norm": 7.71380615234375,
|
| 815 |
+
"learning_rate": 8.715687931944449e-08,
|
| 816 |
+
"loss": 1.6886,
|
| 817 |
+
"mean_token_accuracy": 0.578542947769165,
|
| 818 |
+
"num_tokens": 319900.0,
|
| 819 |
+
"step": 90
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7338709677419355,
|
| 823 |
+
"grad_norm": 7.551061153411865,
|
| 824 |
+
"learning_rate": 8.240378901093034e-08,
|
| 825 |
+
"loss": 1.7469,
|
| 826 |
+
"mean_token_accuracy": 0.570781409740448,
|
| 827 |
+
"num_tokens": 323436.0,
|
| 828 |
+
"step": 91
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.7419354838709677,
|
| 832 |
+
"grad_norm": 9.389298439025879,
|
| 833 |
+
"learning_rate": 7.775827023107834e-08,
|
| 834 |
+
"loss": 1.7162,
|
| 835 |
+
"mean_token_accuracy": 0.571618914604187,
|
| 836 |
+
"num_tokens": 326442.0,
|
| 837 |
+
"step": 92
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.75,
|
| 841 |
+
"grad_norm": 7.125860691070557,
|
| 842 |
+
"learning_rate": 7.322330470336313e-08,
|
| 843 |
+
"loss": 1.7112,
|
| 844 |
+
"mean_token_accuracy": 0.568901777267456,
|
| 845 |
+
"num_tokens": 330234.0,
|
| 846 |
+
"step": 93
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 0.7580645161290323,
|
| 850 |
+
"grad_norm": 7.604709148406982,
|
| 851 |
+
"learning_rate": 6.880180319272006e-08,
|
| 852 |
+
"loss": 1.653,
|
| 853 |
+
"mean_token_accuracy": 0.5836288928985596,
|
| 854 |
+
"num_tokens": 333903.0,
|
| 855 |
+
"step": 94
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 0.7661290322580645,
|
| 859 |
+
"grad_norm": 6.952591896057129,
|
| 860 |
+
"learning_rate": 6.449660363727236e-08,
|
| 861 |
+
"loss": 1.6965,
|
| 862 |
+
"mean_token_accuracy": 0.5798975825309753,
|
| 863 |
+
"num_tokens": 337618.0,
|
| 864 |
+
"step": 95
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.7741935483870968,
|
| 868 |
+
"grad_norm": 10.35627555847168,
|
| 869 |
+
"learning_rate": 6.031046932680229e-08,
|
| 870 |
+
"loss": 1.6945,
|
| 871 |
+
"mean_token_accuracy": 0.5744901299476624,
|
| 872 |
+
"num_tokens": 340515.0,
|
| 873 |
+
"step": 96
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.782258064516129,
|
| 877 |
+
"grad_norm": 7.425570487976074,
|
| 878 |
+
"learning_rate": 5.624608712913531e-08,
|
| 879 |
+
"loss": 1.6154,
|
| 880 |
+
"mean_token_accuracy": 0.5917431116104126,
|
| 881 |
+
"num_tokens": 344007.0,
|
| 882 |
+
"step": 97
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.7903225806451613,
|
| 886 |
+
"grad_norm": 10.059042930603027,
|
| 887 |
+
"learning_rate": 5.230606576557539e-08,
|
| 888 |
+
"loss": 1.6855,
|
| 889 |
+
"mean_token_accuracy": 0.5853906869888306,
|
| 890 |
+
"num_tokens": 347940.0,
|
| 891 |
+
"step": 98
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.7983870967741935,
|
| 895 |
+
"grad_norm": 6.582756042480469,
|
| 896 |
+
"learning_rate": 4.84929341364988e-08,
|
| 897 |
+
"loss": 1.6939,
|
| 898 |
+
"mean_token_accuracy": 0.5809850096702576,
|
| 899 |
+
"num_tokens": 351741.0,
|
| 900 |
+
"step": 99
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.8064516129032258,
|
| 904 |
+
"grad_norm": 6.744507312774658,
|
| 905 |
+
"learning_rate": 4.480913969818098e-08,
|
| 906 |
+
"loss": 1.6739,
|
| 907 |
+
"mean_token_accuracy": 0.585814356803894,
|
| 908 |
+
"num_tokens": 355171.0,
|
| 909 |
+
"step": 100
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.8145161290322581,
|
| 913 |
+
"grad_norm": 6.796554088592529,
|
| 914 |
+
"learning_rate": 4.125704689189818e-08,
|
| 915 |
+
"loss": 1.7039,
|
| 916 |
+
"mean_token_accuracy": 0.5784457325935364,
|
| 917 |
+
"num_tokens": 359267.0,
|
| 918 |
+
"step": 101
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8225806451612904,
|
| 922 |
+
"grad_norm": 6.474706649780273,
|
| 923 |
+
"learning_rate": 3.783893562631224e-08,
|
| 924 |
+
"loss": 1.6828,
|
| 925 |
+
"mean_token_accuracy": 0.5838965773582458,
|
| 926 |
+
"num_tokens": 362674.0,
|
| 927 |
+
"step": 102
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 0.8306451612903226,
|
| 931 |
+
"grad_norm": 7.683028221130371,
|
| 932 |
+
"learning_rate": 3.455699981411259e-08,
|
| 933 |
+
"loss": 1.7309,
|
| 934 |
+
"mean_token_accuracy": 0.5660528540611267,
|
| 935 |
+
"num_tokens": 366425.0,
|
| 936 |
+
"step": 103
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.8387096774193549,
|
| 940 |
+
"grad_norm": 7.123412609100342,
|
| 941 |
+
"learning_rate": 3.141334596385447e-08,
|
| 942 |
+
"loss": 1.6526,
|
| 943 |
+
"mean_token_accuracy": 0.5894578099250793,
|
| 944 |
+
"num_tokens": 369749.0,
|
| 945 |
+
"step": 104
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.8467741935483871,
|
| 949 |
+
"grad_norm": 6.300046920776367,
|
| 950 |
+
"learning_rate": 2.8409991827897968e-08,
|
| 951 |
+
"loss": 1.6927,
|
| 952 |
+
"mean_token_accuracy": 0.5794023275375366,
|
| 953 |
+
"num_tokens": 373802.0,
|
| 954 |
+
"step": 105
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.8548387096774194,
|
| 958 |
+
"grad_norm": 6.947690486907959,
|
| 959 |
+
"learning_rate": 2.5548865107314604e-08,
|
| 960 |
+
"loss": 1.7412,
|
| 961 |
+
"mean_token_accuracy": 0.5656623244285583,
|
| 962 |
+
"num_tokens": 377324.0,
|
| 963 |
+
"step": 106
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.8629032258064516,
|
| 967 |
+
"grad_norm": 7.4605231285095215,
|
| 968 |
+
"learning_rate": 2.283180221459377e-08,
|
| 969 |
+
"loss": 1.7679,
|
| 970 |
+
"mean_token_accuracy": 0.5644617676734924,
|
| 971 |
+
"num_tokens": 381183.0,
|
| 972 |
+
"step": 107
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 0.8709677419354839,
|
| 976 |
+
"grad_norm": 6.687449932098389,
|
| 977 |
+
"learning_rate": 2.0260547094942348e-08,
|
| 978 |
+
"loss": 1.7144,
|
| 979 |
+
"mean_token_accuracy": 0.5790040493011475,
|
| 980 |
+
"num_tokens": 384902.0,
|
| 981 |
+
"step": 108
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.8790322580645161,
|
| 985 |
+
"grad_norm": 9.971823692321777,
|
| 986 |
+
"learning_rate": 1.7836750106934474e-08,
|
| 987 |
+
"loss": 1.6716,
|
| 988 |
+
"mean_token_accuracy": 0.5851125121116638,
|
| 989 |
+
"num_tokens": 388372.0,
|
| 990 |
+
"step": 109
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.8870967741935484,
|
| 994 |
+
"grad_norm": 8.792091369628906,
|
| 995 |
+
"learning_rate": 1.5561966963229923e-08,
|
| 996 |
+
"loss": 1.66,
|
| 997 |
+
"mean_token_accuracy": 0.5835981965065002,
|
| 998 |
+
"num_tokens": 391522.0,
|
| 999 |
+
"step": 110
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.8951612903225806,
|
| 1003 |
+
"grad_norm": 7.2787981033325195,
|
| 1004 |
+
"learning_rate": 1.3437657732040781e-08,
|
| 1005 |
+
"loss": 1.6463,
|
| 1006 |
+
"mean_token_accuracy": 0.5921754837036133,
|
| 1007 |
+
"num_tokens": 394900.0,
|
| 1008 |
+
"step": 111
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.9032258064516129,
|
| 1012 |
+
"grad_norm": 7.135777473449707,
|
| 1013 |
+
"learning_rate": 1.1465185899987794e-08,
|
| 1014 |
+
"loss": 1.7416,
|
| 1015 |
+
"mean_token_accuracy": 0.5735452771186829,
|
| 1016 |
+
"num_tokens": 398616.0,
|
| 1017 |
+
"step": 112
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9112903225806451,
|
| 1021 |
+
"grad_norm": 8.646471977233887,
|
| 1022 |
+
"learning_rate": 9.6458174969469e-09,
|
| 1023 |
+
"loss": 1.7218,
|
| 1024 |
+
"mean_token_accuracy": 0.5801047086715698,
|
| 1025 |
+
"num_tokens": 402440.0,
|
| 1026 |
+
"step": 113
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.9193548387096774,
|
| 1030 |
+
"grad_norm": 6.620112895965576,
|
| 1031 |
+
"learning_rate": 7.980720283448955e-09,
|
| 1032 |
+
"loss": 1.8393,
|
| 1033 |
+
"mean_token_accuracy": 0.5539906024932861,
|
| 1034 |
+
"num_tokens": 406278.0,
|
| 1035 |
+
"step": 114
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.9274193548387096,
|
| 1039 |
+
"grad_norm": 8.354401588439941,
|
| 1040 |
+
"learning_rate": 6.470963001153268e-09,
|
| 1041 |
+
"loss": 1.7239,
|
| 1042 |
+
"mean_token_accuracy": 0.5888278484344482,
|
| 1043 |
+
"num_tokens": 409558.0,
|
| 1044 |
+
"step": 115
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.9354838709677419,
|
| 1048 |
+
"grad_norm": 7.131601810455322,
|
| 1049 |
+
"learning_rate": 5.117514686876378e-09,
|
| 1050 |
+
"loss": 1.6136,
|
| 1051 |
+
"mean_token_accuracy": 0.5943509340286255,
|
| 1052 |
+
"num_tokens": 412890.0,
|
| 1053 |
+
"step": 116
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.9435483870967742,
|
| 1057 |
+
"grad_norm": 7.938460826873779,
|
| 1058 |
+
"learning_rate": 3.921244050616446e-09,
|
| 1059 |
+
"loss": 1.7333,
|
| 1060 |
+
"mean_token_accuracy": 0.5793216824531555,
|
| 1061 |
+
"num_tokens": 416550.0,
|
| 1062 |
+
"step": 117
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.9516129032258065,
|
| 1066 |
+
"grad_norm": 7.657175540924072,
|
| 1067 |
+
"learning_rate": 2.8829189179721547e-09,
|
| 1068 |
+
"loss": 1.7309,
|
| 1069 |
+
"mean_token_accuracy": 0.5766178369522095,
|
| 1070 |
+
"num_tokens": 419830.0,
|
| 1071 |
+
"step": 118
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.9596774193548387,
|
| 1075 |
+
"grad_norm": 7.384486198425293,
|
| 1076 |
+
"learning_rate": 2.0032057373142453e-09,
|
| 1077 |
+
"loss": 1.6211,
|
| 1078 |
+
"mean_token_accuracy": 0.5845938324928284,
|
| 1079 |
+
"num_tokens": 423404.0,
|
| 1080 |
+
"step": 119
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.967741935483871,
|
| 1084 |
+
"grad_norm": 8.56894588470459,
|
| 1085 |
+
"learning_rate": 1.2826691520262112e-09,
|
| 1086 |
+
"loss": 1.7963,
|
| 1087 |
+
"mean_token_accuracy": 0.57413250207901,
|
| 1088 |
+
"num_tokens": 426895.0,
|
| 1089 |
+
"step": 120
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.9758064516129032,
|
| 1093 |
+
"grad_norm": 6.813057899475098,
|
| 1094 |
+
"learning_rate": 7.217716380881477e-10,
|
| 1095 |
+
"loss": 1.726,
|
| 1096 |
+
"mean_token_accuracy": 0.5653586387634277,
|
| 1097 |
+
"num_tokens": 430900.0,
|
| 1098 |
+
"step": 121
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9838709677419355,
|
| 1102 |
+
"grad_norm": 6.8685431480407715,
|
| 1103 |
+
"learning_rate": 3.2087320723681033e-10,
|
| 1104 |
+
"loss": 1.6799,
|
| 1105 |
+
"mean_token_accuracy": 0.5889589190483093,
|
| 1106 |
+
"num_tokens": 434237.0,
|
| 1107 |
+
"step": 122
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.9919354838709677,
|
| 1111 |
+
"grad_norm": 7.024580955505371,
|
| 1112 |
+
"learning_rate": 8.023117589237016e-11,
|
| 1113 |
+
"loss": 1.7037,
|
| 1114 |
+
"mean_token_accuracy": 0.5847502946853638,
|
| 1115 |
+
"num_tokens": 438005.0,
|
| 1116 |
+
"step": 123
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 1.0,
|
| 1120 |
+
"grad_norm": 8.294002532958984,
|
| 1121 |
+
"learning_rate": 0.0,
|
| 1122 |
+
"loss": 1.6386,
|
| 1123 |
+
"mean_token_accuracy": 0.5888568758964539,
|
| 1124 |
+
"num_tokens": 441132.0,
|
| 1125 |
+
"step": 124
|
| 1126 |
+
}
|
| 1127 |
+
],
|
| 1128 |
+
"logging_steps": 1,
|
| 1129 |
+
"max_steps": 124,
|
| 1130 |
+
"num_input_tokens_seen": 0,
|
| 1131 |
+
"num_train_epochs": 1,
|
| 1132 |
+
"save_steps": 3000,
|
| 1133 |
+
"stateful_callbacks": {
|
| 1134 |
+
"TrainerControl": {
|
| 1135 |
+
"args": {
|
| 1136 |
+
"should_epoch_stop": false,
|
| 1137 |
+
"should_evaluate": false,
|
| 1138 |
+
"should_log": false,
|
| 1139 |
+
"should_save": true,
|
| 1140 |
+
"should_training_stop": true
|
| 1141 |
+
},
|
| 1142 |
+
"attributes": {}
|
| 1143 |
+
}
|
| 1144 |
+
},
|
| 1145 |
+
"total_flos": 1.0678265542868992e+16,
|
| 1146 |
+
"train_batch_size": 2,
|
| 1147 |
+
"trial_name": null,
|
| 1148 |
+
"trial_params": null
|
| 1149 |
+
}
|