Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +987 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,987 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 135,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.022222222222222223,
|
| 13 |
+
"grad_norm": 4.167962551116943,
|
| 14 |
+
"learning_rate": 4.000000000000001e-06,
|
| 15 |
+
"loss": 1.4459,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.044444444444444446,
|
| 20 |
+
"grad_norm": 4.384549140930176,
|
| 21 |
+
"learning_rate": 8.000000000000001e-06,
|
| 22 |
+
"loss": 1.4241,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.06666666666666667,
|
| 27 |
+
"grad_norm": 3.540264368057251,
|
| 28 |
+
"learning_rate": 1.2e-05,
|
| 29 |
+
"loss": 1.463,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.08888888888888889,
|
| 34 |
+
"grad_norm": 2.7690279483795166,
|
| 35 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 36 |
+
"loss": 1.3358,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.1111111111111111,
|
| 41 |
+
"grad_norm": 2.213526964187622,
|
| 42 |
+
"learning_rate": 2e-05,
|
| 43 |
+
"loss": 1.2881,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.13333333333333333,
|
| 48 |
+
"grad_norm": 2.0195415019989014,
|
| 49 |
+
"learning_rate": 1.9997080140801932e-05,
|
| 50 |
+
"loss": 1.2891,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.15555555555555556,
|
| 55 |
+
"grad_norm": 1.530819296836853,
|
| 56 |
+
"learning_rate": 1.998832226832327e-05,
|
| 57 |
+
"loss": 1.198,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.17777777777777778,
|
| 62 |
+
"grad_norm": 1.6055681705474854,
|
| 63 |
+
"learning_rate": 1.9973731496914914e-05,
|
| 64 |
+
"loss": 1.2906,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2,
|
| 69 |
+
"grad_norm": 1.5212966203689575,
|
| 70 |
+
"learning_rate": 1.995331634717649e-05,
|
| 71 |
+
"loss": 1.2806,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.2222222222222222,
|
| 76 |
+
"grad_norm": 1.3411786556243896,
|
| 77 |
+
"learning_rate": 1.992708874098054e-05,
|
| 78 |
+
"loss": 1.2865,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.24444444444444444,
|
| 83 |
+
"grad_norm": 1.575100302696228,
|
| 84 |
+
"learning_rate": 1.9895063994510512e-05,
|
| 85 |
+
"loss": 1.3072,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.26666666666666666,
|
| 90 |
+
"grad_norm": 1.2674833536148071,
|
| 91 |
+
"learning_rate": 1.985726080931651e-05,
|
| 92 |
+
"loss": 1.253,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.28888888888888886,
|
| 97 |
+
"grad_norm": 1.2866424322128296,
|
| 98 |
+
"learning_rate": 1.9813701261394136e-05,
|
| 99 |
+
"loss": 1.3403,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3111111111111111,
|
| 104 |
+
"grad_norm": 1.2735956907272339,
|
| 105 |
+
"learning_rate": 1.9764410788292724e-05,
|
| 106 |
+
"loss": 1.2685,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3333333333333333,
|
| 111 |
+
"grad_norm": 1.264608383178711,
|
| 112 |
+
"learning_rate": 1.9709418174260523e-05,
|
| 113 |
+
"loss": 1.2837,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.35555555555555557,
|
| 118 |
+
"grad_norm": 1.261456847190857,
|
| 119 |
+
"learning_rate": 1.9648755533435517e-05,
|
| 120 |
+
"loss": 1.1798,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.37777777777777777,
|
| 125 |
+
"grad_norm": 1.2654298543930054,
|
| 126 |
+
"learning_rate": 1.9582458291091664e-05,
|
| 127 |
+
"loss": 1.253,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4,
|
| 132 |
+
"grad_norm": 1.3421834707260132,
|
| 133 |
+
"learning_rate": 1.9510565162951538e-05,
|
| 134 |
+
"loss": 1.3176,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.4222222222222222,
|
| 139 |
+
"grad_norm": 1.3380440473556519,
|
| 140 |
+
"learning_rate": 1.9433118132577432e-05,
|
| 141 |
+
"loss": 1.2472,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.4444444444444444,
|
| 146 |
+
"grad_norm": 1.357919454574585,
|
| 147 |
+
"learning_rate": 1.9350162426854152e-05,
|
| 148 |
+
"loss": 1.3272,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.4666666666666667,
|
| 153 |
+
"grad_norm": 1.3505072593688965,
|
| 154 |
+
"learning_rate": 1.9261746489577767e-05,
|
| 155 |
+
"loss": 1.2913,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4888888888888889,
|
| 160 |
+
"grad_norm": 1.2780150175094604,
|
| 161 |
+
"learning_rate": 1.9167921953165827e-05,
|
| 162 |
+
"loss": 1.306,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.5111111111111111,
|
| 167 |
+
"grad_norm": 1.277637243270874,
|
| 168 |
+
"learning_rate": 1.9068743608505454e-05,
|
| 169 |
+
"loss": 1.2674,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.5333333333333333,
|
| 174 |
+
"grad_norm": 1.2741228342056274,
|
| 175 |
+
"learning_rate": 1.896426937295704e-05,
|
| 176 |
+
"loss": 1.3755,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.5555555555555556,
|
| 181 |
+
"grad_norm": 1.2366023063659668,
|
| 182 |
+
"learning_rate": 1.8854560256532098e-05,
|
| 183 |
+
"loss": 1.2629,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.5777777777777777,
|
| 188 |
+
"grad_norm": 1.1380127668380737,
|
| 189 |
+
"learning_rate": 1.873968032626518e-05,
|
| 190 |
+
"loss": 1.2493,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.6,
|
| 195 |
+
"grad_norm": 1.2305244207382202,
|
| 196 |
+
"learning_rate": 1.8619696668800494e-05,
|
| 197 |
+
"loss": 1.3068,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.6222222222222222,
|
| 202 |
+
"grad_norm": 1.2032471895217896,
|
| 203 |
+
"learning_rate": 1.8494679351215212e-05,
|
| 204 |
+
"loss": 1.2883,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.6444444444444445,
|
| 209 |
+
"grad_norm": 1.2741798162460327,
|
| 210 |
+
"learning_rate": 1.8364701380102267e-05,
|
| 211 |
+
"loss": 1.3237,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.6666666666666666,
|
| 216 |
+
"grad_norm": 1.2616007328033447,
|
| 217 |
+
"learning_rate": 1.8229838658936566e-05,
|
| 218 |
+
"loss": 1.3214,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.6888888888888889,
|
| 223 |
+
"grad_norm": 1.2144700288772583,
|
| 224 |
+
"learning_rate": 1.8090169943749477e-05,
|
| 225 |
+
"loss": 1.3056,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.7111111111111111,
|
| 230 |
+
"grad_norm": 1.2606492042541504,
|
| 231 |
+
"learning_rate": 1.7945776797137544e-05,
|
| 232 |
+
"loss": 1.2833,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7333333333333333,
|
| 237 |
+
"grad_norm": 1.2066459655761719,
|
| 238 |
+
"learning_rate": 1.7796743540632226e-05,
|
| 239 |
+
"loss": 1.2788,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7555555555555555,
|
| 244 |
+
"grad_norm": 1.284441351890564,
|
| 245 |
+
"learning_rate": 1.7643157205458483e-05,
|
| 246 |
+
"loss": 1.35,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.7777777777777778,
|
| 251 |
+
"grad_norm": 1.1287873983383179,
|
| 252 |
+
"learning_rate": 1.7485107481711014e-05,
|
| 253 |
+
"loss": 1.2327,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.8,
|
| 258 |
+
"grad_norm": 1.2790271043777466,
|
| 259 |
+
"learning_rate": 1.7322686665977738e-05,
|
| 260 |
+
"loss": 1.3085,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.8222222222222222,
|
| 265 |
+
"grad_norm": 1.2326244115829468,
|
| 266 |
+
"learning_rate": 1.715598960744121e-05,
|
| 267 |
+
"loss": 1.2799,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.8444444444444444,
|
| 272 |
+
"grad_norm": 1.2302358150482178,
|
| 273 |
+
"learning_rate": 1.6985113652489374e-05,
|
| 274 |
+
"loss": 1.3319,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.8666666666666667,
|
| 279 |
+
"grad_norm": 1.19487464427948,
|
| 280 |
+
"learning_rate": 1.6810158587867973e-05,
|
| 281 |
+
"loss": 1.2241,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.8888888888888888,
|
| 286 |
+
"grad_norm": 1.169329285621643,
|
| 287 |
+
"learning_rate": 1.6631226582407954e-05,
|
| 288 |
+
"loss": 1.2861,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.9111111111111111,
|
| 293 |
+
"grad_norm": 1.2315469980239868,
|
| 294 |
+
"learning_rate": 1.6448422127361707e-05,
|
| 295 |
+
"loss": 1.2595,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.9333333333333333,
|
| 300 |
+
"grad_norm": 1.2835931777954102,
|
| 301 |
+
"learning_rate": 1.626185197538314e-05,
|
| 302 |
+
"loss": 1.2831,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.9555555555555556,
|
| 307 |
+
"grad_norm": 1.2636137008666992,
|
| 308 |
+
"learning_rate": 1.6071625078187113e-05,
|
| 309 |
+
"loss": 1.242,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.9777777777777777,
|
| 314 |
+
"grad_norm": 1.2494031190872192,
|
| 315 |
+
"learning_rate": 1.5877852522924733e-05,
|
| 316 |
+
"loss": 1.2975,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.0,
|
| 321 |
+
"grad_norm": 1.2741174697875977,
|
| 322 |
+
"learning_rate": 1.568064746731156e-05,
|
| 323 |
+
"loss": 1.1633,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.0222222222222221,
|
| 328 |
+
"grad_norm": 1.650080919265747,
|
| 329 |
+
"learning_rate": 1.5480125073546705e-05,
|
| 330 |
+
"loss": 0.925,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.0444444444444445,
|
| 335 |
+
"grad_norm": 1.5442618131637573,
|
| 336 |
+
"learning_rate": 1.527640244106133e-05,
|
| 337 |
+
"loss": 0.9015,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.0666666666666667,
|
| 342 |
+
"grad_norm": 1.4545583724975586,
|
| 343 |
+
"learning_rate": 1.5069598538135905e-05,
|
| 344 |
+
"loss": 0.8485,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.0888888888888888,
|
| 349 |
+
"grad_norm": 1.3318345546722412,
|
| 350 |
+
"learning_rate": 1.485983413242606e-05,
|
| 351 |
+
"loss": 0.8942,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.1111111111111112,
|
| 356 |
+
"grad_norm": 1.353122353553772,
|
| 357 |
+
"learning_rate": 1.4647231720437687e-05,
|
| 358 |
+
"loss": 0.8117,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.1333333333333333,
|
| 363 |
+
"grad_norm": 1.383474588394165,
|
| 364 |
+
"learning_rate": 1.4431915455992416e-05,
|
| 365 |
+
"loss": 0.8371,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.1555555555555554,
|
| 370 |
+
"grad_norm": 1.4984554052352905,
|
| 371 |
+
"learning_rate": 1.4214011077725293e-05,
|
| 372 |
+
"loss": 0.7971,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.1777777777777778,
|
| 377 |
+
"grad_norm": 1.38037109375,
|
| 378 |
+
"learning_rate": 1.3993645835656955e-05,
|
| 379 |
+
"loss": 0.8353,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.2,
|
| 384 |
+
"grad_norm": 1.4161505699157715,
|
| 385 |
+
"learning_rate": 1.3770948416883205e-05,
|
| 386 |
+
"loss": 0.8498,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.2222222222222223,
|
| 391 |
+
"grad_norm": 1.277146339416504,
|
| 392 |
+
"learning_rate": 1.3546048870425356e-05,
|
| 393 |
+
"loss": 0.8477,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.2444444444444445,
|
| 398 |
+
"grad_norm": 1.1943016052246094,
|
| 399 |
+
"learning_rate": 1.3319078531285286e-05,
|
| 400 |
+
"loss": 0.7678,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.2666666666666666,
|
| 405 |
+
"grad_norm": 1.2254644632339478,
|
| 406 |
+
"learning_rate": 1.3090169943749475e-05,
|
| 407 |
+
"loss": 0.8474,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.2888888888888888,
|
| 412 |
+
"grad_norm": 1.3301903009414673,
|
| 413 |
+
"learning_rate": 1.2859456783986892e-05,
|
| 414 |
+
"loss": 0.8118,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.3111111111111111,
|
| 419 |
+
"grad_norm": 1.2812258005142212,
|
| 420 |
+
"learning_rate": 1.262707378198587e-05,
|
| 421 |
+
"loss": 0.8071,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.3333333333333333,
|
| 426 |
+
"grad_norm": 1.1742645502090454,
|
| 427 |
+
"learning_rate": 1.2393156642875579e-05,
|
| 428 |
+
"loss": 0.7705,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.3555555555555556,
|
| 433 |
+
"grad_norm": 1.2340425252914429,
|
| 434 |
+
"learning_rate": 1.2157841967678064e-05,
|
| 435 |
+
"loss": 0.7734,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.3777777777777778,
|
| 440 |
+
"grad_norm": 1.213952660560608,
|
| 441 |
+
"learning_rate": 1.1921267173537085e-05,
|
| 442 |
+
"loss": 0.7972,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.4,
|
| 447 |
+
"grad_norm": 1.3048670291900635,
|
| 448 |
+
"learning_rate": 1.1683570413470384e-05,
|
| 449 |
+
"loss": 0.7994,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.4222222222222223,
|
| 454 |
+
"grad_norm": 1.244718074798584,
|
| 455 |
+
"learning_rate": 1.1444890495692214e-05,
|
| 456 |
+
"loss": 0.8186,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.4444444444444444,
|
| 461 |
+
"grad_norm": 1.2908648252487183,
|
| 462 |
+
"learning_rate": 1.1205366802553231e-05,
|
| 463 |
+
"loss": 0.7909,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.4666666666666668,
|
| 468 |
+
"grad_norm": 1.230573058128357,
|
| 469 |
+
"learning_rate": 1.0965139209145153e-05,
|
| 470 |
+
"loss": 0.7782,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.488888888888889,
|
| 475 |
+
"grad_norm": 1.26985764503479,
|
| 476 |
+
"learning_rate": 1.0724348001617626e-05,
|
| 477 |
+
"loss": 0.8129,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.511111111111111,
|
| 482 |
+
"grad_norm": 1.1585290431976318,
|
| 483 |
+
"learning_rate": 1.0483133795255072e-05,
|
| 484 |
+
"loss": 0.8105,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.5333333333333332,
|
| 489 |
+
"grad_norm": 1.2600066661834717,
|
| 490 |
+
"learning_rate": 1.0241637452361323e-05,
|
| 491 |
+
"loss": 0.7895,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.5555555555555556,
|
| 496 |
+
"grad_norm": 1.273323655128479,
|
| 497 |
+
"learning_rate": 1e-05,
|
| 498 |
+
"loss": 0.7391,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.5777777777777777,
|
| 503 |
+
"grad_norm": 1.2754024267196655,
|
| 504 |
+
"learning_rate": 9.75836254763868e-06,
|
| 505 |
+
"loss": 0.7914,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.6,
|
| 510 |
+
"grad_norm": 1.2509865760803223,
|
| 511 |
+
"learning_rate": 9.516866204744932e-06,
|
| 512 |
+
"loss": 0.7721,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.6222222222222222,
|
| 517 |
+
"grad_norm": 1.3182200193405151,
|
| 518 |
+
"learning_rate": 9.275651998382377e-06,
|
| 519 |
+
"loss": 0.7834,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.6444444444444444,
|
| 524 |
+
"grad_norm": 1.2975807189941406,
|
| 525 |
+
"learning_rate": 9.034860790854848e-06,
|
| 526 |
+
"loss": 0.7786,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.6666666666666665,
|
| 531 |
+
"grad_norm": 1.2552589178085327,
|
| 532 |
+
"learning_rate": 8.79463319744677e-06,
|
| 533 |
+
"loss": 0.7629,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.6888888888888889,
|
| 538 |
+
"grad_norm": 1.299580693244934,
|
| 539 |
+
"learning_rate": 8.55510950430779e-06,
|
| 540 |
+
"loss": 0.801,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.7111111111111112,
|
| 545 |
+
"grad_norm": 1.223893642425537,
|
| 546 |
+
"learning_rate": 8.316429586529616e-06,
|
| 547 |
+
"loss": 0.7667,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.7333333333333334,
|
| 552 |
+
"grad_norm": 1.2119156122207642,
|
| 553 |
+
"learning_rate": 8.078732826462917e-06,
|
| 554 |
+
"loss": 0.7704,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.7555555555555555,
|
| 559 |
+
"grad_norm": 1.28359854221344,
|
| 560 |
+
"learning_rate": 7.84215803232194e-06,
|
| 561 |
+
"loss": 0.8196,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.7777777777777777,
|
| 566 |
+
"grad_norm": 1.2682065963745117,
|
| 567 |
+
"learning_rate": 7.606843357124426e-06,
|
| 568 |
+
"loss": 0.7964,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.8,
|
| 573 |
+
"grad_norm": 1.2014245986938477,
|
| 574 |
+
"learning_rate": 7.372926218014131e-06,
|
| 575 |
+
"loss": 0.7564,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.8222222222222222,
|
| 580 |
+
"grad_norm": 1.1668450832366943,
|
| 581 |
+
"learning_rate": 7.140543216013109e-06,
|
| 582 |
+
"loss": 0.7473,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.8444444444444446,
|
| 587 |
+
"grad_norm": 1.1506359577178955,
|
| 588 |
+
"learning_rate": 6.909830056250527e-06,
|
| 589 |
+
"loss": 0.7621,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.8666666666666667,
|
| 594 |
+
"grad_norm": 1.2430121898651123,
|
| 595 |
+
"learning_rate": 6.680921468714718e-06,
|
| 596 |
+
"loss": 0.7751,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.8888888888888888,
|
| 601 |
+
"grad_norm": 1.1884524822235107,
|
| 602 |
+
"learning_rate": 6.453951129574644e-06,
|
| 603 |
+
"loss": 0.7941,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.911111111111111,
|
| 608 |
+
"grad_norm": 1.1808528900146484,
|
| 609 |
+
"learning_rate": 6.229051583116796e-06,
|
| 610 |
+
"loss": 0.7918,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.9333333333333333,
|
| 615 |
+
"grad_norm": 2.18603253364563,
|
| 616 |
+
"learning_rate": 6.006354164343047e-06,
|
| 617 |
+
"loss": 0.8135,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.9555555555555557,
|
| 622 |
+
"grad_norm": 1.168459177017212,
|
| 623 |
+
"learning_rate": 5.785988922274711e-06,
|
| 624 |
+
"loss": 0.7734,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.9777777777777779,
|
| 629 |
+
"grad_norm": 1.1896721124649048,
|
| 630 |
+
"learning_rate": 5.5680845440075885e-06,
|
| 631 |
+
"loss": 0.7521,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.0,
|
| 636 |
+
"grad_norm": 1.2082970142364502,
|
| 637 |
+
"learning_rate": 5.352768279562315e-06,
|
| 638 |
+
"loss": 0.706,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.022222222222222,
|
| 643 |
+
"grad_norm": 1.7667112350463867,
|
| 644 |
+
"learning_rate": 5.14016586757394e-06,
|
| 645 |
+
"loss": 0.5277,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.0444444444444443,
|
| 650 |
+
"grad_norm": 1.5258979797363281,
|
| 651 |
+
"learning_rate": 4.930401461864099e-06,
|
| 652 |
+
"loss": 0.5515,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.066666666666667,
|
| 657 |
+
"grad_norm": 1.4529825448989868,
|
| 658 |
+
"learning_rate": 4.7235975589386715e-06,
|
| 659 |
+
"loss": 0.5588,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.088888888888889,
|
| 664 |
+
"grad_norm": 1.4381386041641235,
|
| 665 |
+
"learning_rate": 4.519874926453303e-06,
|
| 666 |
+
"loss": 0.4883,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.111111111111111,
|
| 671 |
+
"grad_norm": 1.3050932884216309,
|
| 672 |
+
"learning_rate": 4.319352532688444e-06,
|
| 673 |
+
"loss": 0.489,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.1333333333333333,
|
| 678 |
+
"grad_norm": 1.2284200191497803,
|
| 679 |
+
"learning_rate": 4.12214747707527e-06,
|
| 680 |
+
"loss": 0.5041,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.1555555555555554,
|
| 685 |
+
"grad_norm": 1.5800485610961914,
|
| 686 |
+
"learning_rate": 3.9283749218128885e-06,
|
| 687 |
+
"loss": 0.5215,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.1777777777777776,
|
| 692 |
+
"grad_norm": 1.7322880029678345,
|
| 693 |
+
"learning_rate": 3.738148024616863e-06,
|
| 694 |
+
"loss": 0.5004,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 2.2,
|
| 699 |
+
"grad_norm": 1.8849283456802368,
|
| 700 |
+
"learning_rate": 3.5515778726382967e-06,
|
| 701 |
+
"loss": 0.4636,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 2.2222222222222223,
|
| 706 |
+
"grad_norm": 1.875591516494751,
|
| 707 |
+
"learning_rate": 3.3687734175920505e-06,
|
| 708 |
+
"loss": 0.5078,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 2.2444444444444445,
|
| 713 |
+
"grad_norm": 1.6613143682479858,
|
| 714 |
+
"learning_rate": 3.1898414121320277e-06,
|
| 715 |
+
"loss": 0.4932,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 2.2666666666666666,
|
| 720 |
+
"grad_norm": 1.3411990404129028,
|
| 721 |
+
"learning_rate": 3.0148863475106315e-06,
|
| 722 |
+
"loss": 0.4634,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.2888888888888888,
|
| 727 |
+
"grad_norm": 1.3103381395339966,
|
| 728 |
+
"learning_rate": 2.8440103925587904e-06,
|
| 729 |
+
"loss": 0.4846,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 2.311111111111111,
|
| 734 |
+
"grad_norm": 1.2642971277236938,
|
| 735 |
+
"learning_rate": 2.6773133340222677e-06,
|
| 736 |
+
"loss": 0.4878,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 2.3333333333333335,
|
| 741 |
+
"grad_norm": 1.2269350290298462,
|
| 742 |
+
"learning_rate": 2.514892518288988e-06,
|
| 743 |
+
"loss": 0.467,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 2.3555555555555556,
|
| 748 |
+
"grad_norm": 1.2345279455184937,
|
| 749 |
+
"learning_rate": 2.3568427945415163e-06,
|
| 750 |
+
"loss": 0.4638,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 2.3777777777777778,
|
| 755 |
+
"grad_norm": 1.2748271226882935,
|
| 756 |
+
"learning_rate": 2.2032564593677773e-06,
|
| 757 |
+
"loss": 0.5024,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 2.4,
|
| 762 |
+
"grad_norm": 1.2327388525009155,
|
| 763 |
+
"learning_rate": 2.0542232028624585e-06,
|
| 764 |
+
"loss": 0.4876,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 2.422222222222222,
|
| 769 |
+
"grad_norm": 1.2413166761398315,
|
| 770 |
+
"learning_rate": 1.9098300562505266e-06,
|
| 771 |
+
"loss": 0.4937,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 2.4444444444444446,
|
| 776 |
+
"grad_norm": 1.301472544670105,
|
| 777 |
+
"learning_rate": 1.7701613410634367e-06,
|
| 778 |
+
"loss": 0.4663,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 2.466666666666667,
|
| 783 |
+
"grad_norm": 1.3159891366958618,
|
| 784 |
+
"learning_rate": 1.6352986198977327e-06,
|
| 785 |
+
"loss": 0.487,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 2.488888888888889,
|
| 790 |
+
"grad_norm": 1.2749731540679932,
|
| 791 |
+
"learning_rate": 1.5053206487847916e-06,
|
| 792 |
+
"loss": 0.4969,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 2.511111111111111,
|
| 797 |
+
"grad_norm": 1.2180869579315186,
|
| 798 |
+
"learning_rate": 1.3803033311995072e-06,
|
| 799 |
+
"loss": 0.4907,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 2.533333333333333,
|
| 804 |
+
"grad_norm": 1.21835196018219,
|
| 805 |
+
"learning_rate": 1.2603196737348211e-06,
|
| 806 |
+
"loss": 0.4819,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 2.5555555555555554,
|
| 811 |
+
"grad_norm": 1.2181463241577148,
|
| 812 |
+
"learning_rate": 1.1454397434679022e-06,
|
| 813 |
+
"loss": 0.4844,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 2.5777777777777775,
|
| 818 |
+
"grad_norm": 1.2931818962097168,
|
| 819 |
+
"learning_rate": 1.0357306270429623e-06,
|
| 820 |
+
"loss": 0.4904,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 2.6,
|
| 825 |
+
"grad_norm": 1.3804171085357666,
|
| 826 |
+
"learning_rate": 9.312563914945461e-07,
|
| 827 |
+
"loss": 0.4899,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 2.6222222222222222,
|
| 832 |
+
"grad_norm": 1.2506023645401,
|
| 833 |
+
"learning_rate": 8.320780468341761e-07,
|
| 834 |
+
"loss": 0.489,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 2.6444444444444444,
|
| 839 |
+
"grad_norm": 1.3200658559799194,
|
| 840 |
+
"learning_rate": 7.382535104222366e-07,
|
| 841 |
+
"loss": 0.4805,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 2.6666666666666665,
|
| 846 |
+
"grad_norm": 1.2870758771896362,
|
| 847 |
+
"learning_rate": 6.498375731458529e-07,
|
| 848 |
+
"loss": 0.4618,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 2.688888888888889,
|
| 853 |
+
"grad_norm": 1.3506563901901245,
|
| 854 |
+
"learning_rate": 5.668818674225684e-07,
|
| 855 |
+
"loss": 0.4805,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 2.7111111111111112,
|
| 860 |
+
"grad_norm": 1.2517269849777222,
|
| 861 |
+
"learning_rate": 4.894348370484648e-07,
|
| 862 |
+
"loss": 0.501,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 2.7333333333333334,
|
| 867 |
+
"grad_norm": 1.3007408380508423,
|
| 868 |
+
"learning_rate": 4.1754170890833777e-07,
|
| 869 |
+
"loss": 0.4736,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 2.7555555555555555,
|
| 874 |
+
"grad_norm": 1.4205576181411743,
|
| 875 |
+
"learning_rate": 3.5124446656448654e-07,
|
| 876 |
+
"loss": 0.4876,
|
| 877 |
+
"step": 124
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 2.7777777777777777,
|
| 881 |
+
"grad_norm": 1.3209325075149536,
|
| 882 |
+
"learning_rate": 2.905818257394799e-07,
|
| 883 |
+
"loss": 0.4924,
|
| 884 |
+
"step": 125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 2.8,
|
| 888 |
+
"grad_norm": 1.3586435317993164,
|
| 889 |
+
"learning_rate": 2.355892117072789e-07,
|
| 890 |
+
"loss": 0.5009,
|
| 891 |
+
"step": 126
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 2.822222222222222,
|
| 895 |
+
"grad_norm": 1.3130888938903809,
|
| 896 |
+
"learning_rate": 1.8629873860586567e-07,
|
| 897 |
+
"loss": 0.4595,
|
| 898 |
+
"step": 127
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 2.8444444444444446,
|
| 902 |
+
"grad_norm": 1.2706563472747803,
|
| 903 |
+
"learning_rate": 1.4273919068349184e-07,
|
| 904 |
+
"loss": 0.5057,
|
| 905 |
+
"step": 128
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 2.8666666666666667,
|
| 909 |
+
"grad_norm": 1.2369240522384644,
|
| 910 |
+
"learning_rate": 1.0493600548948879e-07,
|
| 911 |
+
"loss": 0.4822,
|
| 912 |
+
"step": 129
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 2.888888888888889,
|
| 916 |
+
"grad_norm": 1.3894768953323364,
|
| 917 |
+
"learning_rate": 7.291125901946027e-08,
|
| 918 |
+
"loss": 0.4819,
|
| 919 |
+
"step": 130
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 2.911111111111111,
|
| 923 |
+
"grad_norm": 1.3319177627563477,
|
| 924 |
+
"learning_rate": 4.6683652823513725e-08,
|
| 925 |
+
"loss": 0.5007,
|
| 926 |
+
"step": 131
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 2.9333333333333336,
|
| 930 |
+
"grad_norm": 1.2581626176834106,
|
| 931 |
+
"learning_rate": 2.6268503085089547e-08,
|
| 932 |
+
"loss": 0.4779,
|
| 933 |
+
"step": 132
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 2.9555555555555557,
|
| 937 |
+
"grad_norm": 1.266819715499878,
|
| 938 |
+
"learning_rate": 1.1677731676733584e-08,
|
| 939 |
+
"loss": 0.4792,
|
| 940 |
+
"step": 133
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 2.977777777777778,
|
| 944 |
+
"grad_norm": 1.2672613859176636,
|
| 945 |
+
"learning_rate": 2.9198591980705847e-09,
|
| 946 |
+
"loss": 0.4984,
|
| 947 |
+
"step": 134
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 3.0,
|
| 951 |
+
"grad_norm": 1.108398675918579,
|
| 952 |
+
"learning_rate": 0.0,
|
| 953 |
+
"loss": 0.4556,
|
| 954 |
+
"step": 135
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 3.0,
|
| 958 |
+
"step": 135,
|
| 959 |
+
"total_flos": 2.880049111724196e+17,
|
| 960 |
+
"train_loss": 0.8609414833563346,
|
| 961 |
+
"train_runtime": 957.8082,
|
| 962 |
+
"train_samples_per_second": 17.866,
|
| 963 |
+
"train_steps_per_second": 0.141
|
| 964 |
+
}
|
| 965 |
+
],
|
| 966 |
+
"logging_steps": 1,
|
| 967 |
+
"max_steps": 135,
|
| 968 |
+
"num_input_tokens_seen": 0,
|
| 969 |
+
"num_train_epochs": 3,
|
| 970 |
+
"save_steps": 0,
|
| 971 |
+
"stateful_callbacks": {
|
| 972 |
+
"TrainerControl": {
|
| 973 |
+
"args": {
|
| 974 |
+
"should_epoch_stop": false,
|
| 975 |
+
"should_evaluate": false,
|
| 976 |
+
"should_log": false,
|
| 977 |
+
"should_save": true,
|
| 978 |
+
"should_training_stop": true
|
| 979 |
+
},
|
| 980 |
+
"attributes": {}
|
| 981 |
+
}
|
| 982 |
+
},
|
| 983 |
+
"total_flos": 2.880049111724196e+17,
|
| 984 |
+
"train_batch_size": 32,
|
| 985 |
+
"trial_name": null,
|
| 986 |
+
"trial_params": null
|
| 987 |
+
}
|