Delete checkpoint-7000/trainer_state.json
Browse files
checkpoint-7000/trainer_state.json
DELETED
|
@@ -1,342 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.7579859231185706,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 7000,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.02707092582566324,
|
| 14 |
-
"grad_norm": 0.9713733196258545,
|
| 15 |
-
"learning_rate": 9.910124526258799e-05,
|
| 16 |
-
"loss": 3.4945,
|
| 17 |
-
"step": 250
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.05414185165132648,
|
| 21 |
-
"grad_norm": 0.9521434307098389,
|
| 22 |
-
"learning_rate": 9.819888106839921e-05,
|
| 23 |
-
"loss": 3.5097,
|
| 24 |
-
"step": 500
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.05414185165132648,
|
| 28 |
-
"eval_loss": 2.95473575592041,
|
| 29 |
-
"eval_runtime": 87.829,
|
| 30 |
-
"eval_samples_per_second": 112.15,
|
| 31 |
-
"eval_steps_per_second": 7.014,
|
| 32 |
-
"step": 500
|
| 33 |
-
},
|
| 34 |
-
{
|
| 35 |
-
"epoch": 0.08121277747698971,
|
| 36 |
-
"grad_norm": 0.9334385395050049,
|
| 37 |
-
"learning_rate": 9.729651687421044e-05,
|
| 38 |
-
"loss": 3.4922,
|
| 39 |
-
"step": 750
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 0.10828370330265295,
|
| 43 |
-
"grad_norm": 1.2239603996276855,
|
| 44 |
-
"learning_rate": 9.639415268002166e-05,
|
| 45 |
-
"loss": 3.477,
|
| 46 |
-
"step": 1000
|
| 47 |
-
},
|
| 48 |
-
{
|
| 49 |
-
"epoch": 0.10828370330265295,
|
| 50 |
-
"eval_loss": 2.9319565296173096,
|
| 51 |
-
"eval_runtime": 87.8964,
|
| 52 |
-
"eval_samples_per_second": 112.064,
|
| 53 |
-
"eval_steps_per_second": 7.008,
|
| 54 |
-
"step": 1000
|
| 55 |
-
},
|
| 56 |
-
{
|
| 57 |
-
"epoch": 0.1353546291283162,
|
| 58 |
-
"grad_norm": 1.1727226972579956,
|
| 59 |
-
"learning_rate": 9.549178848583288e-05,
|
| 60 |
-
"loss": 3.493,
|
| 61 |
-
"step": 1250
|
| 62 |
-
},
|
| 63 |
-
{
|
| 64 |
-
"epoch": 0.16242555495397942,
|
| 65 |
-
"grad_norm": 0.966973602771759,
|
| 66 |
-
"learning_rate": 9.458942429164411e-05,
|
| 67 |
-
"loss": 3.5013,
|
| 68 |
-
"step": 1500
|
| 69 |
-
},
|
| 70 |
-
{
|
| 71 |
-
"epoch": 0.16242555495397942,
|
| 72 |
-
"eval_loss": 2.9275825023651123,
|
| 73 |
-
"eval_runtime": 87.899,
|
| 74 |
-
"eval_samples_per_second": 112.06,
|
| 75 |
-
"eval_steps_per_second": 7.008,
|
| 76 |
-
"step": 1500
|
| 77 |
-
},
|
| 78 |
-
{
|
| 79 |
-
"epoch": 0.18949648077964265,
|
| 80 |
-
"grad_norm": 0.9813922047615051,
|
| 81 |
-
"learning_rate": 9.368706009745533e-05,
|
| 82 |
-
"loss": 3.4907,
|
| 83 |
-
"step": 1750
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 0.2165674066053059,
|
| 87 |
-
"grad_norm": 1.027085542678833,
|
| 88 |
-
"learning_rate": 9.278469590326656e-05,
|
| 89 |
-
"loss": 3.4588,
|
| 90 |
-
"step": 2000
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.2165674066053059,
|
| 94 |
-
"eval_loss": 2.9044992923736572,
|
| 95 |
-
"eval_runtime": 87.8776,
|
| 96 |
-
"eval_samples_per_second": 112.088,
|
| 97 |
-
"eval_steps_per_second": 7.01,
|
| 98 |
-
"step": 2000
|
| 99 |
-
},
|
| 100 |
-
{
|
| 101 |
-
"epoch": 0.24363833243096913,
|
| 102 |
-
"grad_norm": 1.0764214992523193,
|
| 103 |
-
"learning_rate": 9.188233170907778e-05,
|
| 104 |
-
"loss": 3.4531,
|
| 105 |
-
"step": 2250
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 0.2707092582566324,
|
| 109 |
-
"grad_norm": 1.0297119617462158,
|
| 110 |
-
"learning_rate": 9.0979967514889e-05,
|
| 111 |
-
"loss": 3.4445,
|
| 112 |
-
"step": 2500
|
| 113 |
-
},
|
| 114 |
-
{
|
| 115 |
-
"epoch": 0.2707092582566324,
|
| 116 |
-
"eval_loss": 2.8855738639831543,
|
| 117 |
-
"eval_runtime": 87.8728,
|
| 118 |
-
"eval_samples_per_second": 112.094,
|
| 119 |
-
"eval_steps_per_second": 7.01,
|
| 120 |
-
"step": 2500
|
| 121 |
-
},
|
| 122 |
-
{
|
| 123 |
-
"epoch": 0.2977801840822956,
|
| 124 |
-
"grad_norm": 0.9697523713111877,
|
| 125 |
-
"learning_rate": 9.007760332070024e-05,
|
| 126 |
-
"loss": 3.4349,
|
| 127 |
-
"step": 2750
|
| 128 |
-
},
|
| 129 |
-
{
|
| 130 |
-
"epoch": 0.32485110990795885,
|
| 131 |
-
"grad_norm": 0.9611329436302185,
|
| 132 |
-
"learning_rate": 8.917523912651147e-05,
|
| 133 |
-
"loss": 3.4213,
|
| 134 |
-
"step": 3000
|
| 135 |
-
},
|
| 136 |
-
{
|
| 137 |
-
"epoch": 0.32485110990795885,
|
| 138 |
-
"eval_loss": 2.8725759983062744,
|
| 139 |
-
"eval_runtime": 87.9054,
|
| 140 |
-
"eval_samples_per_second": 112.052,
|
| 141 |
-
"eval_steps_per_second": 7.008,
|
| 142 |
-
"step": 3000
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.3519220357336221,
|
| 146 |
-
"grad_norm": 1.000977873802185,
|
| 147 |
-
"learning_rate": 8.827287493232269e-05,
|
| 148 |
-
"loss": 3.4239,
|
| 149 |
-
"step": 3250
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.3789929615592853,
|
| 153 |
-
"grad_norm": 1.0573837757110596,
|
| 154 |
-
"learning_rate": 8.737051073813391e-05,
|
| 155 |
-
"loss": 3.4015,
|
| 156 |
-
"step": 3500
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.3789929615592853,
|
| 160 |
-
"eval_loss": 2.8721706867218018,
|
| 161 |
-
"eval_runtime": 87.8848,
|
| 162 |
-
"eval_samples_per_second": 112.078,
|
| 163 |
-
"eval_steps_per_second": 7.009,
|
| 164 |
-
"step": 3500
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.4060638873849486,
|
| 168 |
-
"grad_norm": 0.9532322883605957,
|
| 169 |
-
"learning_rate": 8.646814654394514e-05,
|
| 170 |
-
"loss": 3.401,
|
| 171 |
-
"step": 3750
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.4331348132106118,
|
| 175 |
-
"grad_norm": 1.0036464929580688,
|
| 176 |
-
"learning_rate": 8.556578234975636e-05,
|
| 177 |
-
"loss": 3.3984,
|
| 178 |
-
"step": 4000
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.4331348132106118,
|
| 182 |
-
"eval_loss": 2.863417863845825,
|
| 183 |
-
"eval_runtime": 87.8316,
|
| 184 |
-
"eval_samples_per_second": 112.146,
|
| 185 |
-
"eval_steps_per_second": 7.013,
|
| 186 |
-
"step": 4000
|
| 187 |
-
},
|
| 188 |
-
{
|
| 189 |
-
"epoch": 0.46020573903627504,
|
| 190 |
-
"grad_norm": 0.8860335350036621,
|
| 191 |
-
"learning_rate": 8.466341815556758e-05,
|
| 192 |
-
"loss": 3.4081,
|
| 193 |
-
"step": 4250
|
| 194 |
-
},
|
| 195 |
-
{
|
| 196 |
-
"epoch": 0.48727666486193827,
|
| 197 |
-
"grad_norm": 1.4852999448776245,
|
| 198 |
-
"learning_rate": 8.376105396137882e-05,
|
| 199 |
-
"loss": 3.4105,
|
| 200 |
-
"step": 4500
|
| 201 |
-
},
|
| 202 |
-
{
|
| 203 |
-
"epoch": 0.48727666486193827,
|
| 204 |
-
"eval_loss": 2.8409109115600586,
|
| 205 |
-
"eval_runtime": 87.9025,
|
| 206 |
-
"eval_samples_per_second": 112.056,
|
| 207 |
-
"eval_steps_per_second": 7.008,
|
| 208 |
-
"step": 4500
|
| 209 |
-
},
|
| 210 |
-
{
|
| 211 |
-
"epoch": 0.5143475906876015,
|
| 212 |
-
"grad_norm": 0.936912477016449,
|
| 213 |
-
"learning_rate": 8.285868976719005e-05,
|
| 214 |
-
"loss": 3.3952,
|
| 215 |
-
"step": 4750
|
| 216 |
-
},
|
| 217 |
-
{
|
| 218 |
-
"epoch": 0.5414185165132648,
|
| 219 |
-
"grad_norm": 1.0409356355667114,
|
| 220 |
-
"learning_rate": 8.195632557300127e-05,
|
| 221 |
-
"loss": 3.4039,
|
| 222 |
-
"step": 5000
|
| 223 |
-
},
|
| 224 |
-
{
|
| 225 |
-
"epoch": 0.5414185165132648,
|
| 226 |
-
"eval_loss": 2.8329713344573975,
|
| 227 |
-
"eval_runtime": 87.8287,
|
| 228 |
-
"eval_samples_per_second": 112.15,
|
| 229 |
-
"eval_steps_per_second": 7.014,
|
| 230 |
-
"step": 5000
|
| 231 |
-
},
|
| 232 |
-
{
|
| 233 |
-
"epoch": 0.568489442338928,
|
| 234 |
-
"grad_norm": 0.9606480002403259,
|
| 235 |
-
"learning_rate": 8.10539613788125e-05,
|
| 236 |
-
"loss": 3.3819,
|
| 237 |
-
"step": 5250
|
| 238 |
-
},
|
| 239 |
-
{
|
| 240 |
-
"epoch": 0.5955603681645912,
|
| 241 |
-
"grad_norm": 1.142527461051941,
|
| 242 |
-
"learning_rate": 8.015159718462372e-05,
|
| 243 |
-
"loss": 3.3898,
|
| 244 |
-
"step": 5500
|
| 245 |
-
},
|
| 246 |
-
{
|
| 247 |
-
"epoch": 0.5955603681645912,
|
| 248 |
-
"eval_loss": 2.8344321250915527,
|
| 249 |
-
"eval_runtime": 87.8057,
|
| 250 |
-
"eval_samples_per_second": 112.18,
|
| 251 |
-
"eval_steps_per_second": 7.015,
|
| 252 |
-
"step": 5500
|
| 253 |
-
},
|
| 254 |
-
{
|
| 255 |
-
"epoch": 0.6226312939902545,
|
| 256 |
-
"grad_norm": 1.2294474840164185,
|
| 257 |
-
"learning_rate": 7.924923299043494e-05,
|
| 258 |
-
"loss": 3.3769,
|
| 259 |
-
"step": 5750
|
| 260 |
-
},
|
| 261 |
-
{
|
| 262 |
-
"epoch": 0.6497022198159177,
|
| 263 |
-
"grad_norm": 1.0300960540771484,
|
| 264 |
-
"learning_rate": 7.834686879624617e-05,
|
| 265 |
-
"loss": 3.3621,
|
| 266 |
-
"step": 6000
|
| 267 |
-
},
|
| 268 |
-
{
|
| 269 |
-
"epoch": 0.6497022198159177,
|
| 270 |
-
"eval_loss": 2.8137757778167725,
|
| 271 |
-
"eval_runtime": 87.8367,
|
| 272 |
-
"eval_samples_per_second": 112.14,
|
| 273 |
-
"eval_steps_per_second": 7.013,
|
| 274 |
-
"step": 6000
|
| 275 |
-
},
|
| 276 |
-
{
|
| 277 |
-
"epoch": 0.676773145641581,
|
| 278 |
-
"grad_norm": 1.076827883720398,
|
| 279 |
-
"learning_rate": 7.74445046020574e-05,
|
| 280 |
-
"loss": 3.3704,
|
| 281 |
-
"step": 6250
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"epoch": 0.7038440714672441,
|
| 285 |
-
"grad_norm": 1.4118528366088867,
|
| 286 |
-
"learning_rate": 7.654214040786863e-05,
|
| 287 |
-
"loss": 3.367,
|
| 288 |
-
"step": 6500
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"epoch": 0.7038440714672441,
|
| 292 |
-
"eval_loss": 2.8218181133270264,
|
| 293 |
-
"eval_runtime": 87.9044,
|
| 294 |
-
"eval_samples_per_second": 112.053,
|
| 295 |
-
"eval_steps_per_second": 7.008,
|
| 296 |
-
"step": 6500
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.7309149972929074,
|
| 300 |
-
"grad_norm": 1.0568158626556396,
|
| 301 |
-
"learning_rate": 7.563977621367985e-05,
|
| 302 |
-
"loss": 3.3509,
|
| 303 |
-
"step": 6750
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.7579859231185706,
|
| 307 |
-
"grad_norm": 1.0436064004898071,
|
| 308 |
-
"learning_rate": 7.473741201949108e-05,
|
| 309 |
-
"loss": 3.3496,
|
| 310 |
-
"step": 7000
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.7579859231185706,
|
| 314 |
-
"eval_loss": 2.8048973083496094,
|
| 315 |
-
"eval_runtime": 87.8448,
|
| 316 |
-
"eval_samples_per_second": 112.13,
|
| 317 |
-
"eval_steps_per_second": 7.012,
|
| 318 |
-
"step": 7000
|
| 319 |
-
}
|
| 320 |
-
],
|
| 321 |
-
"logging_steps": 250,
|
| 322 |
-
"max_steps": 27705,
|
| 323 |
-
"num_input_tokens_seen": 0,
|
| 324 |
-
"num_train_epochs": 3,
|
| 325 |
-
"save_steps": 1000,
|
| 326 |
-
"stateful_callbacks": {
|
| 327 |
-
"TrainerControl": {
|
| 328 |
-
"args": {
|
| 329 |
-
"should_epoch_stop": false,
|
| 330 |
-
"should_evaluate": false,
|
| 331 |
-
"should_log": false,
|
| 332 |
-
"should_save": true,
|
| 333 |
-
"should_training_stop": false
|
| 334 |
-
},
|
| 335 |
-
"attributes": {}
|
| 336 |
-
}
|
| 337 |
-
},
|
| 338 |
-
"total_flos": 2527378538496000.0,
|
| 339 |
-
"train_batch_size": 16,
|
| 340 |
-
"trial_name": null,
|
| 341 |
-
"trial_params": null
|
| 342 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|