Upload trainer_state.json
Browse files- trainer_state.json +344 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,344 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.3929173693086,
|
| 5 |
+
"global_step": 11500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.84,
|
| 12 |
+
"learning_rate": 9.578414839797639e-06,
|
| 13 |
+
"loss": 0.3136,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_bleu": 17.1673,
|
| 19 |
+
"eval_gen_len": 43.5832,
|
| 20 |
+
"eval_loss": 2.7004430294036865,
|
| 21 |
+
"eval_runtime": 263.1603,
|
| 22 |
+
"eval_samples_per_second": 3.792,
|
| 23 |
+
"eval_steps_per_second": 0.475,
|
| 24 |
+
"step": 593
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.69,
|
| 28 |
+
"learning_rate": 9.156829679595279e-06,
|
| 29 |
+
"loss": 0.2962,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.0,
|
| 34 |
+
"eval_bleu": 17.2012,
|
| 35 |
+
"eval_gen_len": 43.6693,
|
| 36 |
+
"eval_loss": 2.719139814376831,
|
| 37 |
+
"eval_runtime": 263.247,
|
| 38 |
+
"eval_samples_per_second": 3.791,
|
| 39 |
+
"eval_steps_per_second": 0.475,
|
| 40 |
+
"step": 1186
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 2.53,
|
| 44 |
+
"learning_rate": 8.735244519392918e-06,
|
| 45 |
+
"loss": 0.2927,
|
| 46 |
+
"step": 1500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 3.0,
|
| 50 |
+
"eval_bleu": 17.2291,
|
| 51 |
+
"eval_gen_len": 43.482,
|
| 52 |
+
"eval_loss": 2.7412936687469482,
|
| 53 |
+
"eval_runtime": 260.7955,
|
| 54 |
+
"eval_samples_per_second": 3.827,
|
| 55 |
+
"eval_steps_per_second": 0.479,
|
| 56 |
+
"step": 1779
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 3.37,
|
| 60 |
+
"learning_rate": 8.313659359190556e-06,
|
| 61 |
+
"loss": 0.2677,
|
| 62 |
+
"step": 2000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"eval_bleu": 17.135,
|
| 67 |
+
"eval_gen_len": 43.5862,
|
| 68 |
+
"eval_loss": 2.7617862224578857,
|
| 69 |
+
"eval_runtime": 261.1749,
|
| 70 |
+
"eval_samples_per_second": 3.821,
|
| 71 |
+
"eval_steps_per_second": 0.479,
|
| 72 |
+
"step": 2372
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 4.22,
|
| 76 |
+
"learning_rate": 7.892074198988196e-06,
|
| 77 |
+
"loss": 0.2591,
|
| 78 |
+
"step": 2500
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 5.0,
|
| 82 |
+
"eval_bleu": 17.5543,
|
| 83 |
+
"eval_gen_len": 43.5922,
|
| 84 |
+
"eval_loss": 2.7780115604400635,
|
| 85 |
+
"eval_runtime": 262.647,
|
| 86 |
+
"eval_samples_per_second": 3.8,
|
| 87 |
+
"eval_steps_per_second": 0.476,
|
| 88 |
+
"step": 2965
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 5.06,
|
| 92 |
+
"learning_rate": 7.470489038785835e-06,
|
| 93 |
+
"loss": 0.2473,
|
| 94 |
+
"step": 3000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 5.9,
|
| 98 |
+
"learning_rate": 7.048903878583474e-06,
|
| 99 |
+
"loss": 0.2282,
|
| 100 |
+
"step": 3500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 6.0,
|
| 104 |
+
"eval_bleu": 17.226,
|
| 105 |
+
"eval_gen_len": 43.6703,
|
| 106 |
+
"eval_loss": 2.794311761856079,
|
| 107 |
+
"eval_runtime": 263.8826,
|
| 108 |
+
"eval_samples_per_second": 3.782,
|
| 109 |
+
"eval_steps_per_second": 0.474,
|
| 110 |
+
"step": 3558
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 6.75,
|
| 114 |
+
"learning_rate": 6.6273187183811136e-06,
|
| 115 |
+
"loss": 0.2244,
|
| 116 |
+
"step": 4000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 7.0,
|
| 120 |
+
"eval_bleu": 17.615,
|
| 121 |
+
"eval_gen_len": 43.6934,
|
| 122 |
+
"eval_loss": 2.808680295944214,
|
| 123 |
+
"eval_runtime": 264.0527,
|
| 124 |
+
"eval_samples_per_second": 3.78,
|
| 125 |
+
"eval_steps_per_second": 0.473,
|
| 126 |
+
"step": 4151
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 7.59,
|
| 130 |
+
"learning_rate": 6.2057335581787524e-06,
|
| 131 |
+
"loss": 0.2196,
|
| 132 |
+
"step": 4500
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 8.0,
|
| 136 |
+
"eval_bleu": 17.3227,
|
| 137 |
+
"eval_gen_len": 43.7715,
|
| 138 |
+
"eval_loss": 2.825133800506592,
|
| 139 |
+
"eval_runtime": 281.974,
|
| 140 |
+
"eval_samples_per_second": 3.539,
|
| 141 |
+
"eval_steps_per_second": 0.443,
|
| 142 |
+
"step": 4744
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 8.43,
|
| 146 |
+
"learning_rate": 5.784148397976391e-06,
|
| 147 |
+
"loss": 0.2101,
|
| 148 |
+
"step": 5000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 9.0,
|
| 152 |
+
"eval_bleu": 17.5072,
|
| 153 |
+
"eval_gen_len": 43.7084,
|
| 154 |
+
"eval_loss": 2.834676742553711,
|
| 155 |
+
"eval_runtime": 263.2231,
|
| 156 |
+
"eval_samples_per_second": 3.791,
|
| 157 |
+
"eval_steps_per_second": 0.475,
|
| 158 |
+
"step": 5337
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 9.27,
|
| 162 |
+
"learning_rate": 5.362563237774031e-06,
|
| 163 |
+
"loss": 0.2077,
|
| 164 |
+
"step": 5500
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 10.0,
|
| 168 |
+
"eval_bleu": 17.5712,
|
| 169 |
+
"eval_gen_len": 43.8597,
|
| 170 |
+
"eval_loss": 2.842376708984375,
|
| 171 |
+
"eval_runtime": 270.4686,
|
| 172 |
+
"eval_samples_per_second": 3.69,
|
| 173 |
+
"eval_steps_per_second": 0.462,
|
| 174 |
+
"step": 5930
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 10.12,
|
| 178 |
+
"learning_rate": 4.94097807757167e-06,
|
| 179 |
+
"loss": 0.2034,
|
| 180 |
+
"step": 6000
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 10.96,
|
| 184 |
+
"learning_rate": 4.519392917369309e-06,
|
| 185 |
+
"loss": 0.1968,
|
| 186 |
+
"step": 6500
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 11.0,
|
| 190 |
+
"eval_bleu": 17.6007,
|
| 191 |
+
"eval_gen_len": 43.6994,
|
| 192 |
+
"eval_loss": 2.851884365081787,
|
| 193 |
+
"eval_runtime": 261.9767,
|
| 194 |
+
"eval_samples_per_second": 3.809,
|
| 195 |
+
"eval_steps_per_second": 0.477,
|
| 196 |
+
"step": 6523
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 11.8,
|
| 200 |
+
"learning_rate": 4.097807757166948e-06,
|
| 201 |
+
"loss": 0.1902,
|
| 202 |
+
"step": 7000
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 12.0,
|
| 206 |
+
"eval_bleu": 17.6333,
|
| 207 |
+
"eval_gen_len": 43.6924,
|
| 208 |
+
"eval_loss": 2.8614132404327393,
|
| 209 |
+
"eval_runtime": 263.2972,
|
| 210 |
+
"eval_samples_per_second": 3.79,
|
| 211 |
+
"eval_steps_per_second": 0.475,
|
| 212 |
+
"step": 7116
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 12.65,
|
| 216 |
+
"learning_rate": 3.676222596964587e-06,
|
| 217 |
+
"loss": 0.198,
|
| 218 |
+
"step": 7500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 13.0,
|
| 222 |
+
"eval_bleu": 17.6153,
|
| 223 |
+
"eval_gen_len": 43.7034,
|
| 224 |
+
"eval_loss": 2.865877866744995,
|
| 225 |
+
"eval_runtime": 261.219,
|
| 226 |
+
"eval_samples_per_second": 3.821,
|
| 227 |
+
"eval_steps_per_second": 0.479,
|
| 228 |
+
"step": 7709
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 13.49,
|
| 232 |
+
"learning_rate": 3.2546374367622263e-06,
|
| 233 |
+
"loss": 0.1861,
|
| 234 |
+
"step": 8000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 14.0,
|
| 238 |
+
"eval_bleu": 17.5959,
|
| 239 |
+
"eval_gen_len": 43.7154,
|
| 240 |
+
"eval_loss": 2.873347043991089,
|
| 241 |
+
"eval_runtime": 260.1505,
|
| 242 |
+
"eval_samples_per_second": 3.836,
|
| 243 |
+
"eval_steps_per_second": 0.48,
|
| 244 |
+
"step": 8302
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 14.33,
|
| 248 |
+
"learning_rate": 2.8330522765598656e-06,
|
| 249 |
+
"loss": 0.1956,
|
| 250 |
+
"step": 8500
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 15.0,
|
| 254 |
+
"eval_bleu": 17.6169,
|
| 255 |
+
"eval_gen_len": 43.7164,
|
| 256 |
+
"eval_loss": 2.876323938369751,
|
| 257 |
+
"eval_runtime": 261.1714,
|
| 258 |
+
"eval_samples_per_second": 3.821,
|
| 259 |
+
"eval_steps_per_second": 0.479,
|
| 260 |
+
"step": 8895
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 15.18,
|
| 264 |
+
"learning_rate": 2.4114671163575045e-06,
|
| 265 |
+
"loss": 0.1924,
|
| 266 |
+
"step": 9000
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 16.0,
|
| 270 |
+
"eval_bleu": 17.5443,
|
| 271 |
+
"eval_gen_len": 43.7194,
|
| 272 |
+
"eval_loss": 2.880269765853882,
|
| 273 |
+
"eval_runtime": 261.8101,
|
| 274 |
+
"eval_samples_per_second": 3.812,
|
| 275 |
+
"eval_steps_per_second": 0.477,
|
| 276 |
+
"step": 9488
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 16.02,
|
| 280 |
+
"learning_rate": 1.9898819561551434e-06,
|
| 281 |
+
"loss": 0.1946,
|
| 282 |
+
"step": 9500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 16.86,
|
| 286 |
+
"learning_rate": 1.5682967959527825e-06,
|
| 287 |
+
"loss": 0.1946,
|
| 288 |
+
"step": 10000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 17.0,
|
| 292 |
+
"eval_bleu": 17.577,
|
| 293 |
+
"eval_gen_len": 43.6142,
|
| 294 |
+
"eval_loss": 2.8834807872772217,
|
| 295 |
+
"eval_runtime": 259.2401,
|
| 296 |
+
"eval_samples_per_second": 3.85,
|
| 297 |
+
"eval_steps_per_second": 0.482,
|
| 298 |
+
"step": 10081
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 17.71,
|
| 302 |
+
"learning_rate": 1.1467116357504218e-06,
|
| 303 |
+
"loss": 0.1987,
|
| 304 |
+
"step": 10500
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 18.0,
|
| 308 |
+
"eval_bleu": 17.5677,
|
| 309 |
+
"eval_gen_len": 43.6623,
|
| 310 |
+
"eval_loss": 2.8818464279174805,
|
| 311 |
+
"eval_runtime": 269.8962,
|
| 312 |
+
"eval_samples_per_second": 3.698,
|
| 313 |
+
"eval_steps_per_second": 0.463,
|
| 314 |
+
"step": 10674
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 18.55,
|
| 318 |
+
"learning_rate": 7.251264755480608e-07,
|
| 319 |
+
"loss": 0.2011,
|
| 320 |
+
"step": 11000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 19.0,
|
| 324 |
+
"eval_bleu": 17.6118,
|
| 325 |
+
"eval_gen_len": 43.7395,
|
| 326 |
+
"eval_loss": 2.882765531539917,
|
| 327 |
+
"eval_runtime": 265.3175,
|
| 328 |
+
"eval_samples_per_second": 3.762,
|
| 329 |
+
"eval_steps_per_second": 0.471,
|
| 330 |
+
"step": 11267
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 19.39,
|
| 334 |
+
"learning_rate": 3.0354131534569986e-07,
|
| 335 |
+
"loss": 0.2049,
|
| 336 |
+
"step": 11500
|
| 337 |
+
}
|
| 338 |
+
],
|
| 339 |
+
"max_steps": 11860,
|
| 340 |
+
"num_train_epochs": 20,
|
| 341 |
+
"total_flos": 1623592378957824.0,
|
| 342 |
+
"trial_name": null,
|
| 343 |
+
"trial_params": null
|
| 344 |
+
}
|