Update model
Browse files- trainer_state.json +901 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,901 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 124,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.008064516129032258,
|
| 13 |
+
"grad_norm": 20.875,
|
| 14 |
+
"learning_rate": 1.25e-06,
|
| 15 |
+
"loss": 8.243,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.016129032258064516,
|
| 20 |
+
"grad_norm": 20.375,
|
| 21 |
+
"learning_rate": 2.5e-06,
|
| 22 |
+
"loss": 8.1865,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.024193548387096774,
|
| 27 |
+
"grad_norm": 20.0,
|
| 28 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 29 |
+
"loss": 8.0673,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03225806451612903,
|
| 34 |
+
"grad_norm": 20.625,
|
| 35 |
+
"learning_rate": 5e-06,
|
| 36 |
+
"loss": 8.0871,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04032258064516129,
|
| 41 |
+
"grad_norm": 20.75,
|
| 42 |
+
"learning_rate": 4.999143312438893e-06,
|
| 43 |
+
"loss": 8.0484,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04838709677419355,
|
| 48 |
+
"grad_norm": 20.5,
|
| 49 |
+
"learning_rate": 4.9965738368864345e-06,
|
| 50 |
+
"loss": 7.7429,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.056451612903225805,
|
| 55 |
+
"grad_norm": 18.0,
|
| 56 |
+
"learning_rate": 4.992293334332821e-06,
|
| 57 |
+
"loss": 8.0224,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.06451612903225806,
|
| 62 |
+
"grad_norm": 18.0,
|
| 63 |
+
"learning_rate": 4.986304738420684e-06,
|
| 64 |
+
"loss": 7.6741,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.07258064516129033,
|
| 69 |
+
"grad_norm": 18.625,
|
| 70 |
+
"learning_rate": 4.978612153434527e-06,
|
| 71 |
+
"loss": 7.7528,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.08064516129032258,
|
| 76 |
+
"grad_norm": 17.875,
|
| 77 |
+
"learning_rate": 4.9692208514878445e-06,
|
| 78 |
+
"loss": 7.5786,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.08870967741935484,
|
| 83 |
+
"grad_norm": 17.625,
|
| 84 |
+
"learning_rate": 4.958137268909887e-06,
|
| 85 |
+
"loss": 7.5182,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.0967741935483871,
|
| 90 |
+
"grad_norm": 16.0,
|
| 91 |
+
"learning_rate": 4.9453690018345144e-06,
|
| 92 |
+
"loss": 7.5339,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.10483870967741936,
|
| 97 |
+
"grad_norm": 16.5,
|
| 98 |
+
"learning_rate": 4.930924800994192e-06,
|
| 99 |
+
"loss": 7.5517,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.11290322580645161,
|
| 104 |
+
"grad_norm": 16.375,
|
| 105 |
+
"learning_rate": 4.914814565722671e-06,
|
| 106 |
+
"loss": 7.3944,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.12096774193548387,
|
| 111 |
+
"grad_norm": 15.6875,
|
| 112 |
+
"learning_rate": 4.897049337170483e-06,
|
| 113 |
+
"loss": 7.1319,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.12903225806451613,
|
| 118 |
+
"grad_norm": 15.125,
|
| 119 |
+
"learning_rate": 4.8776412907378845e-06,
|
| 120 |
+
"loss": 7.0313,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.13709677419354838,
|
| 125 |
+
"grad_norm": 15.3125,
|
| 126 |
+
"learning_rate": 4.856603727730446e-06,
|
| 127 |
+
"loss": 6.9869,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.14516129032258066,
|
| 132 |
+
"grad_norm": 14.125,
|
| 133 |
+
"learning_rate": 4.833951066243004e-06,
|
| 134 |
+
"loss": 7.131,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1532258064516129,
|
| 139 |
+
"grad_norm": 13.0625,
|
| 140 |
+
"learning_rate": 4.809698831278217e-06,
|
| 141 |
+
"loss": 6.9202,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.16129032258064516,
|
| 146 |
+
"grad_norm": 13.625,
|
| 147 |
+
"learning_rate": 4.783863644106502e-06,
|
| 148 |
+
"loss": 6.9279,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.1693548387096774,
|
| 153 |
+
"grad_norm": 13.0,
|
| 154 |
+
"learning_rate": 4.7564632108746524e-06,
|
| 155 |
+
"loss": 6.6326,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1774193548387097,
|
| 160 |
+
"grad_norm": 12.875,
|
| 161 |
+
"learning_rate": 4.72751631047092e-06,
|
| 162 |
+
"loss": 6.7503,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.18548387096774194,
|
| 167 |
+
"grad_norm": 12.25,
|
| 168 |
+
"learning_rate": 4.697042781654913e-06,
|
| 169 |
+
"loss": 6.6069,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1935483870967742,
|
| 174 |
+
"grad_norm": 10.0,
|
| 175 |
+
"learning_rate": 4.665063509461098e-06,
|
| 176 |
+
"loss": 6.4378,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.20161290322580644,
|
| 181 |
+
"grad_norm": 10.125,
|
| 182 |
+
"learning_rate": 4.631600410885231e-06,
|
| 183 |
+
"loss": 6.6957,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.20967741935483872,
|
| 188 |
+
"grad_norm": 9.75,
|
| 189 |
+
"learning_rate": 4.596676419863561e-06,
|
| 190 |
+
"loss": 6.4625,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.21774193548387097,
|
| 195 |
+
"grad_norm": 10.125,
|
| 196 |
+
"learning_rate": 4.560315471555039e-06,
|
| 197 |
+
"loss": 6.5304,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.22580645161290322,
|
| 202 |
+
"grad_norm": 9.9375,
|
| 203 |
+
"learning_rate": 4.522542485937369e-06,
|
| 204 |
+
"loss": 6.3632,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.23387096774193547,
|
| 209 |
+
"grad_norm": 8.75,
|
| 210 |
+
"learning_rate": 4.4833833507280884e-06,
|
| 211 |
+
"loss": 6.3071,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.24193548387096775,
|
| 216 |
+
"grad_norm": 8.5,
|
| 217 |
+
"learning_rate": 4.442864903642428e-06,
|
| 218 |
+
"loss": 6.428,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.25,
|
| 223 |
+
"grad_norm": 9.125,
|
| 224 |
+
"learning_rate": 4.401014914000078e-06,
|
| 225 |
+
"loss": 6.3937,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.25806451612903225,
|
| 230 |
+
"grad_norm": 8.375,
|
| 231 |
+
"learning_rate": 4.357862063693486e-06,
|
| 232 |
+
"loss": 6.4951,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2661290322580645,
|
| 237 |
+
"grad_norm": 7.625,
|
| 238 |
+
"learning_rate": 4.313435927530719e-06,
|
| 239 |
+
"loss": 6.1423,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.27419354838709675,
|
| 244 |
+
"grad_norm": 8.125,
|
| 245 |
+
"learning_rate": 4.267766952966369e-06,
|
| 246 |
+
"loss": 6.0947,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.28225806451612906,
|
| 251 |
+
"grad_norm": 7.75,
|
| 252 |
+
"learning_rate": 4.220886439234385e-06,
|
| 253 |
+
"loss": 6.1428,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2903225806451613,
|
| 258 |
+
"grad_norm": 7.375,
|
| 259 |
+
"learning_rate": 4.172826515897146e-06,
|
| 260 |
+
"loss": 6.1118,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.29838709677419356,
|
| 265 |
+
"grad_norm": 7.4375,
|
| 266 |
+
"learning_rate": 4.123620120825459e-06,
|
| 267 |
+
"loss": 5.9917,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3064516129032258,
|
| 272 |
+
"grad_norm": 7.53125,
|
| 273 |
+
"learning_rate": 4.073300977624594e-06,
|
| 274 |
+
"loss": 5.7956,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.31451612903225806,
|
| 279 |
+
"grad_norm": 6.96875,
|
| 280 |
+
"learning_rate": 4.021903572521802e-06,
|
| 281 |
+
"loss": 5.8347,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.3225806451612903,
|
| 286 |
+
"grad_norm": 7.71875,
|
| 287 |
+
"learning_rate": 3.969463130731183e-06,
|
| 288 |
+
"loss": 6.1252,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.33064516129032256,
|
| 293 |
+
"grad_norm": 6.8125,
|
| 294 |
+
"learning_rate": 3.916015592312083e-06,
|
| 295 |
+
"loss": 6.0318,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3387096774193548,
|
| 300 |
+
"grad_norm": 6.46875,
|
| 301 |
+
"learning_rate": 3.861597587537568e-06,
|
| 302 |
+
"loss": 5.902,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.3467741935483871,
|
| 307 |
+
"grad_norm": 6.5,
|
| 308 |
+
"learning_rate": 3.806246411789872e-06,
|
| 309 |
+
"loss": 5.9344,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.3548387096774194,
|
| 314 |
+
"grad_norm": 6.25,
|
| 315 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 316 |
+
"loss": 5.7523,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3629032258064516,
|
| 321 |
+
"grad_norm": 6.46875,
|
| 322 |
+
"learning_rate": 3.6928969006490212e-06,
|
| 323 |
+
"loss": 5.9654,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.3709677419354839,
|
| 328 |
+
"grad_norm": 6.3125,
|
| 329 |
+
"learning_rate": 3.634976249348867e-06,
|
| 330 |
+
"loss": 5.8617,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.3790322580645161,
|
| 335 |
+
"grad_norm": 6.09375,
|
| 336 |
+
"learning_rate": 3.5762777420207382e-06,
|
| 337 |
+
"loss": 5.7341,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3870967741935484,
|
| 342 |
+
"grad_norm": 5.71875,
|
| 343 |
+
"learning_rate": 3.516841607689501e-06,
|
| 344 |
+
"loss": 5.896,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3951612903225806,
|
| 349 |
+
"grad_norm": 5.625,
|
| 350 |
+
"learning_rate": 3.4567085809127247e-06,
|
| 351 |
+
"loss": 5.8032,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4032258064516129,
|
| 356 |
+
"grad_norm": 5.9375,
|
| 357 |
+
"learning_rate": 3.39591987386325e-06,
|
| 358 |
+
"loss": 5.8156,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.4112903225806452,
|
| 363 |
+
"grad_norm": 6.1875,
|
| 364 |
+
"learning_rate": 3.3345171480844275e-06,
|
| 365 |
+
"loss": 5.7496,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.41935483870967744,
|
| 370 |
+
"grad_norm": 5.90625,
|
| 371 |
+
"learning_rate": 3.272542485937369e-06,
|
| 372 |
+
"loss": 5.7032,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.4274193548387097,
|
| 377 |
+
"grad_norm": 5.6875,
|
| 378 |
+
"learning_rate": 3.2100383617598075e-06,
|
| 379 |
+
"loss": 5.688,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.43548387096774194,
|
| 384 |
+
"grad_norm": 7.0625,
|
| 385 |
+
"learning_rate": 3.147047612756302e-06,
|
| 386 |
+
"loss": 5.9719,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.4435483870967742,
|
| 391 |
+
"grad_norm": 5.09375,
|
| 392 |
+
"learning_rate": 3.0836134096397642e-06,
|
| 393 |
+
"loss": 5.8026,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.45161290322580644,
|
| 398 |
+
"grad_norm": 5.21875,
|
| 399 |
+
"learning_rate": 3.019779227044398e-06,
|
| 400 |
+
"loss": 5.5754,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4596774193548387,
|
| 405 |
+
"grad_norm": 5.5625,
|
| 406 |
+
"learning_rate": 2.9555888137303695e-06,
|
| 407 |
+
"loss": 5.8139,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.46774193548387094,
|
| 412 |
+
"grad_norm": 5.46875,
|
| 413 |
+
"learning_rate": 2.8910861626005774e-06,
|
| 414 |
+
"loss": 5.6032,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.47580645161290325,
|
| 419 |
+
"grad_norm": 5.59375,
|
| 420 |
+
"learning_rate": 2.82631548055013e-06,
|
| 421 |
+
"loss": 5.6906,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4838709677419355,
|
| 426 |
+
"grad_norm": 5.40625,
|
| 427 |
+
"learning_rate": 2.761321158169134e-06,
|
| 428 |
+
"loss": 5.7467,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.49193548387096775,
|
| 433 |
+
"grad_norm": 5.875,
|
| 434 |
+
"learning_rate": 2.696147739319613e-06,
|
| 435 |
+
"loss": 5.9122,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.5,
|
| 440 |
+
"grad_norm": 5.5,
|
| 441 |
+
"learning_rate": 2.6308398906073603e-06,
|
| 442 |
+
"loss": 5.6152,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.5080645161290323,
|
| 447 |
+
"grad_norm": 5.03125,
|
| 448 |
+
"learning_rate": 2.5654423707696834e-06,
|
| 449 |
+
"loss": 5.8207,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5161290322580645,
|
| 454 |
+
"grad_norm": 6.1875,
|
| 455 |
+
"learning_rate": 2.5e-06,
|
| 456 |
+
"loss": 5.8129,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5241935483870968,
|
| 461 |
+
"grad_norm": 5.3125,
|
| 462 |
+
"learning_rate": 2.434557629230318e-06,
|
| 463 |
+
"loss": 5.7729,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.532258064516129,
|
| 468 |
+
"grad_norm": 5.1875,
|
| 469 |
+
"learning_rate": 2.3691601093926406e-06,
|
| 470 |
+
"loss": 5.5761,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5403225806451613,
|
| 475 |
+
"grad_norm": 5.4375,
|
| 476 |
+
"learning_rate": 2.3038522606803882e-06,
|
| 477 |
+
"loss": 5.6013,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5483870967741935,
|
| 482 |
+
"grad_norm": 4.96875,
|
| 483 |
+
"learning_rate": 2.238678841830867e-06,
|
| 484 |
+
"loss": 5.5553,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5564516129032258,
|
| 489 |
+
"grad_norm": 5.0625,
|
| 490 |
+
"learning_rate": 2.173684519449872e-06,
|
| 491 |
+
"loss": 5.5334,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.5645161290322581,
|
| 496 |
+
"grad_norm": 4.78125,
|
| 497 |
+
"learning_rate": 2.1089138373994226e-06,
|
| 498 |
+
"loss": 5.5125,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.5725806451612904,
|
| 503 |
+
"grad_norm": 5.34375,
|
| 504 |
+
"learning_rate": 2.0444111862696313e-06,
|
| 505 |
+
"loss": 5.704,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.5806451612903226,
|
| 510 |
+
"grad_norm": 5.03125,
|
| 511 |
+
"learning_rate": 1.9802207729556023e-06,
|
| 512 |
+
"loss": 5.6541,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.5887096774193549,
|
| 517 |
+
"grad_norm": 4.84375,
|
| 518 |
+
"learning_rate": 1.9163865903602374e-06,
|
| 519 |
+
"loss": 5.5812,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5967741935483871,
|
| 524 |
+
"grad_norm": 5.21875,
|
| 525 |
+
"learning_rate": 1.852952387243698e-06,
|
| 526 |
+
"loss": 5.597,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.6048387096774194,
|
| 531 |
+
"grad_norm": 5.0625,
|
| 532 |
+
"learning_rate": 1.7899616382401935e-06,
|
| 533 |
+
"loss": 5.692,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.6129032258064516,
|
| 538 |
+
"grad_norm": 4.75,
|
| 539 |
+
"learning_rate": 1.7274575140626318e-06,
|
| 540 |
+
"loss": 5.6938,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.6209677419354839,
|
| 545 |
+
"grad_norm": 4.96875,
|
| 546 |
+
"learning_rate": 1.665482851915573e-06,
|
| 547 |
+
"loss": 5.5474,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.6290322580645161,
|
| 552 |
+
"grad_norm": 4.84375,
|
| 553 |
+
"learning_rate": 1.6040801261367494e-06,
|
| 554 |
+
"loss": 5.681,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.6370967741935484,
|
| 559 |
+
"grad_norm": 4.875,
|
| 560 |
+
"learning_rate": 1.5432914190872757e-06,
|
| 561 |
+
"loss": 5.6655,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6451612903225806,
|
| 566 |
+
"grad_norm": 4.8125,
|
| 567 |
+
"learning_rate": 1.4831583923105e-06,
|
| 568 |
+
"loss": 5.3145,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.6532258064516129,
|
| 573 |
+
"grad_norm": 4.59375,
|
| 574 |
+
"learning_rate": 1.4237222579792618e-06,
|
| 575 |
+
"loss": 5.5851,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.6612903225806451,
|
| 580 |
+
"grad_norm": 5.5625,
|
| 581 |
+
"learning_rate": 1.3650237506511333e-06,
|
| 582 |
+
"loss": 5.6114,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.6693548387096774,
|
| 587 |
+
"grad_norm": 5.78125,
|
| 588 |
+
"learning_rate": 1.307103099350979e-06,
|
| 589 |
+
"loss": 5.7918,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.6774193548387096,
|
| 594 |
+
"grad_norm": 6.46875,
|
| 595 |
+
"learning_rate": 1.2500000000000007e-06,
|
| 596 |
+
"loss": 5.6083,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.6854838709677419,
|
| 601 |
+
"grad_norm": 5.78125,
|
| 602 |
+
"learning_rate": 1.193753588210128e-06,
|
| 603 |
+
"loss": 5.6588,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.6935483870967742,
|
| 608 |
+
"grad_norm": 4.6875,
|
| 609 |
+
"learning_rate": 1.1384024124624324e-06,
|
| 610 |
+
"loss": 5.552,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7016129032258065,
|
| 615 |
+
"grad_norm": 4.9375,
|
| 616 |
+
"learning_rate": 1.0839844076879186e-06,
|
| 617 |
+
"loss": 5.4991,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7096774193548387,
|
| 622 |
+
"grad_norm": 4.4375,
|
| 623 |
+
"learning_rate": 1.0305368692688175e-06,
|
| 624 |
+
"loss": 5.6138,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.717741935483871,
|
| 629 |
+
"grad_norm": 4.90625,
|
| 630 |
+
"learning_rate": 9.780964274781984e-07,
|
| 631 |
+
"loss": 5.5677,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.7258064516129032,
|
| 636 |
+
"grad_norm": 5.125,
|
| 637 |
+
"learning_rate": 9.266990223754069e-07,
|
| 638 |
+
"loss": 5.5496,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.7338709677419355,
|
| 643 |
+
"grad_norm": 4.9375,
|
| 644 |
+
"learning_rate": 8.763798791745413e-07,
|
| 645 |
+
"loss": 5.6754,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.7419354838709677,
|
| 650 |
+
"grad_norm": 4.90625,
|
| 651 |
+
"learning_rate": 8.271734841028553e-07,
|
| 652 |
+
"loss": 5.3941,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.75,
|
| 657 |
+
"grad_norm": 4.65625,
|
| 658 |
+
"learning_rate": 7.791135607656147e-07,
|
| 659 |
+
"loss": 5.5303,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.7580645161290323,
|
| 664 |
+
"grad_norm": 4.625,
|
| 665 |
+
"learning_rate": 7.322330470336314e-07,
|
| 666 |
+
"loss": 5.5491,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.7661290322580645,
|
| 671 |
+
"grad_norm": 5.9375,
|
| 672 |
+
"learning_rate": 6.865640724692815e-07,
|
| 673 |
+
"loss": 5.7337,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7741935483870968,
|
| 678 |
+
"grad_norm": 4.9375,
|
| 679 |
+
"learning_rate": 6.421379363065142e-07,
|
| 680 |
+
"loss": 5.4986,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.782258064516129,
|
| 685 |
+
"grad_norm": 4.84375,
|
| 686 |
+
"learning_rate": 5.989850859999227e-07,
|
| 687 |
+
"loss": 5.5802,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.7903225806451613,
|
| 692 |
+
"grad_norm": 5.09375,
|
| 693 |
+
"learning_rate": 5.571350963575728e-07,
|
| 694 |
+
"loss": 5.6567,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.7983870967741935,
|
| 699 |
+
"grad_norm": 5.6875,
|
| 700 |
+
"learning_rate": 5.166166492719124e-07,
|
| 701 |
+
"loss": 5.6235,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.8064516129032258,
|
| 706 |
+
"grad_norm": 4.75,
|
| 707 |
+
"learning_rate": 4.774575140626317e-07,
|
| 708 |
+
"loss": 5.6329,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.8145161290322581,
|
| 713 |
+
"grad_norm": 4.96875,
|
| 714 |
+
"learning_rate": 4.396845284449608e-07,
|
| 715 |
+
"loss": 5.5273,
|
| 716 |
+
"step": 101
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.8225806451612904,
|
| 720 |
+
"grad_norm": 4.625,
|
| 721 |
+
"learning_rate": 4.033235801364402e-07,
|
| 722 |
+
"loss": 5.4076,
|
| 723 |
+
"step": 102
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.8306451612903226,
|
| 727 |
+
"grad_norm": 4.5,
|
| 728 |
+
"learning_rate": 3.683995891147696e-07,
|
| 729 |
+
"loss": 5.5664,
|
| 730 |
+
"step": 103
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.8387096774193549,
|
| 734 |
+
"grad_norm": 5.71875,
|
| 735 |
+
"learning_rate": 3.3493649053890325e-07,
|
| 736 |
+
"loss": 5.6994,
|
| 737 |
+
"step": 104
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.8467741935483871,
|
| 741 |
+
"grad_norm": 5.46875,
|
| 742 |
+
"learning_rate": 3.0295721834508686e-07,
|
| 743 |
+
"loss": 5.6093,
|
| 744 |
+
"step": 105
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.8548387096774194,
|
| 748 |
+
"grad_norm": 4.65625,
|
| 749 |
+
"learning_rate": 2.7248368952908055e-07,
|
| 750 |
+
"loss": 5.4977,
|
| 751 |
+
"step": 106
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.8629032258064516,
|
| 755 |
+
"grad_norm": 5.0,
|
| 756 |
+
"learning_rate": 2.43536789125349e-07,
|
| 757 |
+
"loss": 5.5823,
|
| 758 |
+
"step": 107
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.8709677419354839,
|
| 762 |
+
"grad_norm": 7.625,
|
| 763 |
+
"learning_rate": 2.1613635589349756e-07,
|
| 764 |
+
"loss": 5.5065,
|
| 765 |
+
"step": 108
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.8790322580645161,
|
| 769 |
+
"grad_norm": 5.0625,
|
| 770 |
+
"learning_rate": 1.9030116872178317e-07,
|
| 771 |
+
"loss": 5.5458,
|
| 772 |
+
"step": 109
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.8870967741935484,
|
| 776 |
+
"grad_norm": 5.375,
|
| 777 |
+
"learning_rate": 1.6604893375699594e-07,
|
| 778 |
+
"loss": 5.5587,
|
| 779 |
+
"step": 110
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.8951612903225806,
|
| 783 |
+
"grad_norm": 5.25,
|
| 784 |
+
"learning_rate": 1.4339627226955394e-07,
|
| 785 |
+
"loss": 5.6694,
|
| 786 |
+
"step": 111
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.9032258064516129,
|
| 790 |
+
"grad_norm": 5.28125,
|
| 791 |
+
"learning_rate": 1.223587092621162e-07,
|
| 792 |
+
"loss": 5.5847,
|
| 793 |
+
"step": 112
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9112903225806451,
|
| 797 |
+
"grad_norm": 4.6875,
|
| 798 |
+
"learning_rate": 1.0295066282951738e-07,
|
| 799 |
+
"loss": 5.5174,
|
| 800 |
+
"step": 113
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.9193548387096774,
|
| 804 |
+
"grad_norm": 4.75,
|
| 805 |
+
"learning_rate": 8.518543427732951e-08,
|
| 806 |
+
"loss": 5.6295,
|
| 807 |
+
"step": 114
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.9274193548387096,
|
| 811 |
+
"grad_norm": 5.375,
|
| 812 |
+
"learning_rate": 6.907519900580862e-08,
|
| 813 |
+
"loss": 5.5846,
|
| 814 |
+
"step": 115
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.9354838709677419,
|
| 818 |
+
"grad_norm": 4.875,
|
| 819 |
+
"learning_rate": 5.463099816548578e-08,
|
| 820 |
+
"loss": 5.5643,
|
| 821 |
+
"step": 116
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.9435483870967742,
|
| 825 |
+
"grad_norm": 4.59375,
|
| 826 |
+
"learning_rate": 4.186273109011374e-08,
|
| 827 |
+
"loss": 5.5587,
|
| 828 |
+
"step": 117
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.9516129032258065,
|
| 832 |
+
"grad_norm": 4.9375,
|
| 833 |
+
"learning_rate": 3.077914851215585e-08,
|
| 834 |
+
"loss": 5.7314,
|
| 835 |
+
"step": 118
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9596774193548387,
|
| 839 |
+
"grad_norm": 5.0,
|
| 840 |
+
"learning_rate": 2.1387846565474047e-08,
|
| 841 |
+
"loss": 5.5992,
|
| 842 |
+
"step": 119
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.967741935483871,
|
| 846 |
+
"grad_norm": 5.0,
|
| 847 |
+
"learning_rate": 1.3695261579316776e-08,
|
| 848 |
+
"loss": 5.7444,
|
| 849 |
+
"step": 120
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.9758064516129032,
|
| 853 |
+
"grad_norm": 4.6875,
|
| 854 |
+
"learning_rate": 7.70666566718009e-09,
|
| 855 |
+
"loss": 5.4073,
|
| 856 |
+
"step": 121
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.9838709677419355,
|
| 860 |
+
"grad_norm": 4.78125,
|
| 861 |
+
"learning_rate": 3.4261631135654174e-09,
|
| 862 |
+
"loss": 5.6541,
|
| 863 |
+
"step": 122
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.9919354838709677,
|
| 867 |
+
"grad_norm": 4.96875,
|
| 868 |
+
"learning_rate": 8.566875611068503e-10,
|
| 869 |
+
"loss": 5.6101,
|
| 870 |
+
"step": 123
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.0,
|
| 874 |
+
"grad_norm": 5.40625,
|
| 875 |
+
"learning_rate": 0.0,
|
| 876 |
+
"loss": 5.6066,
|
| 877 |
+
"step": 124
|
| 878 |
+
}
|
| 879 |
+
],
|
| 880 |
+
"logging_steps": 1,
|
| 881 |
+
"max_steps": 124,
|
| 882 |
+
"num_input_tokens_seen": 0,
|
| 883 |
+
"num_train_epochs": 1,
|
| 884 |
+
"save_steps": 1000,
|
| 885 |
+
"stateful_callbacks": {
|
| 886 |
+
"TrainerControl": {
|
| 887 |
+
"args": {
|
| 888 |
+
"should_epoch_stop": false,
|
| 889 |
+
"should_evaluate": false,
|
| 890 |
+
"should_log": false,
|
| 891 |
+
"should_save": true,
|
| 892 |
+
"should_training_stop": true
|
| 893 |
+
},
|
| 894 |
+
"attributes": {}
|
| 895 |
+
}
|
| 896 |
+
},
|
| 897 |
+
"total_flos": 7570556190720000.0,
|
| 898 |
+
"train_batch_size": 1,
|
| 899 |
+
"trial_name": null,
|
| 900 |
+
"trial_params": null
|
| 901 |
+
}
|