Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +779 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,779 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 510,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00392156862745098,
|
| 13 |
+
"grad_norm": 16.544786089270207,
|
| 14 |
+
"learning_rate": 3.921568627450981e-07,
|
| 15 |
+
"loss": 1.6547,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0196078431372549,
|
| 20 |
+
"grad_norm": 14.762176989308287,
|
| 21 |
+
"learning_rate": 1.96078431372549e-06,
|
| 22 |
+
"loss": 1.617,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0392156862745098,
|
| 27 |
+
"grad_norm": 5.1165863939415654,
|
| 28 |
+
"learning_rate": 3.92156862745098e-06,
|
| 29 |
+
"loss": 1.4078,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.058823529411764705,
|
| 34 |
+
"grad_norm": 2.2787880179843194,
|
| 35 |
+
"learning_rate": 5.882352941176471e-06,
|
| 36 |
+
"loss": 1.2895,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.0784313725490196,
|
| 41 |
+
"grad_norm": 1.832234050305862,
|
| 42 |
+
"learning_rate": 7.84313725490196e-06,
|
| 43 |
+
"loss": 1.2313,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09803921568627451,
|
| 48 |
+
"grad_norm": 1.2251164606770868,
|
| 49 |
+
"learning_rate": 9.803921568627451e-06,
|
| 50 |
+
"loss": 1.1832,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11764705882352941,
|
| 55 |
+
"grad_norm": 1.51446787757622,
|
| 56 |
+
"learning_rate": 1.1764705882352942e-05,
|
| 57 |
+
"loss": 1.1558,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.13725490196078433,
|
| 62 |
+
"grad_norm": 1.1054530221391963,
|
| 63 |
+
"learning_rate": 1.3725490196078432e-05,
|
| 64 |
+
"loss": 1.1678,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1568627450980392,
|
| 69 |
+
"grad_norm": 1.1955722975383631,
|
| 70 |
+
"learning_rate": 1.568627450980392e-05,
|
| 71 |
+
"loss": 1.162,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.17647058823529413,
|
| 76 |
+
"grad_norm": 1.1601166405223855,
|
| 77 |
+
"learning_rate": 1.7647058823529414e-05,
|
| 78 |
+
"loss": 1.1303,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19607843137254902,
|
| 83 |
+
"grad_norm": 1.2824519717268024,
|
| 84 |
+
"learning_rate": 1.9607843137254903e-05,
|
| 85 |
+
"loss": 1.1272,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21568627450980393,
|
| 90 |
+
"grad_norm": 1.1682993723393238,
|
| 91 |
+
"learning_rate": 1.999625253802851e-05,
|
| 92 |
+
"loss": 1.1209,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23529411764705882,
|
| 97 |
+
"grad_norm": 1.302075491367261,
|
| 98 |
+
"learning_rate": 1.9981033287370443e-05,
|
| 99 |
+
"loss": 1.0887,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2549019607843137,
|
| 104 |
+
"grad_norm": 1.4556930650649318,
|
| 105 |
+
"learning_rate": 1.9954125840299165e-05,
|
| 106 |
+
"loss": 1.1226,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.27450980392156865,
|
| 111 |
+
"grad_norm": 1.289748942480489,
|
| 112 |
+
"learning_rate": 1.9915561706530882e-05,
|
| 113 |
+
"loss": 1.1327,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.29411764705882354,
|
| 118 |
+
"grad_norm": 1.1271075422611054,
|
| 119 |
+
"learning_rate": 1.9865386046236597e-05,
|
| 120 |
+
"loss": 1.1238,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3137254901960784,
|
| 125 |
+
"grad_norm": 1.2436214858595933,
|
| 126 |
+
"learning_rate": 1.9803657617157693e-05,
|
| 127 |
+
"loss": 1.08,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.3333333333333333,
|
| 132 |
+
"grad_norm": 1.1690258085683884,
|
| 133 |
+
"learning_rate": 1.973044870579824e-05,
|
| 134 |
+
"loss": 1.111,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.35294117647058826,
|
| 139 |
+
"grad_norm": 1.0669536039386904,
|
| 140 |
+
"learning_rate": 1.9645845042774555e-05,
|
| 141 |
+
"loss": 1.1074,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.37254901960784315,
|
| 146 |
+
"grad_norm": 1.2617056965745426,
|
| 147 |
+
"learning_rate": 1.9549945702421144e-05,
|
| 148 |
+
"loss": 1.1073,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.39215686274509803,
|
| 153 |
+
"grad_norm": 1.0667234116768218,
|
| 154 |
+
"learning_rate": 1.9442862986770645e-05,
|
| 155 |
+
"loss": 1.1018,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4117647058823529,
|
| 160 |
+
"grad_norm": 1.1310556155759037,
|
| 161 |
+
"learning_rate": 1.932472229404356e-05,
|
| 162 |
+
"loss": 1.1108,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.43137254901960786,
|
| 167 |
+
"grad_norm": 1.0707842591727832,
|
| 168 |
+
"learning_rate": 1.9195661971801825e-05,
|
| 169 |
+
"loss": 1.0978,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45098039215686275,
|
| 174 |
+
"grad_norm": 1.1964147323589545,
|
| 175 |
+
"learning_rate": 1.9055833154938208e-05,
|
| 176 |
+
"loss": 1.0952,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.47058823529411764,
|
| 181 |
+
"grad_norm": 1.0532587111658636,
|
| 182 |
+
"learning_rate": 1.8905399588691165e-05,
|
| 183 |
+
"loss": 1.0901,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.49019607843137253,
|
| 188 |
+
"grad_norm": 1.0534305550044942,
|
| 189 |
+
"learning_rate": 1.8744537436892517e-05,
|
| 190 |
+
"loss": 1.0739,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5098039215686274,
|
| 195 |
+
"grad_norm": 1.032737951798708,
|
| 196 |
+
"learning_rate": 1.8573435075672422e-05,
|
| 197 |
+
"loss": 1.0871,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5294117647058824,
|
| 202 |
+
"grad_norm": 1.0758071447152808,
|
| 203 |
+
"learning_rate": 1.839229287286327e-05,
|
| 204 |
+
"loss": 1.0945,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5490196078431373,
|
| 209 |
+
"grad_norm": 1.109162566034754,
|
| 210 |
+
"learning_rate": 1.8201322953360758e-05,
|
| 211 |
+
"loss": 1.0708,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5686274509803921,
|
| 216 |
+
"grad_norm": 1.317192816494283,
|
| 217 |
+
"learning_rate": 1.800074895071704e-05,
|
| 218 |
+
"loss": 1.0764,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5882352941176471,
|
| 223 |
+
"grad_norm": 1.0795334451744183,
|
| 224 |
+
"learning_rate": 1.7790805745256703e-05,
|
| 225 |
+
"loss": 1.0791,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6078431372549019,
|
| 230 |
+
"grad_norm": 1.0411588891000574,
|
| 231 |
+
"learning_rate": 1.7571739189022365e-05,
|
| 232 |
+
"loss": 1.091,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6274509803921569,
|
| 237 |
+
"grad_norm": 1.0108445010492009,
|
| 238 |
+
"learning_rate": 1.7343805817871885e-05,
|
| 239 |
+
"loss": 1.0809,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6470588235294118,
|
| 244 |
+
"grad_norm": 1.0901231094990715,
|
| 245 |
+
"learning_rate": 1.710727255106447e-05,
|
| 246 |
+
"loss": 1.0602,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6666666666666666,
|
| 251 |
+
"grad_norm": 1.0945695101274928,
|
| 252 |
+
"learning_rate": 1.686241637868734e-05,
|
| 253 |
+
"loss": 1.0777,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.6862745098039216,
|
| 258 |
+
"grad_norm": 1.193395829329031,
|
| 259 |
+
"learning_rate": 1.660952403728902e-05,
|
| 260 |
+
"loss": 1.0444,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.7058823529411765,
|
| 265 |
+
"grad_norm": 1.1047822048211582,
|
| 266 |
+
"learning_rate": 1.634889167409923e-05,
|
| 267 |
+
"loss": 1.0534,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7254901960784313,
|
| 272 |
+
"grad_norm": 1.114058959698183,
|
| 273 |
+
"learning_rate": 1.6080824500228367e-05,
|
| 274 |
+
"loss": 1.0529,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7450980392156863,
|
| 279 |
+
"grad_norm": 0.9904799394022883,
|
| 280 |
+
"learning_rate": 1.5805636433252892e-05,
|
| 281 |
+
"loss": 1.0793,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7647058823529411,
|
| 286 |
+
"grad_norm": 1.2525818789945302,
|
| 287 |
+
"learning_rate": 1.552364972960506e-05,
|
| 288 |
+
"loss": 1.0708,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7843137254901961,
|
| 293 |
+
"grad_norm": 1.1154866308925613,
|
| 294 |
+
"learning_rate": 1.5235194607197508e-05,
|
| 295 |
+
"loss": 1.0241,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.803921568627451,
|
| 300 |
+
"grad_norm": 1.0785210608261893,
|
| 301 |
+
"learning_rate": 1.494060885872464e-05,
|
| 302 |
+
"loss": 1.0733,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.8235294117647058,
|
| 307 |
+
"grad_norm": 1.1856668561642758,
|
| 308 |
+
"learning_rate": 1.4640237456093636e-05,
|
| 309 |
+
"loss": 1.0704,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8431372549019608,
|
| 314 |
+
"grad_norm": 1.1422476630742018,
|
| 315 |
+
"learning_rate": 1.4334432146448272e-05,
|
| 316 |
+
"loss": 1.0655,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8627450980392157,
|
| 321 |
+
"grad_norm": 0.9839617489985488,
|
| 322 |
+
"learning_rate": 1.4023551040258726e-05,
|
| 323 |
+
"loss": 1.0551,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8823529411764706,
|
| 328 |
+
"grad_norm": 1.0486919008457822,
|
| 329 |
+
"learning_rate": 1.3707958191959609e-05,
|
| 330 |
+
"loss": 1.0479,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.9019607843137255,
|
| 335 |
+
"grad_norm": 1.0279452846404875,
|
| 336 |
+
"learning_rate": 1.3388023173627413e-05,
|
| 337 |
+
"loss": 1.0466,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9215686274509803,
|
| 342 |
+
"grad_norm": 1.0670615822796066,
|
| 343 |
+
"learning_rate": 1.3064120642196549e-05,
|
| 344 |
+
"loss": 1.0554,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.9411764705882353,
|
| 349 |
+
"grad_norm": 1.0549543392774092,
|
| 350 |
+
"learning_rate": 1.2736629900720832e-05,
|
| 351 |
+
"loss": 1.0648,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9607843137254902,
|
| 356 |
+
"grad_norm": 0.9674738336064063,
|
| 357 |
+
"learning_rate": 1.2405934454194146e-05,
|
| 358 |
+
"loss": 1.0197,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9803921568627451,
|
| 363 |
+
"grad_norm": 0.9981821541554068,
|
| 364 |
+
"learning_rate": 1.2072421560450497e-05,
|
| 365 |
+
"loss": 1.0355,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.0,
|
| 370 |
+
"grad_norm": 0.9908436174645702,
|
| 371 |
+
"learning_rate": 1.1736481776669307e-05,
|
| 372 |
+
"loss": 1.0409,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.0,
|
| 377 |
+
"eval_loss": 1.0343828201293945,
|
| 378 |
+
"eval_runtime": 4.2299,
|
| 379 |
+
"eval_samples_per_second": 37.826,
|
| 380 |
+
"eval_steps_per_second": 0.709,
|
| 381 |
+
"step": 255
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0196078431372548,
|
| 385 |
+
"grad_norm": 2.78732105665186,
|
| 386 |
+
"learning_rate": 1.1398508502017047e-05,
|
| 387 |
+
"loss": 0.7588,
|
| 388 |
+
"step": 260
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0392156862745099,
|
| 392 |
+
"grad_norm": 1.5405892203726659,
|
| 393 |
+
"learning_rate": 1.1058897516960817e-05,
|
| 394 |
+
"loss": 0.7294,
|
| 395 |
+
"step": 265
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0588235294117647,
|
| 399 |
+
"grad_norm": 1.1503472595942978,
|
| 400 |
+
"learning_rate": 1.0718046519793276e-05,
|
| 401 |
+
"loss": 0.7224,
|
| 402 |
+
"step": 270
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0784313725490196,
|
| 406 |
+
"grad_norm": 1.1159446663960584,
|
| 407 |
+
"learning_rate": 1.0376354660911772e-05,
|
| 408 |
+
"loss": 0.7354,
|
| 409 |
+
"step": 275
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0980392156862746,
|
| 413 |
+
"grad_norm": 1.0502332790639606,
|
| 414 |
+
"learning_rate": 1.0034222075396954e-05,
|
| 415 |
+
"loss": 0.7255,
|
| 416 |
+
"step": 280
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1176470588235294,
|
| 420 |
+
"grad_norm": 0.9509421475398706,
|
| 421 |
+
"learning_rate": 9.692049414438298e-06,
|
| 422 |
+
"loss": 0.7154,
|
| 423 |
+
"step": 285
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1372549019607843,
|
| 427 |
+
"grad_norm": 0.988479903928271,
|
| 428 |
+
"learning_rate": 9.350237376155269e-06,
|
| 429 |
+
"loss": 0.7218,
|
| 430 |
+
"step": 290
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.156862745098039,
|
| 434 |
+
"grad_norm": 1.0085376248525226,
|
| 435 |
+
"learning_rate": 9.00918623636349e-06,
|
| 436 |
+
"loss": 0.7366,
|
| 437 |
+
"step": 295
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1764705882352942,
|
| 441 |
+
"grad_norm": 0.9961601641040383,
|
| 442 |
+
"learning_rate": 8.669295379835467e-06,
|
| 443 |
+
"loss": 0.7184,
|
| 444 |
+
"step": 300
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.196078431372549,
|
| 448 |
+
"grad_norm": 0.9921390197234601,
|
| 449 |
+
"learning_rate": 8.330962832604747e-06,
|
| 450 |
+
"loss": 0.7313,
|
| 451 |
+
"step": 305
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.215686274509804,
|
| 455 |
+
"grad_norm": 1.0706453846667212,
|
| 456 |
+
"learning_rate": 7.994584795861248e-06,
|
| 457 |
+
"loss": 0.7257,
|
| 458 |
+
"step": 310
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.2352941176470589,
|
| 462 |
+
"grad_norm": 1.0297863168482826,
|
| 463 |
+
"learning_rate": 7.660555181983517e-06,
|
| 464 |
+
"loss": 0.7468,
|
| 465 |
+
"step": 315
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2549019607843137,
|
| 469 |
+
"grad_norm": 0.9995332244709463,
|
| 470 |
+
"learning_rate": 7.329265153251285e-06,
|
| 471 |
+
"loss": 0.7439,
|
| 472 |
+
"step": 320
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2745098039215685,
|
| 476 |
+
"grad_norm": 1.0186580227806559,
|
| 477 |
+
"learning_rate": 7.001102663778533e-06,
|
| 478 |
+
"loss": 0.7172,
|
| 479 |
+
"step": 325
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2941176470588236,
|
| 483 |
+
"grad_norm": 1.014034224454933,
|
| 484 |
+
"learning_rate": 6.6764520052034054e-06,
|
| 485 |
+
"loss": 0.7309,
|
| 486 |
+
"step": 330
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.3137254901960784,
|
| 490 |
+
"grad_norm": 1.0151160383199134,
|
| 491 |
+
"learning_rate": 6.3556933566670656e-06,
|
| 492 |
+
"loss": 0.719,
|
| 493 |
+
"step": 335
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3333333333333333,
|
| 497 |
+
"grad_norm": 0.9563996745028378,
|
| 498 |
+
"learning_rate": 6.039202339608432e-06,
|
| 499 |
+
"loss": 0.7448,
|
| 500 |
+
"step": 340
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3529411764705883,
|
| 504 |
+
"grad_norm": 0.9430025085055084,
|
| 505 |
+
"learning_rate": 5.727349577896194e-06,
|
| 506 |
+
"loss": 0.715,
|
| 507 |
+
"step": 345
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3725490196078431,
|
| 511 |
+
"grad_norm": 0.9520104411677209,
|
| 512 |
+
"learning_rate": 5.420500263813141e-06,
|
| 513 |
+
"loss": 0.7423,
|
| 514 |
+
"step": 350
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.392156862745098,
|
| 518 |
+
"grad_norm": 0.9533462340832225,
|
| 519 |
+
"learning_rate": 5.119013730401152e-06,
|
| 520 |
+
"loss": 0.7206,
|
| 521 |
+
"step": 355
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.4117647058823528,
|
| 525 |
+
"grad_norm": 0.944013138090968,
|
| 526 |
+
"learning_rate": 4.823243030667576e-06,
|
| 527 |
+
"loss": 0.7112,
|
| 528 |
+
"step": 360
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4313725490196079,
|
| 532 |
+
"grad_norm": 0.9654739136152124,
|
| 533 |
+
"learning_rate": 4.533534524145756e-06,
|
| 534 |
+
"loss": 0.7092,
|
| 535 |
+
"step": 365
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.4509803921568627,
|
| 539 |
+
"grad_norm": 1.0325541079170395,
|
| 540 |
+
"learning_rate": 4.2502274712939355e-06,
|
| 541 |
+
"loss": 0.7218,
|
| 542 |
+
"step": 370
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4705882352941178,
|
| 546 |
+
"grad_norm": 0.945807759954701,
|
| 547 |
+
"learning_rate": 3.973653636207437e-06,
|
| 548 |
+
"loss": 0.7304,
|
| 549 |
+
"step": 375
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4901960784313726,
|
| 553 |
+
"grad_norm": 0.9339565355678595,
|
| 554 |
+
"learning_rate": 3.704136898109403e-06,
|
| 555 |
+
"loss": 0.723,
|
| 556 |
+
"step": 380
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.5098039215686274,
|
| 560 |
+
"grad_norm": 0.9680921345423967,
|
| 561 |
+
"learning_rate": 3.4419928720750274e-06,
|
| 562 |
+
"loss": 0.7382,
|
| 563 |
+
"step": 385
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5294117647058822,
|
| 567 |
+
"grad_norm": 0.9738097860413395,
|
| 568 |
+
"learning_rate": 3.1875285394334575e-06,
|
| 569 |
+
"loss": 0.7129,
|
| 570 |
+
"step": 390
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5490196078431373,
|
| 574 |
+
"grad_norm": 0.9541739057939187,
|
| 575 |
+
"learning_rate": 2.9410418882801682e-06,
|
| 576 |
+
"loss": 0.7392,
|
| 577 |
+
"step": 395
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.5686274509803921,
|
| 581 |
+
"grad_norm": 0.9315613334411681,
|
| 582 |
+
"learning_rate": 2.702821564520732e-06,
|
| 583 |
+
"loss": 0.7323,
|
| 584 |
+
"step": 400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5882352941176472,
|
| 588 |
+
"grad_norm": 0.9959745037027158,
|
| 589 |
+
"learning_rate": 2.4731465338547556e-06,
|
| 590 |
+
"loss": 0.7501,
|
| 591 |
+
"step": 405
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.607843137254902,
|
| 595 |
+
"grad_norm": 0.9359344044277552,
|
| 596 |
+
"learning_rate": 2.252285755095652e-06,
|
| 597 |
+
"loss": 0.733,
|
| 598 |
+
"step": 410
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.6274509803921569,
|
| 602 |
+
"grad_norm": 0.9222172843750491,
|
| 603 |
+
"learning_rate": 2.0404978652089325e-06,
|
| 604 |
+
"loss": 0.6955,
|
| 605 |
+
"step": 415
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.6470588235294117,
|
| 609 |
+
"grad_norm": 0.9295279337365954,
|
| 610 |
+
"learning_rate": 1.8380308764377841e-06,
|
| 611 |
+
"loss": 0.7209,
|
| 612 |
+
"step": 420
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6666666666666665,
|
| 616 |
+
"grad_norm": 0.9978286311378533,
|
| 617 |
+
"learning_rate": 1.6451218858706374e-06,
|
| 618 |
+
"loss": 0.7229,
|
| 619 |
+
"step": 425
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6862745098039216,
|
| 623 |
+
"grad_norm": 0.9673841209469143,
|
| 624 |
+
"learning_rate": 1.4619967977908157e-06,
|
| 625 |
+
"loss": 0.7197,
|
| 626 |
+
"step": 430
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.7058823529411766,
|
| 630 |
+
"grad_norm": 0.9537359526264119,
|
| 631 |
+
"learning_rate": 1.2888700591334225e-06,
|
| 632 |
+
"loss": 0.7144,
|
| 633 |
+
"step": 435
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.7254901960784315,
|
| 637 |
+
"grad_norm": 0.9009373079713732,
|
| 638 |
+
"learning_rate": 1.1259444083592585e-06,
|
| 639 |
+
"loss": 0.7078,
|
| 640 |
+
"step": 440
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7450980392156863,
|
| 644 |
+
"grad_norm": 0.9618775347330766,
|
| 645 |
+
"learning_rate": 9.734106380398022e-07,
|
| 646 |
+
"loss": 0.743,
|
| 647 |
+
"step": 445
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7647058823529411,
|
| 651 |
+
"grad_norm": 0.951269838967216,
|
| 652 |
+
"learning_rate": 8.31447371431372e-07,
|
| 653 |
+
"loss": 0.7278,
|
| 654 |
+
"step": 450
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.784313725490196,
|
| 658 |
+
"grad_norm": 0.912258612261103,
|
| 659 |
+
"learning_rate": 7.002208532999933e-07,
|
| 660 |
+
"loss": 0.7241,
|
| 661 |
+
"step": 455
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.803921568627451,
|
| 665 |
+
"grad_norm": 0.9276161963363886,
|
| 666 |
+
"learning_rate": 5.798847552420184e-07,
|
| 667 |
+
"loss": 0.7143,
|
| 668 |
+
"step": 460
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.8235294117647058,
|
| 672 |
+
"grad_norm": 0.9666247110758565,
|
| 673 |
+
"learning_rate": 4.7057999572843516e-07,
|
| 674 |
+
"loss": 0.7165,
|
| 675 |
+
"step": 465
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.843137254901961,
|
| 679 |
+
"grad_norm": 0.9040376476718873,
|
| 680 |
+
"learning_rate": 3.7243457508358784e-07,
|
| 681 |
+
"loss": 0.7192,
|
| 682 |
+
"step": 470
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8627450980392157,
|
| 686 |
+
"grad_norm": 0.9529298475220304,
|
| 687 |
+
"learning_rate": 2.8556342559159513e-07,
|
| 688 |
+
"loss": 0.7213,
|
| 689 |
+
"step": 475
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8823529411764706,
|
| 693 |
+
"grad_norm": 0.9373675816598103,
|
| 694 |
+
"learning_rate": 2.1006827690595478e-07,
|
| 695 |
+
"loss": 0.7247,
|
| 696 |
+
"step": 480
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.9019607843137254,
|
| 700 |
+
"grad_norm": 0.9049110221925867,
|
| 701 |
+
"learning_rate": 1.4603753691998735e-07,
|
| 702 |
+
"loss": 0.7217,
|
| 703 |
+
"step": 485
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.9215686274509802,
|
| 707 |
+
"grad_norm": 1.0200136766624364,
|
| 708 |
+
"learning_rate": 9.354618823758654e-08,
|
| 709 |
+
"loss": 0.712,
|
| 710 |
+
"step": 490
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.9411764705882353,
|
| 714 |
+
"grad_norm": 1.0048113834797883,
|
| 715 |
+
"learning_rate": 5.265570036553813e-08,
|
| 716 |
+
"loss": 0.7337,
|
| 717 |
+
"step": 495
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9607843137254903,
|
| 721 |
+
"grad_norm": 0.9199685223066211,
|
| 722 |
+
"learning_rate": 2.3413957730226144e-08,
|
| 723 |
+
"loss": 0.7121,
|
| 724 |
+
"step": 500
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9803921568627452,
|
| 728 |
+
"grad_norm": 0.8894954442490418,
|
| 729 |
+
"learning_rate": 5.855203603017945e-09,
|
| 730 |
+
"loss": 0.7124,
|
| 731 |
+
"step": 505
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 2.0,
|
| 735 |
+
"grad_norm": 0.9330570772451235,
|
| 736 |
+
"learning_rate": 0.0,
|
| 737 |
+
"loss": 0.7276,
|
| 738 |
+
"step": 510
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.0,
|
| 742 |
+
"eval_loss": 1.0487182140350342,
|
| 743 |
+
"eval_runtime": 5.9345,
|
| 744 |
+
"eval_samples_per_second": 26.961,
|
| 745 |
+
"eval_steps_per_second": 0.506,
|
| 746 |
+
"step": 510
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 2.0,
|
| 750 |
+
"step": 510,
|
| 751 |
+
"total_flos": 106783624396800.0,
|
| 752 |
+
"train_loss": 0.9181881465163886,
|
| 753 |
+
"train_runtime": 3659.086,
|
| 754 |
+
"train_samples_per_second": 8.919,
|
| 755 |
+
"train_steps_per_second": 0.139
|
| 756 |
+
}
|
| 757 |
+
],
|
| 758 |
+
"logging_steps": 5,
|
| 759 |
+
"max_steps": 510,
|
| 760 |
+
"num_input_tokens_seen": 0,
|
| 761 |
+
"num_train_epochs": 2,
|
| 762 |
+
"save_steps": 500,
|
| 763 |
+
"stateful_callbacks": {
|
| 764 |
+
"TrainerControl": {
|
| 765 |
+
"args": {
|
| 766 |
+
"should_epoch_stop": false,
|
| 767 |
+
"should_evaluate": false,
|
| 768 |
+
"should_log": false,
|
| 769 |
+
"should_save": true,
|
| 770 |
+
"should_training_stop": true
|
| 771 |
+
},
|
| 772 |
+
"attributes": {}
|
| 773 |
+
}
|
| 774 |
+
},
|
| 775 |
+
"total_flos": 106783624396800.0,
|
| 776 |
+
"train_batch_size": 8,
|
| 777 |
+
"trial_name": null,
|
| 778 |
+
"trial_params": null
|
| 779 |
+
}
|