Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +570 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,570 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 90,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"learning_rate": 1.3333333333333333e-05,
|
| 14 |
+
"loss": 0.166,
|
| 15 |
+
"step": 1
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.02,
|
| 19 |
+
"learning_rate": 2.6666666666666667e-05,
|
| 20 |
+
"loss": 0.3088,
|
| 21 |
+
"step": 2
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.03,
|
| 25 |
+
"learning_rate": 4e-05,
|
| 26 |
+
"loss": 0.2176,
|
| 27 |
+
"step": 3
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.04,
|
| 31 |
+
"learning_rate": 3.998696190779354e-05,
|
| 32 |
+
"loss": 0.2843,
|
| 33 |
+
"step": 4
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.06,
|
| 37 |
+
"learning_rate": 3.9947864630359005e-05,
|
| 38 |
+
"loss": 0.1867,
|
| 39 |
+
"step": 5
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.07,
|
| 43 |
+
"learning_rate": 3.9882759143087194e-05,
|
| 44 |
+
"loss": 0.7367,
|
| 45 |
+
"step": 6
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.08,
|
| 49 |
+
"learning_rate": 3.979173033111275e-05,
|
| 50 |
+
"loss": 0.3016,
|
| 51 |
+
"step": 7
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.09,
|
| 55 |
+
"learning_rate": 3.9674896878640054e-05,
|
| 56 |
+
"loss": 0.1187,
|
| 57 |
+
"step": 8
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.1,
|
| 61 |
+
"learning_rate": 3.953241111420174e-05,
|
| 62 |
+
"loss": 0.2443,
|
| 63 |
+
"step": 9
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.11,
|
| 67 |
+
"learning_rate": 3.936445881205127e-05,
|
| 68 |
+
"loss": 0.2212,
|
| 69 |
+
"step": 10
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.12,
|
| 73 |
+
"learning_rate": 3.9171258949948827e-05,
|
| 74 |
+
"loss": 0.1678,
|
| 75 |
+
"step": 11
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.13,
|
| 79 |
+
"learning_rate": 3.8953063423656055e-05,
|
| 80 |
+
"loss": 0.1301,
|
| 81 |
+
"step": 12
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.14,
|
| 85 |
+
"learning_rate": 3.871015671851202e-05,
|
| 86 |
+
"loss": 0.44,
|
| 87 |
+
"step": 13
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.16,
|
| 91 |
+
"learning_rate": 3.8442855538518667e-05,
|
| 92 |
+
"loss": 0.277,
|
| 93 |
+
"step": 14
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.17,
|
| 97 |
+
"learning_rate": 3.815150839341915e-05,
|
| 98 |
+
"loss": 0.2494,
|
| 99 |
+
"step": 15
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.18,
|
| 103 |
+
"learning_rate": 3.7836495144307644e-05,
|
| 104 |
+
"loss": 0.4562,
|
| 105 |
+
"step": 16
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.19,
|
| 109 |
+
"learning_rate": 3.7498226508362996e-05,
|
| 110 |
+
"loss": 0.0884,
|
| 111 |
+
"step": 17
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.2,
|
| 115 |
+
"learning_rate": 3.7137143523351787e-05,
|
| 116 |
+
"loss": 2.1156,
|
| 117 |
+
"step": 18
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.21,
|
| 121 |
+
"learning_rate": 3.67537169725993e-05,
|
| 122 |
+
"loss": 0.7894,
|
| 123 |
+
"step": 19
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.22,
|
| 127 |
+
"learning_rate": 3.634844677117784e-05,
|
| 128 |
+
"loss": 0.1293,
|
| 129 |
+
"step": 20
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.23,
|
| 133 |
+
"learning_rate": 3.592186131411288e-05,
|
| 134 |
+
"loss": 1.03,
|
| 135 |
+
"step": 21
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.24,
|
| 139 |
+
"learning_rate": 3.547451678745673e-05,
|
| 140 |
+
"loss": 0.066,
|
| 141 |
+
"step": 22
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.26,
|
| 145 |
+
"learning_rate": 3.500699644312805e-05,
|
| 146 |
+
"loss": 0.3185,
|
| 147 |
+
"step": 23
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.27,
|
| 151 |
+
"learning_rate": 3.451990983846262e-05,
|
| 152 |
+
"loss": 0.155,
|
| 153 |
+
"step": 24
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.28,
|
| 157 |
+
"learning_rate": 3.401389204146685e-05,
|
| 158 |
+
"loss": 0.1746,
|
| 159 |
+
"step": 25
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.29,
|
| 163 |
+
"learning_rate": 3.3489602802810276e-05,
|
| 164 |
+
"loss": 0.0698,
|
| 165 |
+
"step": 26
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.3,
|
| 169 |
+
"learning_rate": 3.294772569563656e-05,
|
| 170 |
+
"loss": 0.139,
|
| 171 |
+
"step": 27
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.31,
|
| 175 |
+
"learning_rate": 3.2388967224314464e-05,
|
| 176 |
+
"loss": 0.0764,
|
| 177 |
+
"step": 28
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.32,
|
| 181 |
+
"learning_rate": 3.181405590329102e-05,
|
| 182 |
+
"loss": 0.2437,
|
| 183 |
+
"step": 29
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.33,
|
| 187 |
+
"learning_rate": 3.122374130724765e-05,
|
| 188 |
+
"loss": 0.3985,
|
| 189 |
+
"step": 30
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.34,
|
| 193 |
+
"learning_rate": 3.061879309379774e-05,
|
| 194 |
+
"loss": 0.2214,
|
| 195 |
+
"step": 31
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.36,
|
| 199 |
+
"learning_rate": 3.0000000000000004e-05,
|
| 200 |
+
"loss": 0.4577,
|
| 201 |
+
"step": 32
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.37,
|
| 205 |
+
"learning_rate": 2.9368168813995806e-05,
|
| 206 |
+
"loss": 0.202,
|
| 207 |
+
"step": 33
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.38,
|
| 211 |
+
"learning_rate": 2.872412332311135e-05,
|
| 212 |
+
"loss": 0.339,
|
| 213 |
+
"step": 34
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.39,
|
| 217 |
+
"learning_rate": 2.8068703239796175e-05,
|
| 218 |
+
"loss": 0.1432,
|
| 219 |
+
"step": 35
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4,
|
| 223 |
+
"learning_rate": 2.7402763106798295e-05,
|
| 224 |
+
"loss": 0.518,
|
| 225 |
+
"step": 36
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.41,
|
| 229 |
+
"learning_rate": 2.6727171183003502e-05,
|
| 230 |
+
"loss": 0.1572,
|
| 231 |
+
"step": 37
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.42,
|
| 235 |
+
"learning_rate": 2.6042808311391456e-05,
|
| 236 |
+
"loss": 0.4009,
|
| 237 |
+
"step": 38
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.43,
|
| 241 |
+
"learning_rate": 2.5350566770584423e-05,
|
| 242 |
+
"loss": 0.4428,
|
| 243 |
+
"step": 39
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.44,
|
| 247 |
+
"learning_rate": 2.4651349111486212e-05,
|
| 248 |
+
"loss": 0.1903,
|
| 249 |
+
"step": 40
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.46,
|
| 253 |
+
"learning_rate": 2.3946066980528e-05,
|
| 254 |
+
"loss": 0.2332,
|
| 255 |
+
"step": 41
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.47,
|
| 259 |
+
"learning_rate": 2.32356399310553e-05,
|
| 260 |
+
"loss": 1.3527,
|
| 261 |
+
"step": 42
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.48,
|
| 265 |
+
"learning_rate": 2.252099422440579e-05,
|
| 266 |
+
"loss": 0.1223,
|
| 267 |
+
"step": 43
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.49,
|
| 271 |
+
"learning_rate": 2.180306162224131e-05,
|
| 272 |
+
"loss": 0.1199,
|
| 273 |
+
"step": 44
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.5,
|
| 277 |
+
"learning_rate": 2.1082778171708355e-05,
|
| 278 |
+
"loss": 0.2352,
|
| 279 |
+
"step": 45
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.51,
|
| 283 |
+
"learning_rate": 2.036108298501121e-05,
|
| 284 |
+
"loss": 0.1032,
|
| 285 |
+
"step": 46
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.52,
|
| 289 |
+
"learning_rate": 1.963891701498879e-05,
|
| 290 |
+
"loss": 0.0865,
|
| 291 |
+
"step": 47
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 0.53,
|
| 295 |
+
"learning_rate": 1.8917221828291652e-05,
|
| 296 |
+
"loss": 0.2237,
|
| 297 |
+
"step": 48
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.54,
|
| 301 |
+
"learning_rate": 1.8196938377758696e-05,
|
| 302 |
+
"loss": 0.1759,
|
| 303 |
+
"step": 49
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.56,
|
| 307 |
+
"learning_rate": 1.7479005775594216e-05,
|
| 308 |
+
"loss": 0.3111,
|
| 309 |
+
"step": 50
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.57,
|
| 313 |
+
"learning_rate": 1.6764360068944706e-05,
|
| 314 |
+
"loss": 0.0737,
|
| 315 |
+
"step": 51
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 0.58,
|
| 319 |
+
"learning_rate": 1.6053933019472003e-05,
|
| 320 |
+
"loss": 0.242,
|
| 321 |
+
"step": 52
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.59,
|
| 325 |
+
"learning_rate": 1.5348650888513798e-05,
|
| 326 |
+
"loss": 0.1971,
|
| 327 |
+
"step": 53
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.6,
|
| 331 |
+
"learning_rate": 1.4649433229415588e-05,
|
| 332 |
+
"loss": 0.2123,
|
| 333 |
+
"step": 54
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.61,
|
| 337 |
+
"learning_rate": 1.3957191688608544e-05,
|
| 338 |
+
"loss": 0.0686,
|
| 339 |
+
"step": 55
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.62,
|
| 343 |
+
"learning_rate": 1.3272828816996498e-05,
|
| 344 |
+
"loss": 0.2889,
|
| 345 |
+
"step": 56
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.63,
|
| 349 |
+
"learning_rate": 1.2597236893201712e-05,
|
| 350 |
+
"loss": 0.2579,
|
| 351 |
+
"step": 57
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.64,
|
| 355 |
+
"learning_rate": 1.1931296760203831e-05,
|
| 356 |
+
"loss": 0.2627,
|
| 357 |
+
"step": 58
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.66,
|
| 361 |
+
"learning_rate": 1.127587667688865e-05,
|
| 362 |
+
"loss": 0.5957,
|
| 363 |
+
"step": 59
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.67,
|
| 367 |
+
"learning_rate": 1.06318311860042e-05,
|
| 368 |
+
"loss": 0.1272,
|
| 369 |
+
"step": 60
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.68,
|
| 373 |
+
"learning_rate": 1.0000000000000006e-05,
|
| 374 |
+
"loss": 0.2622,
|
| 375 |
+
"step": 61
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 0.69,
|
| 379 |
+
"learning_rate": 9.381206906202268e-06,
|
| 380 |
+
"loss": 0.1471,
|
| 381 |
+
"step": 62
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.7,
|
| 385 |
+
"learning_rate": 8.776258692752355e-06,
|
| 386 |
+
"loss": 0.1139,
|
| 387 |
+
"step": 63
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.71,
|
| 391 |
+
"learning_rate": 8.185944096708982e-06,
|
| 392 |
+
"loss": 0.3799,
|
| 393 |
+
"step": 64
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.72,
|
| 397 |
+
"learning_rate": 7.611032775685541e-06,
|
| 398 |
+
"loss": 0.7215,
|
| 399 |
+
"step": 65
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.73,
|
| 403 |
+
"learning_rate": 7.052274304363449e-06,
|
| 404 |
+
"loss": 0.1838,
|
| 405 |
+
"step": 66
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 0.74,
|
| 409 |
+
"learning_rate": 6.510397197189724e-06,
|
| 410 |
+
"loss": 0.0487,
|
| 411 |
+
"step": 67
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.76,
|
| 415 |
+
"learning_rate": 5.9861079585331535e-06,
|
| 416 |
+
"loss": 0.1821,
|
| 417 |
+
"step": 68
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.77,
|
| 421 |
+
"learning_rate": 5.480090161537388e-06,
|
| 422 |
+
"loss": 0.1827,
|
| 423 |
+
"step": 69
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.78,
|
| 427 |
+
"learning_rate": 4.993003556871954e-06,
|
| 428 |
+
"loss": 0.1902,
|
| 429 |
+
"step": 70
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.79,
|
| 433 |
+
"learning_rate": 4.525483212543273e-06,
|
| 434 |
+
"loss": 0.1518,
|
| 435 |
+
"step": 71
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.8,
|
| 439 |
+
"learning_rate": 4.078138685887125e-06,
|
| 440 |
+
"loss": 0.629,
|
| 441 |
+
"step": 72
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.81,
|
| 445 |
+
"learning_rate": 3.6515532288221646e-06,
|
| 446 |
+
"loss": 0.5371,
|
| 447 |
+
"step": 73
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 0.82,
|
| 451 |
+
"learning_rate": 3.2462830274007073e-06,
|
| 452 |
+
"loss": 0.1234,
|
| 453 |
+
"step": 74
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.83,
|
| 457 |
+
"learning_rate": 2.8628564766482193e-06,
|
| 458 |
+
"loss": 0.0766,
|
| 459 |
+
"step": 75
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.84,
|
| 463 |
+
"learning_rate": 2.5017734916370073e-06,
|
| 464 |
+
"loss": 0.1043,
|
| 465 |
+
"step": 76
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.86,
|
| 469 |
+
"learning_rate": 2.1635048556923555e-06,
|
| 470 |
+
"loss": 0.2326,
|
| 471 |
+
"step": 77
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.87,
|
| 475 |
+
"learning_rate": 1.8484916065808622e-06,
|
| 476 |
+
"loss": 0.5545,
|
| 477 |
+
"step": 78
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.88,
|
| 481 |
+
"learning_rate": 1.557144461481337e-06,
|
| 482 |
+
"loss": 1.5451,
|
| 483 |
+
"step": 79
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 0.89,
|
| 487 |
+
"learning_rate": 1.2898432814879813e-06,
|
| 488 |
+
"loss": 0.1618,
|
| 489 |
+
"step": 80
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.9,
|
| 493 |
+
"learning_rate": 1.0469365763439532e-06,
|
| 494 |
+
"loss": 0.1308,
|
| 495 |
+
"step": 81
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.91,
|
| 499 |
+
"learning_rate": 8.287410500511739e-07,
|
| 500 |
+
"loss": 0.0495,
|
| 501 |
+
"step": 82
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.92,
|
| 505 |
+
"learning_rate": 6.355411879487339e-07,
|
| 506 |
+
"loss": 0.1456,
|
| 507 |
+
"step": 83
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.93,
|
| 511 |
+
"learning_rate": 4.675888857982669e-07,
|
| 512 |
+
"loss": 0.1224,
|
| 513 |
+
"step": 84
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.94,
|
| 517 |
+
"learning_rate": 3.251031213599465e-07,
|
| 518 |
+
"loss": 0.139,
|
| 519 |
+
"step": 85
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.96,
|
| 523 |
+
"learning_rate": 2.082696688872554e-07,
|
| 524 |
+
"loss": 0.1965,
|
| 525 |
+
"step": 86
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.97,
|
| 529 |
+
"learning_rate": 1.1724085691280806e-07,
|
| 530 |
+
"loss": 0.1047,
|
| 531 |
+
"step": 87
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.98,
|
| 535 |
+
"learning_rate": 5.21353696410043e-08,
|
| 536 |
+
"loss": 0.1254,
|
| 537 |
+
"step": 88
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.99,
|
| 541 |
+
"learning_rate": 1.3038092206461017e-08,
|
| 542 |
+
"loss": 0.1941,
|
| 543 |
+
"step": 89
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.0,
|
| 547 |
+
"learning_rate": 0.0,
|
| 548 |
+
"loss": 0.9494,
|
| 549 |
+
"step": 90
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.0,
|
| 553 |
+
"step": 90,
|
| 554 |
+
"total_flos": 1.4124649781447885e+18,
|
| 555 |
+
"train_loss": 0.3038743817557891,
|
| 556 |
+
"train_runtime": 288.0759,
|
| 557 |
+
"train_samples_per_second": 0.312,
|
| 558 |
+
"train_steps_per_second": 0.312
|
| 559 |
+
}
|
| 560 |
+
],
|
| 561 |
+
"logging_steps": 1.0,
|
| 562 |
+
"max_steps": 90,
|
| 563 |
+
"num_input_tokens_seen": 0,
|
| 564 |
+
"num_train_epochs": 1,
|
| 565 |
+
"save_steps": 200,
|
| 566 |
+
"total_flos": 1.4124649781447885e+18,
|
| 567 |
+
"train_batch_size": 1,
|
| 568 |
+
"trial_name": null,
|
| 569 |
+
"trial_params": null
|
| 570 |
+
}
|