Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +1013 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,1013 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.7692307692307693,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 140,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.005494505494505495,
|
| 13 |
+
"grad_norm": 0.29632216691970825,
|
| 14 |
+
"learning_rate": 4e-05,
|
| 15 |
+
"loss": 1.5561,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01098901098901099,
|
| 20 |
+
"grad_norm": 0.3113888204097748,
|
| 21 |
+
"learning_rate": 8e-05,
|
| 22 |
+
"loss": 1.5458,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.016483516483516484,
|
| 27 |
+
"grad_norm": 0.3019237518310547,
|
| 28 |
+
"learning_rate": 0.00012,
|
| 29 |
+
"loss": 1.5507,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02197802197802198,
|
| 34 |
+
"grad_norm": 0.29094433784484863,
|
| 35 |
+
"learning_rate": 0.00016,
|
| 36 |
+
"loss": 1.4652,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.027472527472527472,
|
| 41 |
+
"grad_norm": 0.27828261256217957,
|
| 42 |
+
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 1.4454,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03296703296703297,
|
| 48 |
+
"grad_norm": 0.3085501492023468,
|
| 49 |
+
"learning_rate": 0.0001999842488663838,
|
| 50 |
+
"loss": 1.1417,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.038461538461538464,
|
| 55 |
+
"grad_norm": 0.30620744824409485,
|
| 56 |
+
"learning_rate": 0.00019993700042749937,
|
| 57 |
+
"loss": 1.225,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04395604395604396,
|
| 62 |
+
"grad_norm": 0.41407617926597595,
|
| 63 |
+
"learning_rate": 0.0001998582695676762,
|
| 64 |
+
"loss": 0.9375,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.04945054945054945,
|
| 69 |
+
"grad_norm": 0.3822758197784424,
|
| 70 |
+
"learning_rate": 0.00019974808108892016,
|
| 71 |
+
"loss": 0.8788,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.054945054945054944,
|
| 76 |
+
"grad_norm": 0.40259653329849243,
|
| 77 |
+
"learning_rate": 0.00019960646970310027,
|
| 78 |
+
"loss": 0.7664,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.06043956043956044,
|
| 83 |
+
"grad_norm": 0.5633108019828796,
|
| 84 |
+
"learning_rate": 0.00019943348002101371,
|
| 85 |
+
"loss": 0.6699,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06593406593406594,
|
| 90 |
+
"grad_norm": 0.37629789113998413,
|
| 91 |
+
"learning_rate": 0.00019922916653833248,
|
| 92 |
+
"loss": 0.5892,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.07142857142857142,
|
| 97 |
+
"grad_norm": 0.31924572587013245,
|
| 98 |
+
"learning_rate": 0.0001989935936184358,
|
| 99 |
+
"loss": 0.5198,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.07692307692307693,
|
| 104 |
+
"grad_norm": 0.38061216473579407,
|
| 105 |
+
"learning_rate": 0.00019872683547213446,
|
| 106 |
+
"loss": 0.5894,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.08241758241758242,
|
| 111 |
+
"grad_norm": 0.2537476122379303,
|
| 112 |
+
"learning_rate": 0.00019842897613429262,
|
| 113 |
+
"loss": 0.5341,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.08791208791208792,
|
| 118 |
+
"grad_norm": 0.24057163298130035,
|
| 119 |
+
"learning_rate": 0.00019810010943735479,
|
| 120 |
+
"loss": 0.3744,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.09340659340659341,
|
| 125 |
+
"grad_norm": 0.2103324830532074,
|
| 126 |
+
"learning_rate": 0.00019774033898178667,
|
| 127 |
+
"loss": 0.5054,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.0989010989010989,
|
| 132 |
+
"grad_norm": 0.18747441470623016,
|
| 133 |
+
"learning_rate": 0.00019734977810343865,
|
| 134 |
+
"loss": 0.3734,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1043956043956044,
|
| 139 |
+
"grad_norm": 0.19926677644252777,
|
| 140 |
+
"learning_rate": 0.00019692854983784235,
|
| 141 |
+
"loss": 0.399,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.10989010989010989,
|
| 146 |
+
"grad_norm": 0.20050929486751556,
|
| 147 |
+
"learning_rate": 0.0001964767868814516,
|
| 148 |
+
"loss": 0.4754,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.11538461538461539,
|
| 153 |
+
"grad_norm": 0.18073034286499023,
|
| 154 |
+
"learning_rate": 0.0001959946315498402,
|
| 155 |
+
"loss": 0.4375,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.12087912087912088,
|
| 160 |
+
"grad_norm": 0.20278413593769073,
|
| 161 |
+
"learning_rate": 0.0001954822357328692,
|
| 162 |
+
"loss": 0.4896,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.12637362637362637,
|
| 167 |
+
"grad_norm": 0.16931243240833282,
|
| 168 |
+
"learning_rate": 0.00019493976084683813,
|
| 169 |
+
"loss": 0.4137,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.13186813186813187,
|
| 174 |
+
"grad_norm": 0.18224675953388214,
|
| 175 |
+
"learning_rate": 0.00019436737778363527,
|
| 176 |
+
"loss": 0.3895,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.13736263736263737,
|
| 181 |
+
"grad_norm": 0.16282841563224792,
|
| 182 |
+
"learning_rate": 0.0001937652668569028,
|
| 183 |
+
"loss": 0.364,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.14285714285714285,
|
| 188 |
+
"grad_norm": 0.19350826740264893,
|
| 189 |
+
"learning_rate": 0.00019313361774523385,
|
| 190 |
+
"loss": 0.3578,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.14835164835164835,
|
| 195 |
+
"grad_norm": 0.20993147790431976,
|
| 196 |
+
"learning_rate": 0.0001924726294324196,
|
| 197 |
+
"loss": 0.4351,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.15384615384615385,
|
| 202 |
+
"grad_norm": 0.20188163220882416,
|
| 203 |
+
"learning_rate": 0.00019178251014476466,
|
| 204 |
+
"loss": 0.3965,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.15934065934065933,
|
| 209 |
+
"grad_norm": 0.2008114606142044,
|
| 210 |
+
"learning_rate": 0.00019106347728549135,
|
| 211 |
+
"loss": 0.3193,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.16483516483516483,
|
| 216 |
+
"grad_norm": 0.1973695605993271,
|
| 217 |
+
"learning_rate": 0.00019031575736625238,
|
| 218 |
+
"loss": 0.298,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.17032967032967034,
|
| 223 |
+
"grad_norm": 0.20849934220314026,
|
| 224 |
+
"learning_rate": 0.00018953958593577493,
|
| 225 |
+
"loss": 0.2995,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.17582417582417584,
|
| 230 |
+
"grad_norm": 0.20263291895389557,
|
| 231 |
+
"learning_rate": 0.00018873520750565718,
|
| 232 |
+
"loss": 0.3021,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.1813186813186813,
|
| 237 |
+
"grad_norm": 0.20541806519031525,
|
| 238 |
+
"learning_rate": 0.00018790287547334176,
|
| 239 |
+
"loss": 0.2883,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.18681318681318682,
|
| 244 |
+
"grad_norm": 0.24142159521579742,
|
| 245 |
+
"learning_rate": 0.00018704285204228973,
|
| 246 |
+
"loss": 0.3396,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.19230769230769232,
|
| 251 |
+
"grad_norm": 0.20562608540058136,
|
| 252 |
+
"learning_rate": 0.0001861554081393806,
|
| 253 |
+
"loss": 0.2773,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.1978021978021978,
|
| 258 |
+
"grad_norm": 0.26048898696899414,
|
| 259 |
+
"learning_rate": 0.00018524082332956428,
|
| 260 |
+
"loss": 0.3024,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2032967032967033,
|
| 265 |
+
"grad_norm": 0.23103846609592438,
|
| 266 |
+
"learning_rate": 0.00018429938572779152,
|
| 267 |
+
"loss": 0.2843,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2087912087912088,
|
| 272 |
+
"grad_norm": 0.23378689587116241,
|
| 273 |
+
"learning_rate": 0.0001833313919082515,
|
| 274 |
+
"loss": 0.2963,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.21428571428571427,
|
| 279 |
+
"grad_norm": 0.27318716049194336,
|
| 280 |
+
"learning_rate": 0.00018233714681094404,
|
| 281 |
+
"loss": 0.3863,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.21978021978021978,
|
| 286 |
+
"grad_norm": 0.219524085521698,
|
| 287 |
+
"learning_rate": 0.00018131696364561667,
|
| 288 |
+
"loss": 0.2703,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.22527472527472528,
|
| 293 |
+
"grad_norm": 0.21837233006954193,
|
| 294 |
+
"learning_rate": 0.00018027116379309638,
|
| 295 |
+
"loss": 0.2695,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.23076923076923078,
|
| 300 |
+
"grad_norm": 0.2384708672761917,
|
| 301 |
+
"learning_rate": 0.0001792000767040474,
|
| 302 |
+
"loss": 0.2819,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.23626373626373626,
|
| 307 |
+
"grad_norm": 0.20765846967697144,
|
| 308 |
+
"learning_rate": 0.00017810403979518681,
|
| 309 |
+
"loss": 0.2658,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.24175824175824176,
|
| 314 |
+
"grad_norm": 0.22932475805282593,
|
| 315 |
+
"learning_rate": 0.00017698339834299061,
|
| 316 |
+
"loss": 0.2276,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.24725274725274726,
|
| 321 |
+
"grad_norm": 0.2080792635679245,
|
| 322 |
+
"learning_rate": 0.00017583850537492387,
|
| 323 |
+
"loss": 0.1902,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.25274725274725274,
|
| 328 |
+
"grad_norm": 0.2754678428173065,
|
| 329 |
+
"learning_rate": 0.0001746697215582288,
|
| 330 |
+
"loss": 0.2797,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.25824175824175827,
|
| 335 |
+
"grad_norm": 0.22715018689632416,
|
| 336 |
+
"learning_rate": 0.00017347741508630672,
|
| 337 |
+
"loss": 0.1703,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.26373626373626374,
|
| 342 |
+
"grad_norm": 0.31407058238983154,
|
| 343 |
+
"learning_rate": 0.00017226196156272874,
|
| 344 |
+
"loss": 0.281,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.2692307692307692,
|
| 349 |
+
"grad_norm": 0.2527458071708679,
|
| 350 |
+
"learning_rate": 0.00017102374388291183,
|
| 351 |
+
"loss": 0.2579,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.27472527472527475,
|
| 356 |
+
"grad_norm": 0.2112707644701004,
|
| 357 |
+
"learning_rate": 0.0001697631521134985,
|
| 358 |
+
"loss": 0.264,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.2802197802197802,
|
| 363 |
+
"grad_norm": 0.20465674996376038,
|
| 364 |
+
"learning_rate": 0.00016848058336947657,
|
| 365 |
+
"loss": 0.2882,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.2857142857142857,
|
| 370 |
+
"grad_norm": 0.2256706804037094,
|
| 371 |
+
"learning_rate": 0.0001671764416890793,
|
| 372 |
+
"loss": 0.1847,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.29120879120879123,
|
| 377 |
+
"grad_norm": 0.20961597561836243,
|
| 378 |
+
"learning_rate": 0.00016585113790650388,
|
| 379 |
+
"loss": 0.196,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.2967032967032967,
|
| 384 |
+
"grad_norm": 0.2462984025478363,
|
| 385 |
+
"learning_rate": 0.00016450508952248956,
|
| 386 |
+
"loss": 0.2032,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.3021978021978022,
|
| 391 |
+
"grad_norm": 0.22251084446907043,
|
| 392 |
+
"learning_rate": 0.00016313872057279534,
|
| 393 |
+
"loss": 0.203,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3076923076923077,
|
| 398 |
+
"grad_norm": 0.23791608214378357,
|
| 399 |
+
"learning_rate": 0.0001617524614946192,
|
| 400 |
+
"loss": 0.2458,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.3131868131868132,
|
| 405 |
+
"grad_norm": 0.23427598178386688,
|
| 406 |
+
"learning_rate": 0.0001603467489910004,
|
| 407 |
+
"loss": 0.149,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.31868131868131866,
|
| 412 |
+
"grad_norm": 0.23513749241828918,
|
| 413 |
+
"learning_rate": 0.00015892202589324835,
|
| 414 |
+
"loss": 0.1449,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.3241758241758242,
|
| 419 |
+
"grad_norm": 0.23785251379013062,
|
| 420 |
+
"learning_rate": 0.0001574787410214407,
|
| 421 |
+
"loss": 0.1938,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.32967032967032966,
|
| 426 |
+
"grad_norm": 0.20233368873596191,
|
| 427 |
+
"learning_rate": 0.0001560173490430346,
|
| 428 |
+
"loss": 0.1226,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.33516483516483514,
|
| 433 |
+
"grad_norm": 0.2437211126089096,
|
| 434 |
+
"learning_rate": 0.0001545383103296365,
|
| 435 |
+
"loss": 0.1892,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.34065934065934067,
|
| 440 |
+
"grad_norm": 0.2462146282196045,
|
| 441 |
+
"learning_rate": 0.00015304209081197425,
|
| 442 |
+
"loss": 0.1581,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.34615384615384615,
|
| 447 |
+
"grad_norm": 0.3353044092655182,
|
| 448 |
+
"learning_rate": 0.0001515291618331188,
|
| 449 |
+
"loss": 0.2195,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.3516483516483517,
|
| 454 |
+
"grad_norm": 0.22197988629341125,
|
| 455 |
+
"learning_rate": 0.00015000000000000001,
|
| 456 |
+
"loss": 0.1739,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.35714285714285715,
|
| 461 |
+
"grad_norm": 0.2737789750099182,
|
| 462 |
+
"learning_rate": 0.00014845508703326504,
|
| 463 |
+
"loss": 0.1717,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.3626373626373626,
|
| 468 |
+
"grad_norm": 0.22385859489440918,
|
| 469 |
+
"learning_rate": 0.00014689490961552513,
|
| 470 |
+
"loss": 0.1724,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.36813186813186816,
|
| 475 |
+
"grad_norm": 0.27109456062316895,
|
| 476 |
+
"learning_rate": 0.00014531995923803973,
|
| 477 |
+
"loss": 0.196,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.37362637362637363,
|
| 482 |
+
"grad_norm": 0.20588435232639313,
|
| 483 |
+
"learning_rate": 0.00014373073204588556,
|
| 484 |
+
"loss": 0.1351,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.3791208791208791,
|
| 489 |
+
"grad_norm": 0.22479820251464844,
|
| 490 |
+
"learning_rate": 0.00014212772868165958,
|
| 491 |
+
"loss": 0.1083,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.38461538461538464,
|
| 496 |
+
"grad_norm": 0.2374928891658783,
|
| 497 |
+
"learning_rate": 0.00014051145412776535,
|
| 498 |
+
"loss": 0.136,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.3901098901098901,
|
| 503 |
+
"grad_norm": 0.25980404019355774,
|
| 504 |
+
"learning_rate": 0.00013888241754733208,
|
| 505 |
+
"loss": 0.1505,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.3956043956043956,
|
| 510 |
+
"grad_norm": 0.3704344928264618,
|
| 511 |
+
"learning_rate": 0.0001372411321238166,
|
| 512 |
+
"loss": 0.2365,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.4010989010989011,
|
| 517 |
+
"grad_norm": 0.2545515298843384,
|
| 518 |
+
"learning_rate": 0.00013558811489933908,
|
| 519 |
+
"loss": 0.1645,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.4065934065934066,
|
| 524 |
+
"grad_norm": 0.2218877524137497,
|
| 525 |
+
"learning_rate": 0.00013392388661180303,
|
| 526 |
+
"loss": 0.1178,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.41208791208791207,
|
| 531 |
+
"grad_norm": 0.19732226431369781,
|
| 532 |
+
"learning_rate": 0.0001322489715308509,
|
| 533 |
+
"loss": 0.1001,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.4175824175824176,
|
| 538 |
+
"grad_norm": 0.28702080249786377,
|
| 539 |
+
"learning_rate": 0.00013056389729270738,
|
| 540 |
+
"loss": 0.1437,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4230769230769231,
|
| 545 |
+
"grad_norm": 0.27200955152511597,
|
| 546 |
+
"learning_rate": 0.0001288691947339621,
|
| 547 |
+
"loss": 0.1753,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.42857142857142855,
|
| 552 |
+
"grad_norm": 0.20188194513320923,
|
| 553 |
+
"learning_rate": 0.00012716539772434388,
|
| 554 |
+
"loss": 0.0998,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.4340659340659341,
|
| 559 |
+
"grad_norm": 0.2388327270746231,
|
| 560 |
+
"learning_rate": 0.00012545304299853977,
|
| 561 |
+
"loss": 0.1339,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.43956043956043955,
|
| 566 |
+
"grad_norm": 0.21373318135738373,
|
| 567 |
+
"learning_rate": 0.0001237326699871115,
|
| 568 |
+
"loss": 0.1205,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.44505494505494503,
|
| 573 |
+
"grad_norm": 0.29054194688796997,
|
| 574 |
+
"learning_rate": 0.00012200482064656248,
|
| 575 |
+
"loss": 0.1632,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.45054945054945056,
|
| 580 |
+
"grad_norm": 0.24942484498023987,
|
| 581 |
+
"learning_rate": 0.00012027003928860937,
|
| 582 |
+
"loss": 0.1282,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.45604395604395603,
|
| 587 |
+
"grad_norm": 0.20856431126594543,
|
| 588 |
+
"learning_rate": 0.00011852887240871145,
|
| 589 |
+
"loss": 0.1168,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.46153846153846156,
|
| 594 |
+
"grad_norm": 0.2552940249443054,
|
| 595 |
+
"learning_rate": 0.00011678186851391218,
|
| 596 |
+
"loss": 0.1269,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.46703296703296704,
|
| 601 |
+
"grad_norm": 0.2256869375705719,
|
| 602 |
+
"learning_rate": 0.00011502957795004705,
|
| 603 |
+
"loss": 0.1043,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.4725274725274725,
|
| 608 |
+
"grad_norm": 0.20186680555343628,
|
| 609 |
+
"learning_rate": 0.00011327255272837221,
|
| 610 |
+
"loss": 0.0929,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.47802197802197804,
|
| 615 |
+
"grad_norm": 0.2341729998588562,
|
| 616 |
+
"learning_rate": 0.00011151134635166829,
|
| 617 |
+
"loss": 0.0981,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.4835164835164835,
|
| 622 |
+
"grad_norm": 0.23417356610298157,
|
| 623 |
+
"learning_rate": 0.00010974651363987465,
|
| 624 |
+
"loss": 0.1066,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.489010989010989,
|
| 629 |
+
"grad_norm": 0.18209996819496155,
|
| 630 |
+
"learning_rate": 0.00010797861055530831,
|
| 631 |
+
"loss": 0.1032,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.4945054945054945,
|
| 636 |
+
"grad_norm": 0.18998686969280243,
|
| 637 |
+
"learning_rate": 0.0001062081940275234,
|
| 638 |
+
"loss": 0.0965,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.5,
|
| 643 |
+
"grad_norm": 0.20998521149158478,
|
| 644 |
+
"learning_rate": 0.00010443582177786564,
|
| 645 |
+
"loss": 0.144,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5054945054945055,
|
| 650 |
+
"grad_norm": 0.22074103355407715,
|
| 651 |
+
"learning_rate": 0.00010266205214377748,
|
| 652 |
+
"loss": 0.1184,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.510989010989011,
|
| 657 |
+
"grad_norm": 0.15477561950683594,
|
| 658 |
+
"learning_rate": 0.0001008874439029091,
|
| 659 |
+
"loss": 0.0796,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.5164835164835165,
|
| 664 |
+
"grad_norm": 0.2496633678674698,
|
| 665 |
+
"learning_rate": 9.91125560970909e-05,
|
| 666 |
+
"loss": 0.142,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.521978021978022,
|
| 671 |
+
"grad_norm": 0.3139898180961609,
|
| 672 |
+
"learning_rate": 9.733794785622253e-05,
|
| 673 |
+
"loss": 0.1407,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.5274725274725275,
|
| 678 |
+
"grad_norm": 0.21977433562278748,
|
| 679 |
+
"learning_rate": 9.556417822213435e-05,
|
| 680 |
+
"loss": 0.1164,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.532967032967033,
|
| 685 |
+
"grad_norm": 0.19279415905475616,
|
| 686 |
+
"learning_rate": 9.379180597247661e-05,
|
| 687 |
+
"loss": 0.0876,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.5384615384615384,
|
| 692 |
+
"grad_norm": 0.23302212357521057,
|
| 693 |
+
"learning_rate": 9.202138944469168e-05,
|
| 694 |
+
"loss": 0.1269,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.5439560439560439,
|
| 699 |
+
"grad_norm": 0.1825583428144455,
|
| 700 |
+
"learning_rate": 9.025348636012536e-05,
|
| 701 |
+
"loss": 0.0936,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.5494505494505495,
|
| 706 |
+
"grad_norm": 0.19986358284950256,
|
| 707 |
+
"learning_rate": 8.84886536483317e-05,
|
| 708 |
+
"loss": 0.113,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.554945054945055,
|
| 713 |
+
"grad_norm": 0.19531255960464478,
|
| 714 |
+
"learning_rate": 8.672744727162781e-05,
|
| 715 |
+
"loss": 0.0941,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.5604395604395604,
|
| 720 |
+
"grad_norm": 0.2396000772714615,
|
| 721 |
+
"learning_rate": 8.497042204995299e-05,
|
| 722 |
+
"loss": 0.0992,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.5659340659340659,
|
| 727 |
+
"grad_norm": 0.196890726685524,
|
| 728 |
+
"learning_rate": 8.321813148608783e-05,
|
| 729 |
+
"loss": 0.0833,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.5714285714285714,
|
| 734 |
+
"grad_norm": 0.2729422450065613,
|
| 735 |
+
"learning_rate": 8.147112759128859e-05,
|
| 736 |
+
"loss": 0.1452,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.5769230769230769,
|
| 741 |
+
"grad_norm": 0.1796470433473587,
|
| 742 |
+
"learning_rate": 7.972996071139064e-05,
|
| 743 |
+
"loss": 0.0799,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.5824175824175825,
|
| 748 |
+
"grad_norm": 0.14928919076919556,
|
| 749 |
+
"learning_rate": 7.799517935343757e-05,
|
| 750 |
+
"loss": 0.0481,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.5879120879120879,
|
| 755 |
+
"grad_norm": 0.21496616303920746,
|
| 756 |
+
"learning_rate": 7.626733001288851e-05,
|
| 757 |
+
"loss": 0.0693,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.5934065934065934,
|
| 762 |
+
"grad_norm": 0.24284154176712036,
|
| 763 |
+
"learning_rate": 7.454695700146024e-05,
|
| 764 |
+
"loss": 0.1201,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.5989010989010989,
|
| 769 |
+
"grad_norm": 0.2560741603374481,
|
| 770 |
+
"learning_rate": 7.283460227565613e-05,
|
| 771 |
+
"loss": 0.1095,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.6043956043956044,
|
| 776 |
+
"grad_norm": 0.13009899854660034,
|
| 777 |
+
"learning_rate": 7.113080526603792e-05,
|
| 778 |
+
"loss": 0.0512,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.6098901098901099,
|
| 783 |
+
"grad_norm": 0.20721551775932312,
|
| 784 |
+
"learning_rate": 6.94361027072926e-05,
|
| 785 |
+
"loss": 0.1132,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6153846153846154,
|
| 790 |
+
"grad_norm": 0.154499813914299,
|
| 791 |
+
"learning_rate": 6.775102846914911e-05,
|
| 792 |
+
"loss": 0.0755,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6208791208791209,
|
| 797 |
+
"grad_norm": 0.19335274398326874,
|
| 798 |
+
"learning_rate": 6.607611338819697e-05,
|
| 799 |
+
"loss": 0.0847,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.6263736263736264,
|
| 804 |
+
"grad_norm": 0.17462438344955444,
|
| 805 |
+
"learning_rate": 6.441188510066091e-05,
|
| 806 |
+
"loss": 0.0569,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.6318681318681318,
|
| 811 |
+
"grad_norm": 0.33881354331970215,
|
| 812 |
+
"learning_rate": 6.275886787618339e-05,
|
| 813 |
+
"loss": 0.1436,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.6373626373626373,
|
| 818 |
+
"grad_norm": 0.14344196021556854,
|
| 819 |
+
"learning_rate": 6.111758245266794e-05,
|
| 820 |
+
"loss": 0.0628,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.6428571428571429,
|
| 825 |
+
"grad_norm": 0.10723075270652771,
|
| 826 |
+
"learning_rate": 5.9488545872234645e-05,
|
| 827 |
+
"loss": 0.0462,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.6483516483516484,
|
| 832 |
+
"grad_norm": 0.19554950296878815,
|
| 833 |
+
"learning_rate": 5.787227131834043e-05,
|
| 834 |
+
"loss": 0.1092,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.6538461538461539,
|
| 839 |
+
"grad_norm": 0.13956406712532043,
|
| 840 |
+
"learning_rate": 5.626926795411447e-05,
|
| 841 |
+
"loss": 0.0674,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.6593406593406593,
|
| 846 |
+
"grad_norm": 0.2507220506668091,
|
| 847 |
+
"learning_rate": 5.468004076196029e-05,
|
| 848 |
+
"loss": 0.1,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.6648351648351648,
|
| 853 |
+
"grad_norm": 0.1673005074262619,
|
| 854 |
+
"learning_rate": 5.310509038447492e-05,
|
| 855 |
+
"loss": 0.1039,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.6703296703296703,
|
| 860 |
+
"grad_norm": 0.12703153491020203,
|
| 861 |
+
"learning_rate": 5.1544912966734994e-05,
|
| 862 |
+
"loss": 0.0704,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.6758241758241759,
|
| 867 |
+
"grad_norm": 0.17319592833518982,
|
| 868 |
+
"learning_rate": 5.000000000000002e-05,
|
| 869 |
+
"loss": 0.0724,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.6813186813186813,
|
| 874 |
+
"grad_norm": 0.154402956366539,
|
| 875 |
+
"learning_rate": 4.8470838166881226e-05,
|
| 876 |
+
"loss": 0.047,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.6868131868131868,
|
| 881 |
+
"grad_norm": 0.15553316473960876,
|
| 882 |
+
"learning_rate": 4.695790918802576e-05,
|
| 883 |
+
"loss": 0.0932,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.6923076923076923,
|
| 888 |
+
"grad_norm": 0.1012202799320221,
|
| 889 |
+
"learning_rate": 4.546168967036351e-05,
|
| 890 |
+
"loss": 0.0456,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.6978021978021978,
|
| 895 |
+
"grad_norm": 0.1133142039179802,
|
| 896 |
+
"learning_rate": 4.398265095696539e-05,
|
| 897 |
+
"loss": 0.0482,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.7032967032967034,
|
| 902 |
+
"grad_norm": 0.14953628182411194,
|
| 903 |
+
"learning_rate": 4.252125897855932e-05,
|
| 904 |
+
"loss": 0.0577,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.7087912087912088,
|
| 909 |
+
"grad_norm": 0.145247682929039,
|
| 910 |
+
"learning_rate": 4.107797410675166e-05,
|
| 911 |
+
"loss": 0.0645,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.7142857142857143,
|
| 916 |
+
"grad_norm": 0.09949145466089249,
|
| 917 |
+
"learning_rate": 3.965325100899961e-05,
|
| 918 |
+
"loss": 0.0486,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.7197802197802198,
|
| 923 |
+
"grad_norm": 0.128681942820549,
|
| 924 |
+
"learning_rate": 3.824753850538082e-05,
|
| 925 |
+
"loss": 0.0561,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.7252747252747253,
|
| 930 |
+
"grad_norm": 0.16942492127418518,
|
| 931 |
+
"learning_rate": 3.686127942720463e-05,
|
| 932 |
+
"loss": 0.0614,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.7307692307692307,
|
| 937 |
+
"grad_norm": 0.1377490609884262,
|
| 938 |
+
"learning_rate": 3.5494910477510445e-05,
|
| 939 |
+
"loss": 0.0634,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.7362637362637363,
|
| 944 |
+
"grad_norm": 0.1743365377187729,
|
| 945 |
+
"learning_rate": 3.414886209349615e-05,
|
| 946 |
+
"loss": 0.0695,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.7417582417582418,
|
| 951 |
+
"grad_norm": 0.18100859224796295,
|
| 952 |
+
"learning_rate": 3.282355831092072e-05,
|
| 953 |
+
"loss": 0.0561,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.7472527472527473,
|
| 958 |
+
"grad_norm": 0.1113915666937828,
|
| 959 |
+
"learning_rate": 3.1519416630523444e-05,
|
| 960 |
+
"loss": 0.0609,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.7527472527472527,
|
| 965 |
+
"grad_norm": 0.11914083361625671,
|
| 966 |
+
"learning_rate": 3.0236847886501542e-05,
|
| 967 |
+
"loss": 0.0437,
|
| 968 |
+
"step": 137
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.7582417582417582,
|
| 972 |
+
"grad_norm": 0.16115868091583252,
|
| 973 |
+
"learning_rate": 2.8976256117088195e-05,
|
| 974 |
+
"loss": 0.0668,
|
| 975 |
+
"step": 138
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.7637362637362637,
|
| 979 |
+
"grad_norm": 0.1613149493932724,
|
| 980 |
+
"learning_rate": 2.7738038437271284e-05,
|
| 981 |
+
"loss": 0.0769,
|
| 982 |
+
"step": 139
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.7692307692307693,
|
| 986 |
+
"grad_norm": 0.12686750292778015,
|
| 987 |
+
"learning_rate": 2.6522584913693294e-05,
|
| 988 |
+
"loss": 0.0586,
|
| 989 |
+
"step": 140
|
| 990 |
+
}
|
| 991 |
+
],
|
| 992 |
+
"logging_steps": 1,
|
| 993 |
+
"max_steps": 182,
|
| 994 |
+
"num_input_tokens_seen": 0,
|
| 995 |
+
"num_train_epochs": 1,
|
| 996 |
+
"save_steps": 20,
|
| 997 |
+
"stateful_callbacks": {
|
| 998 |
+
"TrainerControl": {
|
| 999 |
+
"args": {
|
| 1000 |
+
"should_epoch_stop": false,
|
| 1001 |
+
"should_evaluate": false,
|
| 1002 |
+
"should_log": false,
|
| 1003 |
+
"should_save": true,
|
| 1004 |
+
"should_training_stop": false
|
| 1005 |
+
},
|
| 1006 |
+
"attributes": {}
|
| 1007 |
+
}
|
| 1008 |
+
},
|
| 1009 |
+
"total_flos": 4.078551699224986e+16,
|
| 1010 |
+
"train_batch_size": 2,
|
| 1011 |
+
"trial_name": null,
|
| 1012 |
+
"trial_params": null
|
| 1013 |
+
}
|