Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +735 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,735 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 99,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.030303030303030304,
|
| 13 |
+
"grad_norm": 1.5626156330108643,
|
| 14 |
+
"learning_rate": 6.666666666666667e-06,
|
| 15 |
+
"loss": 1.8564,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06060606060606061,
|
| 20 |
+
"grad_norm": 1.567307710647583,
|
| 21 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 22 |
+
"loss": 1.8964,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09090909090909091,
|
| 27 |
+
"grad_norm": 1.4846278429031372,
|
| 28 |
+
"learning_rate": 2e-05,
|
| 29 |
+
"loss": 1.7225,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.12121212121212122,
|
| 34 |
+
"grad_norm": 1.5041078329086304,
|
| 35 |
+
"learning_rate": 1.9994645874763657e-05,
|
| 36 |
+
"loss": 1.7961,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.15151515151515152,
|
| 41 |
+
"grad_norm": 1.3576281070709229,
|
| 42 |
+
"learning_rate": 1.9978589232386036e-05,
|
| 43 |
+
"loss": 1.5684,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18181818181818182,
|
| 48 |
+
"grad_norm": 1.5841476917266846,
|
| 49 |
+
"learning_rate": 1.995184726672197e-05,
|
| 50 |
+
"loss": 1.6809,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.21212121212121213,
|
| 55 |
+
"grad_norm": 1.451629638671875,
|
| 56 |
+
"learning_rate": 1.9914448613738107e-05,
|
| 57 |
+
"loss": 1.6862,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.24242424242424243,
|
| 62 |
+
"grad_norm": 1.3648676872253418,
|
| 63 |
+
"learning_rate": 1.9866433320848793e-05,
|
| 64 |
+
"loss": 1.6725,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.2727272727272727,
|
| 69 |
+
"grad_norm": 1.4545490741729736,
|
| 70 |
+
"learning_rate": 1.9807852804032306e-05,
|
| 71 |
+
"loss": 1.615,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.30303030303030304,
|
| 76 |
+
"grad_norm": 1.4878568649291992,
|
| 77 |
+
"learning_rate": 1.9738769792773338e-05,
|
| 78 |
+
"loss": 1.567,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.3333333333333333,
|
| 83 |
+
"grad_norm": 1.9032509326934814,
|
| 84 |
+
"learning_rate": 1.9659258262890683e-05,
|
| 85 |
+
"loss": 1.6849,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36363636363636365,
|
| 90 |
+
"grad_norm": 1.4631747007369995,
|
| 91 |
+
"learning_rate": 1.956940335732209e-05,
|
| 92 |
+
"loss": 1.7436,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3939393939393939,
|
| 97 |
+
"grad_norm": 1.3237619400024414,
|
| 98 |
+
"learning_rate": 1.946930129495106e-05,
|
| 99 |
+
"loss": 1.7103,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.42424242424242425,
|
| 104 |
+
"grad_norm": 1.4598267078399658,
|
| 105 |
+
"learning_rate": 1.935905926757326e-05,
|
| 106 |
+
"loss": 1.813,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.45454545454545453,
|
| 111 |
+
"grad_norm": 1.6514884233474731,
|
| 112 |
+
"learning_rate": 1.9238795325112867e-05,
|
| 113 |
+
"loss": 1.7952,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.48484848484848486,
|
| 118 |
+
"grad_norm": 1.4365404844284058,
|
| 119 |
+
"learning_rate": 1.910863824921176e-05,
|
| 120 |
+
"loss": 1.7656,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.5151515151515151,
|
| 125 |
+
"grad_norm": 1.253454566001892,
|
| 126 |
+
"learning_rate": 1.8968727415326885e-05,
|
| 127 |
+
"loss": 1.8244,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5454545454545454,
|
| 132 |
+
"grad_norm": 1.189319133758545,
|
| 133 |
+
"learning_rate": 1.881921264348355e-05,
|
| 134 |
+
"loss": 1.6908,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5757575757575758,
|
| 139 |
+
"grad_norm": 1.5841031074523926,
|
| 140 |
+
"learning_rate": 1.866025403784439e-05,
|
| 141 |
+
"loss": 1.7654,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.6060606060606061,
|
| 146 |
+
"grad_norm": 1.2561309337615967,
|
| 147 |
+
"learning_rate": 1.849202181526579e-05,
|
| 148 |
+
"loss": 1.7271,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.6363636363636364,
|
| 153 |
+
"grad_norm": 1.532762885093689,
|
| 154 |
+
"learning_rate": 1.8314696123025456e-05,
|
| 155 |
+
"loss": 1.7545,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.6666666666666666,
|
| 160 |
+
"grad_norm": 1.6211514472961426,
|
| 161 |
+
"learning_rate": 1.8128466845916156e-05,
|
| 162 |
+
"loss": 1.4955,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.696969696969697,
|
| 167 |
+
"grad_norm": 1.5107672214508057,
|
| 168 |
+
"learning_rate": 1.7933533402912354e-05,
|
| 169 |
+
"loss": 1.5387,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.7272727272727273,
|
| 174 |
+
"grad_norm": 1.3635034561157227,
|
| 175 |
+
"learning_rate": 1.773010453362737e-05,
|
| 176 |
+
"loss": 1.5995,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.7575757575757576,
|
| 181 |
+
"grad_norm": 1.3096054792404175,
|
| 182 |
+
"learning_rate": 1.7518398074789776e-05,
|
| 183 |
+
"loss": 1.7549,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.7878787878787878,
|
| 188 |
+
"grad_norm": 1.2032136917114258,
|
| 189 |
+
"learning_rate": 1.7298640726978357e-05,
|
| 190 |
+
"loss": 1.634,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.8181818181818182,
|
| 195 |
+
"grad_norm": 1.1950513124465942,
|
| 196 |
+
"learning_rate": 1.7071067811865477e-05,
|
| 197 |
+
"loss": 1.6768,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.8484848484848485,
|
| 202 |
+
"grad_norm": 1.6178178787231445,
|
| 203 |
+
"learning_rate": 1.6835923020228714e-05,
|
| 204 |
+
"loss": 1.739,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.8787878787878788,
|
| 209 |
+
"grad_norm": 1.3394091129302979,
|
| 210 |
+
"learning_rate": 1.659345815100069e-05,
|
| 211 |
+
"loss": 1.8401,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9090909090909091,
|
| 216 |
+
"grad_norm": 1.3852413892745972,
|
| 217 |
+
"learning_rate": 1.6343932841636455e-05,
|
| 218 |
+
"loss": 1.6928,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.9393939393939394,
|
| 223 |
+
"grad_norm": 1.2562795877456665,
|
| 224 |
+
"learning_rate": 1.608761429008721e-05,
|
| 225 |
+
"loss": 1.7373,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.9696969696969697,
|
| 230 |
+
"grad_norm": 1.1925987005233765,
|
| 231 |
+
"learning_rate": 1.5824776968678024e-05,
|
| 232 |
+
"loss": 1.5096,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.0,
|
| 237 |
+
"grad_norm": 1.7065600156784058,
|
| 238 |
+
"learning_rate": 1.5555702330196024e-05,
|
| 239 |
+
"loss": 1.3249,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.0303030303030303,
|
| 244 |
+
"grad_norm": 1.791105031967163,
|
| 245 |
+
"learning_rate": 1.528067850650368e-05,
|
| 246 |
+
"loss": 1.2349,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.0606060606060606,
|
| 251 |
+
"grad_norm": 1.6474976539611816,
|
| 252 |
+
"learning_rate": 1.5000000000000002e-05,
|
| 253 |
+
"loss": 1.2779,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.0909090909090908,
|
| 258 |
+
"grad_norm": 1.5695178508758545,
|
| 259 |
+
"learning_rate": 1.4713967368259981e-05,
|
| 260 |
+
"loss": 1.2858,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.121212121212121,
|
| 265 |
+
"grad_norm": 1.3161590099334717,
|
| 266 |
+
"learning_rate": 1.4422886902190014e-05,
|
| 267 |
+
"loss": 1.1867,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 1.1515151515151516,
|
| 272 |
+
"grad_norm": 1.5779008865356445,
|
| 273 |
+
"learning_rate": 1.4127070298043949e-05,
|
| 274 |
+
"loss": 1.188,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.1818181818181819,
|
| 279 |
+
"grad_norm": 1.5446536540985107,
|
| 280 |
+
"learning_rate": 1.3826834323650899e-05,
|
| 281 |
+
"loss": 1.1306,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 1.2121212121212122,
|
| 286 |
+
"grad_norm": 1.9311134815216064,
|
| 287 |
+
"learning_rate": 1.3522500479212337e-05,
|
| 288 |
+
"loss": 1.2253,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 1.2424242424242424,
|
| 293 |
+
"grad_norm": 2.0665199756622314,
|
| 294 |
+
"learning_rate": 1.3214394653031616e-05,
|
| 295 |
+
"loss": 1.2405,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.2727272727272727,
|
| 300 |
+
"grad_norm": 1.340359091758728,
|
| 301 |
+
"learning_rate": 1.2902846772544625e-05,
|
| 302 |
+
"loss": 1.0486,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.303030303030303,
|
| 307 |
+
"grad_norm": 1.5210260152816772,
|
| 308 |
+
"learning_rate": 1.2588190451025209e-05,
|
| 309 |
+
"loss": 1.1205,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 1.3333333333333333,
|
| 314 |
+
"grad_norm": 1.4104619026184082,
|
| 315 |
+
"learning_rate": 1.2270762630343734e-05,
|
| 316 |
+
"loss": 1.1148,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.3636363636363638,
|
| 321 |
+
"grad_norm": 1.373587965965271,
|
| 322 |
+
"learning_rate": 1.1950903220161286e-05,
|
| 323 |
+
"loss": 1.1462,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 1.393939393939394,
|
| 328 |
+
"grad_norm": 1.364762306213379,
|
| 329 |
+
"learning_rate": 1.162895473394589e-05,
|
| 330 |
+
"loss": 1.0527,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 1.4242424242424243,
|
| 335 |
+
"grad_norm": 1.7049798965454102,
|
| 336 |
+
"learning_rate": 1.130526192220052e-05,
|
| 337 |
+
"loss": 1.1314,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.4545454545454546,
|
| 342 |
+
"grad_norm": 1.7813589572906494,
|
| 343 |
+
"learning_rate": 1.098017140329561e-05,
|
| 344 |
+
"loss": 1.1012,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.4848484848484849,
|
| 349 |
+
"grad_norm": 1.595564603805542,
|
| 350 |
+
"learning_rate": 1.0654031292301432e-05,
|
| 351 |
+
"loss": 1.0513,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.5151515151515151,
|
| 356 |
+
"grad_norm": 1.5489888191223145,
|
| 357 |
+
"learning_rate": 1.0327190828217763e-05,
|
| 358 |
+
"loss": 1.1117,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.5454545454545454,
|
| 363 |
+
"grad_norm": 1.7180898189544678,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 1.1898,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.5757575757575757,
|
| 370 |
+
"grad_norm": 1.3736799955368042,
|
| 371 |
+
"learning_rate": 9.67280917178224e-06,
|
| 372 |
+
"loss": 1.0677,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.606060606060606,
|
| 377 |
+
"grad_norm": 1.3806053400039673,
|
| 378 |
+
"learning_rate": 9.34596870769857e-06,
|
| 379 |
+
"loss": 1.1602,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6363636363636362,
|
| 384 |
+
"grad_norm": 1.5797897577285767,
|
| 385 |
+
"learning_rate": 9.019828596704394e-06,
|
| 386 |
+
"loss": 1.0706,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.6666666666666665,
|
| 391 |
+
"grad_norm": 1.6175159215927124,
|
| 392 |
+
"learning_rate": 8.694738077799487e-06,
|
| 393 |
+
"loss": 1.147,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.696969696969697,
|
| 398 |
+
"grad_norm": 1.6543636322021484,
|
| 399 |
+
"learning_rate": 8.371045266054114e-06,
|
| 400 |
+
"loss": 0.9402,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.7272727272727273,
|
| 405 |
+
"grad_norm": 1.4852937459945679,
|
| 406 |
+
"learning_rate": 8.04909677983872e-06,
|
| 407 |
+
"loss": 1.0766,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.7575757575757576,
|
| 412 |
+
"grad_norm": 1.6130305528640747,
|
| 413 |
+
"learning_rate": 7.72923736965627e-06,
|
| 414 |
+
"loss": 1.0769,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.7878787878787878,
|
| 419 |
+
"grad_norm": 1.4735853672027588,
|
| 420 |
+
"learning_rate": 7.411809548974792e-06,
|
| 421 |
+
"loss": 0.9402,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8181818181818183,
|
| 426 |
+
"grad_norm": 1.5684136152267456,
|
| 427 |
+
"learning_rate": 7.097153227455379e-06,
|
| 428 |
+
"loss": 1.1553,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.8484848484848486,
|
| 433 |
+
"grad_norm": 1.4093765020370483,
|
| 434 |
+
"learning_rate": 6.785605346968387e-06,
|
| 435 |
+
"loss": 1.0747,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.878787878787879,
|
| 440 |
+
"grad_norm": 1.2906310558319092,
|
| 441 |
+
"learning_rate": 6.4774995207876654e-06,
|
| 442 |
+
"loss": 1.0484,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.9090909090909092,
|
| 447 |
+
"grad_norm": 1.5653777122497559,
|
| 448 |
+
"learning_rate": 6.173165676349103e-06,
|
| 449 |
+
"loss": 1.2066,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.9393939393939394,
|
| 454 |
+
"grad_norm": 1.3291213512420654,
|
| 455 |
+
"learning_rate": 5.872929701956054e-06,
|
| 456 |
+
"loss": 1.0324,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.9696969696969697,
|
| 461 |
+
"grad_norm": 1.45272958278656,
|
| 462 |
+
"learning_rate": 5.5771130978099896e-06,
|
| 463 |
+
"loss": 1.0972,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.0,
|
| 468 |
+
"grad_norm": 1.6642159223556519,
|
| 469 |
+
"learning_rate": 5.286032631740023e-06,
|
| 470 |
+
"loss": 0.893,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.0303030303030303,
|
| 475 |
+
"grad_norm": 1.6590158939361572,
|
| 476 |
+
"learning_rate": 5.000000000000003e-06,
|
| 477 |
+
"loss": 0.8978,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 2.0606060606060606,
|
| 482 |
+
"grad_norm": 2.192969799041748,
|
| 483 |
+
"learning_rate": 4.7193214934963204e-06,
|
| 484 |
+
"loss": 0.891,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 2.090909090909091,
|
| 489 |
+
"grad_norm": 1.7911643981933594,
|
| 490 |
+
"learning_rate": 4.444297669803981e-06,
|
| 491 |
+
"loss": 0.7529,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 2.121212121212121,
|
| 496 |
+
"grad_norm": 1.4908910989761353,
|
| 497 |
+
"learning_rate": 4.17522303132198e-06,
|
| 498 |
+
"loss": 0.7556,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 2.1515151515151514,
|
| 503 |
+
"grad_norm": 1.7466297149658203,
|
| 504 |
+
"learning_rate": 3.912385709912794e-06,
|
| 505 |
+
"loss": 0.8257,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.1818181818181817,
|
| 510 |
+
"grad_norm": 1.5904890298843384,
|
| 511 |
+
"learning_rate": 3.6560671583635467e-06,
|
| 512 |
+
"loss": 0.7562,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.212121212121212,
|
| 517 |
+
"grad_norm": 1.60899817943573,
|
| 518 |
+
"learning_rate": 3.4065418489993118e-06,
|
| 519 |
+
"loss": 0.7925,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 2.242424242424242,
|
| 524 |
+
"grad_norm": 1.9628279209136963,
|
| 525 |
+
"learning_rate": 3.1640769797712865e-06,
|
| 526 |
+
"loss": 0.7809,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.2727272727272725,
|
| 531 |
+
"grad_norm": 1.9455692768096924,
|
| 532 |
+
"learning_rate": 2.9289321881345257e-06,
|
| 533 |
+
"loss": 0.7837,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 2.303030303030303,
|
| 538 |
+
"grad_norm": 1.9199553728103638,
|
| 539 |
+
"learning_rate": 2.7013592730216464e-06,
|
| 540 |
+
"loss": 0.9131,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 2.3333333333333335,
|
| 545 |
+
"grad_norm": 2.134833335876465,
|
| 546 |
+
"learning_rate": 2.4816019252102274e-06,
|
| 547 |
+
"loss": 0.7476,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.3636363636363638,
|
| 552 |
+
"grad_norm": 1.9508200883865356,
|
| 553 |
+
"learning_rate": 2.26989546637263e-06,
|
| 554 |
+
"loss": 0.7747,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.393939393939394,
|
| 559 |
+
"grad_norm": 1.4582135677337646,
|
| 560 |
+
"learning_rate": 2.0664665970876496e-06,
|
| 561 |
+
"loss": 0.8055,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 2.4242424242424243,
|
| 566 |
+
"grad_norm": 1.7351216077804565,
|
| 567 |
+
"learning_rate": 1.8715331540838488e-06,
|
| 568 |
+
"loss": 0.8209,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.4545454545454546,
|
| 573 |
+
"grad_norm": 1.351690411567688,
|
| 574 |
+
"learning_rate": 1.6853038769745466e-06,
|
| 575 |
+
"loss": 0.7738,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 2.484848484848485,
|
| 580 |
+
"grad_norm": 1.5985406637191772,
|
| 581 |
+
"learning_rate": 1.5079781847342122e-06,
|
| 582 |
+
"loss": 0.6174,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 2.515151515151515,
|
| 587 |
+
"grad_norm": 1.7237467765808105,
|
| 588 |
+
"learning_rate": 1.339745962155613e-06,
|
| 589 |
+
"loss": 0.6738,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.5454545454545454,
|
| 594 |
+
"grad_norm": 1.442694067955017,
|
| 595 |
+
"learning_rate": 1.1807873565164507e-06,
|
| 596 |
+
"loss": 0.913,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.5757575757575757,
|
| 601 |
+
"grad_norm": 1.5762035846710205,
|
| 602 |
+
"learning_rate": 1.0312725846731174e-06,
|
| 603 |
+
"loss": 0.8292,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.606060606060606,
|
| 608 |
+
"grad_norm": 1.6882027387619019,
|
| 609 |
+
"learning_rate": 8.91361750788241e-07,
|
| 610 |
+
"loss": 0.7793,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 2.6363636363636362,
|
| 615 |
+
"grad_norm": 1.3225022554397583,
|
| 616 |
+
"learning_rate": 7.612046748871327e-07,
|
| 617 |
+
"loss": 0.7621,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 2.6666666666666665,
|
| 622 |
+
"grad_norm": 1.8009114265441895,
|
| 623 |
+
"learning_rate": 6.409407324267448e-07,
|
| 624 |
+
"loss": 0.7223,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 2.6969696969696972,
|
| 629 |
+
"grad_norm": 1.5908373594284058,
|
| 630 |
+
"learning_rate": 5.306987050489442e-07,
|
| 631 |
+
"loss": 0.801,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.7272727272727275,
|
| 636 |
+
"grad_norm": 1.5530376434326172,
|
| 637 |
+
"learning_rate": 4.305966426779118e-07,
|
| 638 |
+
"loss": 0.735,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.757575757575758,
|
| 643 |
+
"grad_norm": 1.3646830320358276,
|
| 644 |
+
"learning_rate": 3.4074173710931804e-07,
|
| 645 |
+
"loss": 0.7068,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 2.787878787878788,
|
| 650 |
+
"grad_norm": 1.6398669481277466,
|
| 651 |
+
"learning_rate": 2.612302072266637e-07,
|
| 652 |
+
"loss": 0.7385,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 2.8181818181818183,
|
| 657 |
+
"grad_norm": 1.7059106826782227,
|
| 658 |
+
"learning_rate": 1.921471959676957e-07,
|
| 659 |
+
"loss": 0.6993,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 2.8484848484848486,
|
| 664 |
+
"grad_norm": 1.6918838024139404,
|
| 665 |
+
"learning_rate": 1.3356667915121025e-07,
|
| 666 |
+
"loss": 0.8254,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 2.878787878787879,
|
| 671 |
+
"grad_norm": 1.3561716079711914,
|
| 672 |
+
"learning_rate": 8.555138626189619e-08,
|
| 673 |
+
"loss": 0.668,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.909090909090909,
|
| 678 |
+
"grad_norm": 1.593574047088623,
|
| 679 |
+
"learning_rate": 4.815273327803183e-08,
|
| 680 |
+
"loss": 0.7856,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.9393939393939394,
|
| 685 |
+
"grad_norm": 1.6517926454544067,
|
| 686 |
+
"learning_rate": 2.1410767613965212e-08,
|
| 687 |
+
"loss": 0.7489,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 2.9696969696969697,
|
| 692 |
+
"grad_norm": 1.5703673362731934,
|
| 693 |
+
"learning_rate": 5.354125236343155e-09,
|
| 694 |
+
"loss": 0.7786,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 3.0,
|
| 699 |
+
"grad_norm": 1.5764566659927368,
|
| 700 |
+
"learning_rate": 0.0,
|
| 701 |
+
"loss": 0.7356,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 3.0,
|
| 706 |
+
"step": 99,
|
| 707 |
+
"total_flos": 2.3635472472132813e+17,
|
| 708 |
+
"train_loss": 1.1949122909343604,
|
| 709 |
+
"train_runtime": 1081.5376,
|
| 710 |
+
"train_samples_per_second": 2.854,
|
| 711 |
+
"train_steps_per_second": 0.092
|
| 712 |
+
}
|
| 713 |
+
],
|
| 714 |
+
"logging_steps": 1,
|
| 715 |
+
"max_steps": 99,
|
| 716 |
+
"num_input_tokens_seen": 0,
|
| 717 |
+
"num_train_epochs": 3,
|
| 718 |
+
"save_steps": 0,
|
| 719 |
+
"stateful_callbacks": {
|
| 720 |
+
"TrainerControl": {
|
| 721 |
+
"args": {
|
| 722 |
+
"should_epoch_stop": false,
|
| 723 |
+
"should_evaluate": false,
|
| 724 |
+
"should_log": false,
|
| 725 |
+
"should_save": true,
|
| 726 |
+
"should_training_stop": true
|
| 727 |
+
},
|
| 728 |
+
"attributes": {}
|
| 729 |
+
}
|
| 730 |
+
},
|
| 731 |
+
"total_flos": 2.3635472472132813e+17,
|
| 732 |
+
"train_batch_size": 8,
|
| 733 |
+
"trial_name": null,
|
| 734 |
+
"trial_params": null
|
| 735 |
+
}
|