Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +994 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,994 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 16.93658536585366,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 136,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11707317073170732,
|
| 13 |
+
"grad_norm": 1.713741660118103,
|
| 14 |
+
"learning_rate": 7.142857142857143e-07,
|
| 15 |
+
"loss": 0.391,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.23414634146341465,
|
| 20 |
+
"grad_norm": 1.7321815490722656,
|
| 21 |
+
"learning_rate": 1.4285714285714286e-06,
|
| 22 |
+
"loss": 0.3876,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.35121951219512193,
|
| 27 |
+
"grad_norm": 1.869097352027893,
|
| 28 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 29 |
+
"loss": 0.4163,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.4682926829268293,
|
| 34 |
+
"grad_norm": 1.741655707359314,
|
| 35 |
+
"learning_rate": 2.8571428571428573e-06,
|
| 36 |
+
"loss": 0.4025,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.5853658536585366,
|
| 41 |
+
"grad_norm": 1.5887295007705688,
|
| 42 |
+
"learning_rate": 3.5714285714285718e-06,
|
| 43 |
+
"loss": 0.383,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.7024390243902439,
|
| 48 |
+
"grad_norm": 1.5869324207305908,
|
| 49 |
+
"learning_rate": 4.2857142857142855e-06,
|
| 50 |
+
"loss": 0.3968,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.8195121951219512,
|
| 55 |
+
"grad_norm": 1.3928492069244385,
|
| 56 |
+
"learning_rate": 5e-06,
|
| 57 |
+
"loss": 0.3868,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.9365853658536586,
|
| 62 |
+
"grad_norm": 1.317595362663269,
|
| 63 |
+
"learning_rate": 5.7142857142857145e-06,
|
| 64 |
+
"loss": 0.3871,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.1170731707317074,
|
| 69 |
+
"grad_norm": 1.3958970308303833,
|
| 70 |
+
"learning_rate": 6.4285714285714295e-06,
|
| 71 |
+
"loss": 0.7529,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.2341463414634146,
|
| 76 |
+
"grad_norm": 0.588860273361206,
|
| 77 |
+
"learning_rate": 7.1428571428571436e-06,
|
| 78 |
+
"loss": 0.361,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.3512195121951218,
|
| 83 |
+
"grad_norm": 0.5455438494682312,
|
| 84 |
+
"learning_rate": 7.857142857142858e-06,
|
| 85 |
+
"loss": 0.3647,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.4682926829268292,
|
| 90 |
+
"grad_norm": 1.2507004737854004,
|
| 91 |
+
"learning_rate": 8.571428571428571e-06,
|
| 92 |
+
"loss": 0.3795,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.5853658536585367,
|
| 97 |
+
"grad_norm": 1.3338210582733154,
|
| 98 |
+
"learning_rate": 9.285714285714288e-06,
|
| 99 |
+
"loss": 0.3783,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.7024390243902439,
|
| 104 |
+
"grad_norm": 1.2802133560180664,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 0.3637,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.819512195121951,
|
| 111 |
+
"grad_norm": 1.1926186084747314,
|
| 112 |
+
"learning_rate": 9.998342337571566e-06,
|
| 113 |
+
"loss": 0.3663,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.9365853658536585,
|
| 118 |
+
"grad_norm": 0.8796733021736145,
|
| 119 |
+
"learning_rate": 9.993370449424153e-06,
|
| 120 |
+
"loss": 0.3446,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.1170731707317074,
|
| 125 |
+
"grad_norm": 1.2581405639648438,
|
| 126 |
+
"learning_rate": 9.985087632242634e-06,
|
| 127 |
+
"loss": 0.6666,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 2.234146341463415,
|
| 132 |
+
"grad_norm": 0.6097764372825623,
|
| 133 |
+
"learning_rate": 9.973499378072947e-06,
|
| 134 |
+
"loss": 0.3404,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 2.351219512195122,
|
| 139 |
+
"grad_norm": 0.5183754563331604,
|
| 140 |
+
"learning_rate": 9.958613370680507e-06,
|
| 141 |
+
"loss": 0.331,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 2.4682926829268292,
|
| 146 |
+
"grad_norm": 0.5353206992149353,
|
| 147 |
+
"learning_rate": 9.940439480455386e-06,
|
| 148 |
+
"loss": 0.3364,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.5853658536585367,
|
| 153 |
+
"grad_norm": 0.5248769521713257,
|
| 154 |
+
"learning_rate": 9.918989757867584e-06,
|
| 155 |
+
"loss": 0.3289,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.7024390243902436,
|
| 160 |
+
"grad_norm": 0.5104326009750366,
|
| 161 |
+
"learning_rate": 9.89427842547679e-06,
|
| 162 |
+
"loss": 0.3207,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.819512195121951,
|
| 167 |
+
"grad_norm": 0.429746150970459,
|
| 168 |
+
"learning_rate": 9.866321868501914e-06,
|
| 169 |
+
"loss": 0.3354,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.9365853658536585,
|
| 174 |
+
"grad_norm": 0.4242963492870331,
|
| 175 |
+
"learning_rate": 9.835138623956603e-06,
|
| 176 |
+
"loss": 0.3149,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 3.1170731707317074,
|
| 181 |
+
"grad_norm": 0.8024505972862244,
|
| 182 |
+
"learning_rate": 9.80074936835801e-06,
|
| 183 |
+
"loss": 0.6097,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 3.234146341463415,
|
| 188 |
+
"grad_norm": 0.3315274715423584,
|
| 189 |
+
"learning_rate": 9.763176904016914e-06,
|
| 190 |
+
"loss": 0.3016,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 3.351219512195122,
|
| 195 |
+
"grad_norm": 0.3314255475997925,
|
| 196 |
+
"learning_rate": 9.722446143918307e-06,
|
| 197 |
+
"loss": 0.2876,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 3.4682926829268292,
|
| 202 |
+
"grad_norm": 0.30298393964767456,
|
| 203 |
+
"learning_rate": 9.678584095202468e-06,
|
| 204 |
+
"loss": 0.2912,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 3.5853658536585367,
|
| 209 |
+
"grad_norm": 0.3056536614894867,
|
| 210 |
+
"learning_rate": 9.631619841257477e-06,
|
| 211 |
+
"loss": 0.2855,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 3.7024390243902436,
|
| 216 |
+
"grad_norm": 0.3028172552585602,
|
| 217 |
+
"learning_rate": 9.581584522435025e-06,
|
| 218 |
+
"loss": 0.2848,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.819512195121951,
|
| 223 |
+
"grad_norm": 0.3157103657722473,
|
| 224 |
+
"learning_rate": 9.528511315402358e-06,
|
| 225 |
+
"loss": 0.2891,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.9365853658536585,
|
| 230 |
+
"grad_norm": 0.28788891434669495,
|
| 231 |
+
"learning_rate": 9.472435411143979e-06,
|
| 232 |
+
"loss": 0.274,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 4.117073170731707,
|
| 237 |
+
"grad_norm": 0.562315821647644,
|
| 238 |
+
"learning_rate": 9.413393991627737e-06,
|
| 239 |
+
"loss": 0.5517,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 4.234146341463415,
|
| 244 |
+
"grad_norm": 0.24586421251296997,
|
| 245 |
+
"learning_rate": 9.351426205150778e-06,
|
| 246 |
+
"loss": 0.2557,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 4.351219512195122,
|
| 251 |
+
"grad_norm": 0.24129828810691833,
|
| 252 |
+
"learning_rate": 9.286573140381663e-06,
|
| 253 |
+
"loss": 0.2506,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.46829268292683,
|
| 258 |
+
"grad_norm": 0.2212359756231308,
|
| 259 |
+
"learning_rate": 9.218877799115929e-06,
|
| 260 |
+
"loss": 0.2583,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 4.585365853658536,
|
| 265 |
+
"grad_norm": 0.20721423625946045,
|
| 266 |
+
"learning_rate": 9.148385067763094e-06,
|
| 267 |
+
"loss": 0.2509,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 4.702439024390244,
|
| 272 |
+
"grad_norm": 0.22884435951709747,
|
| 273 |
+
"learning_rate": 9.075141687584056e-06,
|
| 274 |
+
"loss": 0.248,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 4.819512195121951,
|
| 279 |
+
"grad_norm": 0.24287782609462738,
|
| 280 |
+
"learning_rate": 8.999196223698599e-06,
|
| 281 |
+
"loss": 0.2443,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 4.9365853658536585,
|
| 286 |
+
"grad_norm": 0.2220289707183838,
|
| 287 |
+
"learning_rate": 8.920599032883553e-06,
|
| 288 |
+
"loss": 0.235,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 5.117073170731707,
|
| 293 |
+
"grad_norm": 0.4429185092449188,
|
| 294 |
+
"learning_rate": 8.839402230183e-06,
|
| 295 |
+
"loss": 0.4562,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 5.234146341463415,
|
| 300 |
+
"grad_norm": 0.2300422042608261,
|
| 301 |
+
"learning_rate": 8.755659654352599e-06,
|
| 302 |
+
"loss": 0.2245,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 5.351219512195122,
|
| 307 |
+
"grad_norm": 0.21638786792755127,
|
| 308 |
+
"learning_rate": 8.669426832160997e-06,
|
| 309 |
+
"loss": 0.2268,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 5.46829268292683,
|
| 314 |
+
"grad_norm": 0.2475624680519104,
|
| 315 |
+
"learning_rate": 8.580760941571968e-06,
|
| 316 |
+
"loss": 0.2026,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 5.585365853658536,
|
| 321 |
+
"grad_norm": 0.2283526360988617,
|
| 322 |
+
"learning_rate": 8.489720773831717e-06,
|
| 323 |
+
"loss": 0.203,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.702439024390244,
|
| 328 |
+
"grad_norm": 0.19043877720832825,
|
| 329 |
+
"learning_rate": 8.396366694486466e-06,
|
| 330 |
+
"loss": 0.2039,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 5.819512195121951,
|
| 335 |
+
"grad_norm": 0.2230452299118042,
|
| 336 |
+
"learning_rate": 8.30076060335616e-06,
|
| 337 |
+
"loss": 0.1957,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 5.9365853658536585,
|
| 342 |
+
"grad_norm": 0.2054775357246399,
|
| 343 |
+
"learning_rate": 8.202965893490877e-06,
|
| 344 |
+
"loss": 0.202,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 6.117073170731707,
|
| 349 |
+
"grad_norm": 0.19972442090511322,
|
| 350 |
+
"learning_rate": 8.103047409137114e-06,
|
| 351 |
+
"loss": 0.3737,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 6.234146341463415,
|
| 356 |
+
"grad_norm": 0.20269611477851868,
|
| 357 |
+
"learning_rate": 8.001071402741843e-06,
|
| 358 |
+
"loss": 0.1804,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 6.351219512195122,
|
| 363 |
+
"grad_norm": 0.2113686352968216,
|
| 364 |
+
"learning_rate": 7.897105491022819e-06,
|
| 365 |
+
"loss": 0.1755,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 6.46829268292683,
|
| 370 |
+
"grad_norm": 0.19944241642951965,
|
| 371 |
+
"learning_rate": 7.791218610134324e-06,
|
| 372 |
+
"loss": 0.1576,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 6.585365853658536,
|
| 377 |
+
"grad_norm": 0.1904984563589096,
|
| 378 |
+
"learning_rate": 7.683480969958005e-06,
|
| 379 |
+
"loss": 0.1586,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 6.702439024390244,
|
| 384 |
+
"grad_norm": 0.19858157634735107,
|
| 385 |
+
"learning_rate": 7.5739640075491546e-06,
|
| 386 |
+
"loss": 0.1628,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 6.819512195121951,
|
| 391 |
+
"grad_norm": 0.1883428692817688,
|
| 392 |
+
"learning_rate": 7.462740339769323e-06,
|
| 393 |
+
"loss": 0.1632,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.9365853658536585,
|
| 398 |
+
"grad_norm": 0.17427878081798553,
|
| 399 |
+
"learning_rate": 7.349883715136601e-06,
|
| 400 |
+
"loss": 0.1653,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 7.117073170731707,
|
| 405 |
+
"grad_norm": 0.37863096594810486,
|
| 406 |
+
"learning_rate": 7.235468964925571e-06,
|
| 407 |
+
"loss": 0.3005,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 7.234146341463415,
|
| 412 |
+
"grad_norm": 0.1933353841304779,
|
| 413 |
+
"learning_rate": 7.119571953549305e-06,
|
| 414 |
+
"loss": 0.1439,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 7.351219512195122,
|
| 419 |
+
"grad_norm": 0.1970248520374298,
|
| 420 |
+
"learning_rate": 7.002269528256334e-06,
|
| 421 |
+
"loss": 0.132,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 7.46829268292683,
|
| 426 |
+
"grad_norm": 0.19500941038131714,
|
| 427 |
+
"learning_rate": 6.883639468175926e-06,
|
| 428 |
+
"loss": 0.1264,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 7.585365853658536,
|
| 433 |
+
"grad_norm": 0.20704524219036102,
|
| 434 |
+
"learning_rate": 6.763760432745475e-06,
|
| 435 |
+
"loss": 0.135,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 7.702439024390244,
|
| 440 |
+
"grad_norm": 0.21069225668907166,
|
| 441 |
+
"learning_rate": 6.6427119095541745e-06,
|
| 442 |
+
"loss": 0.1168,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 7.819512195121951,
|
| 447 |
+
"grad_norm": 0.185212641954422,
|
| 448 |
+
"learning_rate": 6.520574161637591e-06,
|
| 449 |
+
"loss": 0.1252,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 7.9365853658536585,
|
| 454 |
+
"grad_norm": 0.18015943467617035,
|
| 455 |
+
"learning_rate": 6.397428174258048e-06,
|
| 456 |
+
"loss": 0.1249,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 8.117073170731707,
|
| 461 |
+
"grad_norm": 0.5003377199172974,
|
| 462 |
+
"learning_rate": 6.273355601206143e-06,
|
| 463 |
+
"loss": 0.2434,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 8.234146341463415,
|
| 468 |
+
"grad_norm": 0.2011335790157318,
|
| 469 |
+
"learning_rate": 6.148438710658979e-06,
|
| 470 |
+
"loss": 0.0987,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 8.351219512195122,
|
| 475 |
+
"grad_norm": 0.1703123301267624,
|
| 476 |
+
"learning_rate": 6.022760330631006e-06,
|
| 477 |
+
"loss": 0.1057,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 8.46829268292683,
|
| 482 |
+
"grad_norm": 0.17905612289905548,
|
| 483 |
+
"learning_rate": 5.896403794053679e-06,
|
| 484 |
+
"loss": 0.1019,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 8.585365853658537,
|
| 489 |
+
"grad_norm": 0.22483955323696136,
|
| 490 |
+
"learning_rate": 5.76945288352031e-06,
|
| 491 |
+
"loss": 0.1018,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 8.702439024390245,
|
| 496 |
+
"grad_norm": 0.18365544080734253,
|
| 497 |
+
"learning_rate": 5.641991775732756e-06,
|
| 498 |
+
"loss": 0.0906,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 8.819512195121952,
|
| 503 |
+
"grad_norm": 0.16530556976795197,
|
| 504 |
+
"learning_rate": 5.514104985686802e-06,
|
| 505 |
+
"loss": 0.0961,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 8.93658536585366,
|
| 510 |
+
"grad_norm": 0.18940307199954987,
|
| 511 |
+
"learning_rate": 5.385877310633233e-06,
|
| 512 |
+
"loss": 0.0921,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 9.117073170731707,
|
| 517 |
+
"grad_norm": 0.3870483934879303,
|
| 518 |
+
"learning_rate": 5.257393773851733e-06,
|
| 519 |
+
"loss": 0.1844,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 9.234146341463415,
|
| 524 |
+
"grad_norm": 0.19212201237678528,
|
| 525 |
+
"learning_rate": 5.1287395682749444e-06,
|
| 526 |
+
"loss": 0.0799,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 9.351219512195122,
|
| 531 |
+
"grad_norm": 0.19690005481243134,
|
| 532 |
+
"learning_rate": 5e-06,
|
| 533 |
+
"loss": 0.0757,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 9.46829268292683,
|
| 538 |
+
"grad_norm": 0.18531833589076996,
|
| 539 |
+
"learning_rate": 4.871260431725058e-06,
|
| 540 |
+
"loss": 0.0746,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 9.585365853658537,
|
| 545 |
+
"grad_norm": 0.2481042891740799,
|
| 546 |
+
"learning_rate": 4.742606226148268e-06,
|
| 547 |
+
"loss": 0.064,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 9.702439024390245,
|
| 552 |
+
"grad_norm": 0.1442754566669464,
|
| 553 |
+
"learning_rate": 4.614122689366769e-06,
|
| 554 |
+
"loss": 0.0741,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 9.819512195121952,
|
| 559 |
+
"grad_norm": 0.1746453046798706,
|
| 560 |
+
"learning_rate": 4.485895014313198e-06,
|
| 561 |
+
"loss": 0.0712,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 9.93658536585366,
|
| 566 |
+
"grad_norm": 0.19312545657157898,
|
| 567 |
+
"learning_rate": 4.358008224267245e-06,
|
| 568 |
+
"loss": 0.0747,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 10.117073170731707,
|
| 573 |
+
"grad_norm": 0.4135829508304596,
|
| 574 |
+
"learning_rate": 4.230547116479691e-06,
|
| 575 |
+
"loss": 0.1344,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 10.234146341463415,
|
| 580 |
+
"grad_norm": 0.14636236429214478,
|
| 581 |
+
"learning_rate": 4.103596205946323e-06,
|
| 582 |
+
"loss": 0.0564,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 10.351219512195122,
|
| 587 |
+
"grad_norm": 0.18952858448028564,
|
| 588 |
+
"learning_rate": 3.977239669368998e-06,
|
| 589 |
+
"loss": 0.0567,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 10.46829268292683,
|
| 594 |
+
"grad_norm": 0.23925091326236725,
|
| 595 |
+
"learning_rate": 3.851561289341023e-06,
|
| 596 |
+
"loss": 0.0575,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 10.585365853658537,
|
| 601 |
+
"grad_norm": 0.20485427975654602,
|
| 602 |
+
"learning_rate": 3.726644398793857e-06,
|
| 603 |
+
"loss": 0.0571,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 10.702439024390245,
|
| 608 |
+
"grad_norm": 0.15516437590122223,
|
| 609 |
+
"learning_rate": 3.6025718257419532e-06,
|
| 610 |
+
"loss": 0.0515,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 10.819512195121952,
|
| 615 |
+
"grad_norm": 0.20174065232276917,
|
| 616 |
+
"learning_rate": 3.4794258383624115e-06,
|
| 617 |
+
"loss": 0.0538,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 10.93658536585366,
|
| 622 |
+
"grad_norm": 0.23884522914886475,
|
| 623 |
+
"learning_rate": 3.3572880904458267e-06,
|
| 624 |
+
"loss": 0.0557,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 11.117073170731707,
|
| 629 |
+
"grad_norm": 0.3778754770755768,
|
| 630 |
+
"learning_rate": 3.236239567254526e-06,
|
| 631 |
+
"loss": 0.1121,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 11.234146341463415,
|
| 636 |
+
"grad_norm": 0.15436743199825287,
|
| 637 |
+
"learning_rate": 3.116360531824074e-06,
|
| 638 |
+
"loss": 0.0422,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 11.351219512195122,
|
| 643 |
+
"grad_norm": 0.14751587808132172,
|
| 644 |
+
"learning_rate": 2.997730471743667e-06,
|
| 645 |
+
"loss": 0.0417,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 11.46829268292683,
|
| 650 |
+
"grad_norm": 0.23487572371959686,
|
| 651 |
+
"learning_rate": 2.880428046450697e-06,
|
| 652 |
+
"loss": 0.0396,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 11.585365853658537,
|
| 657 |
+
"grad_norm": 0.2069467157125473,
|
| 658 |
+
"learning_rate": 2.7645310350744296e-06,
|
| 659 |
+
"loss": 0.0426,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 11.702439024390245,
|
| 664 |
+
"grad_norm": 0.15420763194561005,
|
| 665 |
+
"learning_rate": 2.6501162848634023e-06,
|
| 666 |
+
"loss": 0.0413,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 11.819512195121952,
|
| 671 |
+
"grad_norm": 0.1196109727025032,
|
| 672 |
+
"learning_rate": 2.537259660230679e-06,
|
| 673 |
+
"loss": 0.0404,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 11.93658536585366,
|
| 678 |
+
"grad_norm": 0.1467064768075943,
|
| 679 |
+
"learning_rate": 2.426035992450848e-06,
|
| 680 |
+
"loss": 0.0407,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 12.117073170731707,
|
| 685 |
+
"grad_norm": 0.19361571967601776,
|
| 686 |
+
"learning_rate": 2.316519030041998e-06,
|
| 687 |
+
"loss": 0.0791,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 12.234146341463415,
|
| 692 |
+
"grad_norm": 0.13011707365512848,
|
| 693 |
+
"learning_rate": 2.2087813898656775e-06,
|
| 694 |
+
"loss": 0.0346,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 12.351219512195122,
|
| 699 |
+
"grad_norm": 0.10667014867067337,
|
| 700 |
+
"learning_rate": 2.102894508977182e-06,
|
| 701 |
+
"loss": 0.0327,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 12.46829268292683,
|
| 706 |
+
"grad_norm": 0.14411939680576324,
|
| 707 |
+
"learning_rate": 1.9989285972581595e-06,
|
| 708 |
+
"loss": 0.0326,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 12.585365853658537,
|
| 713 |
+
"grad_norm": 0.19949135184288025,
|
| 714 |
+
"learning_rate": 1.896952590862886e-06,
|
| 715 |
+
"loss": 0.0343,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 12.702439024390245,
|
| 720 |
+
"grad_norm": 0.18770211935043335,
|
| 721 |
+
"learning_rate": 1.7970341065091246e-06,
|
| 722 |
+
"loss": 0.031,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 12.819512195121952,
|
| 727 |
+
"grad_norm": 0.1456199288368225,
|
| 728 |
+
"learning_rate": 1.699239396643841e-06,
|
| 729 |
+
"loss": 0.0283,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 12.93658536585366,
|
| 734 |
+
"grad_norm": 0.140044167637825,
|
| 735 |
+
"learning_rate": 1.6036333055135345e-06,
|
| 736 |
+
"loss": 0.0331,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 13.117073170731707,
|
| 741 |
+
"grad_norm": 0.3027644157409668,
|
| 742 |
+
"learning_rate": 1.5102792261682813e-06,
|
| 743 |
+
"loss": 0.0651,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 13.234146341463415,
|
| 748 |
+
"grad_norm": 0.16215696930885315,
|
| 749 |
+
"learning_rate": 1.4192390584280347e-06,
|
| 750 |
+
"loss": 0.0296,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 13.351219512195122,
|
| 755 |
+
"grad_norm": 0.1622275710105896,
|
| 756 |
+
"learning_rate": 1.330573167839005e-06,
|
| 757 |
+
"loss": 0.0265,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 13.46829268292683,
|
| 762 |
+
"grad_norm": 0.13922300934791565,
|
| 763 |
+
"learning_rate": 1.2443403456474017e-06,
|
| 764 |
+
"loss": 0.0277,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 13.585365853658537,
|
| 769 |
+
"grad_norm": 0.08746293187141418,
|
| 770 |
+
"learning_rate": 1.1605977698170001e-06,
|
| 771 |
+
"loss": 0.0237,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 13.702439024390245,
|
| 776 |
+
"grad_norm": 0.09428137540817261,
|
| 777 |
+
"learning_rate": 1.0794009671164484e-06,
|
| 778 |
+
"loss": 0.0255,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 13.819512195121952,
|
| 783 |
+
"grad_norm": 0.10664989054203033,
|
| 784 |
+
"learning_rate": 1.0008037763014033e-06,
|
| 785 |
+
"loss": 0.0265,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 13.93658536585366,
|
| 790 |
+
"grad_norm": 0.12778587639331818,
|
| 791 |
+
"learning_rate": 9.248583124159438e-07,
|
| 792 |
+
"loss": 0.0265,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 14.117073170731707,
|
| 797 |
+
"grad_norm": 0.25158995389938354,
|
| 798 |
+
"learning_rate": 8.516149322369055e-07,
|
| 799 |
+
"loss": 0.0504,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 14.234146341463415,
|
| 804 |
+
"grad_norm": 0.08846470713615417,
|
| 805 |
+
"learning_rate": 7.811222008840719e-07,
|
| 806 |
+
"loss": 0.0218,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 14.351219512195122,
|
| 811 |
+
"grad_norm": 0.08048608899116516,
|
| 812 |
+
"learning_rate": 7.13426859618338e-07,
|
| 813 |
+
"loss": 0.0234,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 14.46829268292683,
|
| 818 |
+
"grad_norm": 0.08340264111757278,
|
| 819 |
+
"learning_rate": 6.485737948492237e-07,
|
| 820 |
+
"loss": 0.0237,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 14.585365853658537,
|
| 825 |
+
"grad_norm": 0.08076529949903488,
|
| 826 |
+
"learning_rate": 5.866060083722624e-07,
|
| 827 |
+
"loss": 0.0212,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 14.702439024390245,
|
| 832 |
+
"grad_norm": 0.08138881623744965,
|
| 833 |
+
"learning_rate": 5.275645888560233e-07,
|
| 834 |
+
"loss": 0.0221,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 14.819512195121952,
|
| 839 |
+
"grad_norm": 0.08238395303487778,
|
| 840 |
+
"learning_rate": 4.71488684597643e-07,
|
| 841 |
+
"loss": 0.0222,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 14.93658536585366,
|
| 846 |
+
"grad_norm": 0.08457454293966293,
|
| 847 |
+
"learning_rate": 4.184154775649768e-07,
|
| 848 |
+
"loss": 0.0231,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 15.117073170731707,
|
| 853 |
+
"grad_norm": 0.21646344661712646,
|
| 854 |
+
"learning_rate": 3.683801587425251e-07,
|
| 855 |
+
"loss": 0.0459,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 15.234146341463415,
|
| 860 |
+
"grad_norm": 0.08210847526788712,
|
| 861 |
+
"learning_rate": 3.214159047975324e-07,
|
| 862 |
+
"loss": 0.0243,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 15.351219512195122,
|
| 867 |
+
"grad_norm": 0.0840378999710083,
|
| 868 |
+
"learning_rate": 2.7755385608169374e-07,
|
| 869 |
+
"loss": 0.0205,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 15.46829268292683,
|
| 874 |
+
"grad_norm": 0.07520610839128494,
|
| 875 |
+
"learning_rate": 2.368230959830875e-07,
|
| 876 |
+
"loss": 0.0196,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 15.585365853658537,
|
| 881 |
+
"grad_norm": 0.07876352220773697,
|
| 882 |
+
"learning_rate": 1.992506316419912e-07,
|
| 883 |
+
"loss": 0.0204,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 15.702439024390245,
|
| 888 |
+
"grad_norm": 0.07992194592952728,
|
| 889 |
+
"learning_rate": 1.6486137604339813e-07,
|
| 890 |
+
"loss": 0.0211,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 15.819512195121952,
|
| 895 |
+
"grad_norm": 0.08172820508480072,
|
| 896 |
+
"learning_rate": 1.3367813149808728e-07,
|
| 897 |
+
"loss": 0.0229,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 15.93658536585366,
|
| 902 |
+
"grad_norm": 0.07874112576246262,
|
| 903 |
+
"learning_rate": 1.0572157452321097e-07,
|
| 904 |
+
"loss": 0.0221,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 16.117073170731707,
|
| 909 |
+
"grad_norm": 0.19976511597633362,
|
| 910 |
+
"learning_rate": 8.101024213241826e-08,
|
| 911 |
+
"loss": 0.0391,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 16.234146341463415,
|
| 916 |
+
"grad_norm": 0.0725630447268486,
|
| 917 |
+
"learning_rate": 5.9560519544614725e-08,
|
| 918 |
+
"loss": 0.0225,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 16.351219512195122,
|
| 923 |
+
"grad_norm": 0.07620209455490112,
|
| 924 |
+
"learning_rate": 4.138662931949255e-08,
|
| 925 |
+
"loss": 0.0197,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 16.46829268292683,
|
| 930 |
+
"grad_norm": 0.06960424035787582,
|
| 931 |
+
"learning_rate": 2.6500621927054716e-08,
|
| 932 |
+
"loss": 0.0205,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 16.585365853658537,
|
| 937 |
+
"grad_norm": 0.0744452178478241,
|
| 938 |
+
"learning_rate": 1.4912367757366485e-08,
|
| 939 |
+
"loss": 0.0204,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 16.702439024390245,
|
| 944 |
+
"grad_norm": 0.07866977900266647,
|
| 945 |
+
"learning_rate": 6.629550575847355e-09,
|
| 946 |
+
"loss": 0.0212,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 16.819512195121952,
|
| 951 |
+
"grad_norm": 0.07277688384056091,
|
| 952 |
+
"learning_rate": 1.657662428434792e-09,
|
| 953 |
+
"loss": 0.0186,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 16.93658536585366,
|
| 958 |
+
"grad_norm": 0.06631285697221756,
|
| 959 |
+
"learning_rate": 0.0,
|
| 960 |
+
"loss": 0.0242,
|
| 961 |
+
"step": 136
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 16.93658536585366,
|
| 965 |
+
"step": 136,
|
| 966 |
+
"total_flos": 483897984679936.0,
|
| 967 |
+
"train_loss": 0.16699845409568617,
|
| 968 |
+
"train_runtime": 26245.2704,
|
| 969 |
+
"train_samples_per_second": 0.531,
|
| 970 |
+
"train_steps_per_second": 0.005
|
| 971 |
+
}
|
| 972 |
+
],
|
| 973 |
+
"logging_steps": 1,
|
| 974 |
+
"max_steps": 136,
|
| 975 |
+
"num_input_tokens_seen": 0,
|
| 976 |
+
"num_train_epochs": 17,
|
| 977 |
+
"save_steps": 10000,
|
| 978 |
+
"stateful_callbacks": {
|
| 979 |
+
"TrainerControl": {
|
| 980 |
+
"args": {
|
| 981 |
+
"should_epoch_stop": false,
|
| 982 |
+
"should_evaluate": false,
|
| 983 |
+
"should_log": false,
|
| 984 |
+
"should_save": true,
|
| 985 |
+
"should_training_stop": true
|
| 986 |
+
},
|
| 987 |
+
"attributes": {}
|
| 988 |
+
}
|
| 989 |
+
},
|
| 990 |
+
"total_flos": 483897984679936.0,
|
| 991 |
+
"train_batch_size": 1,
|
| 992 |
+
"trial_name": null,
|
| 993 |
+
"trial_params": null
|
| 994 |
+
}
|