Delete trainer_state.json
Browse files- trainer_state.json +0 -342
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": 1.3249907986750091,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 3600,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0368052999631947,
|
| 14 |
-
"grad_norm": 6.317627906799316,
|
| 15 |
-
"learning_rate": 3.991842559741558e-05,
|
| 16 |
-
"loss": 1.3232,
|
| 17 |
-
"step": 100
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.0736105999263894,
|
| 21 |
-
"grad_norm": 6.4019927978515625,
|
| 22 |
-
"learning_rate": 3.9770967337294366e-05,
|
| 23 |
-
"loss": 1.1983,
|
| 24 |
-
"step": 200
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.1104158998895841,
|
| 28 |
-
"grad_norm": 5.58203649520874,
|
| 29 |
-
"learning_rate": 3.962350907717314e-05,
|
| 30 |
-
"loss": 1.1226,
|
| 31 |
-
"step": 300
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.1472211998527788,
|
| 35 |
-
"grad_norm": 4.505954742431641,
|
| 36 |
-
"learning_rate": 3.947605081705192e-05,
|
| 37 |
-
"loss": 0.7575,
|
| 38 |
-
"step": 400
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.1840264998159735,
|
| 42 |
-
"grad_norm": 5.4588165283203125,
|
| 43 |
-
"learning_rate": 3.93285925569307e-05,
|
| 44 |
-
"loss": 0.684,
|
| 45 |
-
"step": 500
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.1840264998159735,
|
| 49 |
-
"eval_loss": 2.1012208461761475,
|
| 50 |
-
"eval_runtime": 663.3742,
|
| 51 |
-
"eval_samples_per_second": 8.193,
|
| 52 |
-
"eval_steps_per_second": 1.025,
|
| 53 |
-
"step": 500
|
| 54 |
-
},
|
| 55 |
-
{
|
| 56 |
-
"epoch": 0.2208317997791682,
|
| 57 |
-
"grad_norm": 4.757272720336914,
|
| 58 |
-
"learning_rate": 3.918113429680948e-05,
|
| 59 |
-
"loss": 0.6552,
|
| 60 |
-
"step": 600
|
| 61 |
-
},
|
| 62 |
-
{
|
| 63 |
-
"epoch": 0.2576370997423629,
|
| 64 |
-
"grad_norm": 4.794644355773926,
|
| 65 |
-
"learning_rate": 3.903367603668826e-05,
|
| 66 |
-
"loss": 0.6429,
|
| 67 |
-
"step": 700
|
| 68 |
-
},
|
| 69 |
-
{
|
| 70 |
-
"epoch": 0.2944423997055576,
|
| 71 |
-
"grad_norm": 5.151235103607178,
|
| 72 |
-
"learning_rate": 3.888621777656704e-05,
|
| 73 |
-
"loss": 0.6567,
|
| 74 |
-
"step": 800
|
| 75 |
-
},
|
| 76 |
-
{
|
| 77 |
-
"epoch": 0.3312476996687523,
|
| 78 |
-
"grad_norm": 5.006649971008301,
|
| 79 |
-
"learning_rate": 3.8738759516445824e-05,
|
| 80 |
-
"loss": 0.6278,
|
| 81 |
-
"step": 900
|
| 82 |
-
},
|
| 83 |
-
{
|
| 84 |
-
"epoch": 0.368052999631947,
|
| 85 |
-
"grad_norm": 6.711447715759277,
|
| 86 |
-
"learning_rate": 3.85913012563246e-05,
|
| 87 |
-
"loss": 0.637,
|
| 88 |
-
"step": 1000
|
| 89 |
-
},
|
| 90 |
-
{
|
| 91 |
-
"epoch": 0.368052999631947,
|
| 92 |
-
"eval_loss": 2.1562345027923584,
|
| 93 |
-
"eval_runtime": 663.0039,
|
| 94 |
-
"eval_samples_per_second": 8.198,
|
| 95 |
-
"eval_steps_per_second": 1.026,
|
| 96 |
-
"step": 1000
|
| 97 |
-
},
|
| 98 |
-
{
|
| 99 |
-
"epoch": 0.4048582995951417,
|
| 100 |
-
"grad_norm": 5.702431678771973,
|
| 101 |
-
"learning_rate": 3.844384299620338e-05,
|
| 102 |
-
"loss": 0.5867,
|
| 103 |
-
"step": 1100
|
| 104 |
-
},
|
| 105 |
-
{
|
| 106 |
-
"epoch": 0.4416635995583364,
|
| 107 |
-
"grad_norm": 4.582945346832275,
|
| 108 |
-
"learning_rate": 3.829638473608216e-05,
|
| 109 |
-
"loss": 0.6156,
|
| 110 |
-
"step": 1200
|
| 111 |
-
},
|
| 112 |
-
{
|
| 113 |
-
"epoch": 0.4784688995215311,
|
| 114 |
-
"grad_norm": 4.968910217285156,
|
| 115 |
-
"learning_rate": 3.814892647596094e-05,
|
| 116 |
-
"loss": 0.6213,
|
| 117 |
-
"step": 1300
|
| 118 |
-
},
|
| 119 |
-
{
|
| 120 |
-
"epoch": 0.5152741994847257,
|
| 121 |
-
"grad_norm": 5.8628950119018555,
|
| 122 |
-
"learning_rate": 3.800146821583972e-05,
|
| 123 |
-
"loss": 0.6036,
|
| 124 |
-
"step": 1400
|
| 125 |
-
},
|
| 126 |
-
{
|
| 127 |
-
"epoch": 0.5520794994479205,
|
| 128 |
-
"grad_norm": 4.60435676574707,
|
| 129 |
-
"learning_rate": 3.7854009955718494e-05,
|
| 130 |
-
"loss": 0.5947,
|
| 131 |
-
"step": 1500
|
| 132 |
-
},
|
| 133 |
-
{
|
| 134 |
-
"epoch": 0.5520794994479205,
|
| 135 |
-
"eval_loss": 2.2223360538482666,
|
| 136 |
-
"eval_runtime": 662.8599,
|
| 137 |
-
"eval_samples_per_second": 8.199,
|
| 138 |
-
"eval_steps_per_second": 1.026,
|
| 139 |
-
"step": 1500
|
| 140 |
-
},
|
| 141 |
-
{
|
| 142 |
-
"epoch": 0.5888847994111152,
|
| 143 |
-
"grad_norm": 4.219241142272949,
|
| 144 |
-
"learning_rate": 3.7706551695597275e-05,
|
| 145 |
-
"loss": 0.61,
|
| 146 |
-
"step": 1600
|
| 147 |
-
},
|
| 148 |
-
{
|
| 149 |
-
"epoch": 0.6256900993743099,
|
| 150 |
-
"grad_norm": 4.9983344078063965,
|
| 151 |
-
"learning_rate": 3.7559093435476055e-05,
|
| 152 |
-
"loss": 0.5801,
|
| 153 |
-
"step": 1700
|
| 154 |
-
},
|
| 155 |
-
{
|
| 156 |
-
"epoch": 0.6624953993375046,
|
| 157 |
-
"grad_norm": 5.795677185058594,
|
| 158 |
-
"learning_rate": 3.7411635175354836e-05,
|
| 159 |
-
"loss": 0.6016,
|
| 160 |
-
"step": 1800
|
| 161 |
-
},
|
| 162 |
-
{
|
| 163 |
-
"epoch": 0.6993006993006993,
|
| 164 |
-
"grad_norm": 4.981507778167725,
|
| 165 |
-
"learning_rate": 3.726417691523362e-05,
|
| 166 |
-
"loss": 0.5839,
|
| 167 |
-
"step": 1900
|
| 168 |
-
},
|
| 169 |
-
{
|
| 170 |
-
"epoch": 0.736105999263894,
|
| 171 |
-
"grad_norm": 5.115480899810791,
|
| 172 |
-
"learning_rate": 3.71167186551124e-05,
|
| 173 |
-
"loss": 0.5946,
|
| 174 |
-
"step": 2000
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 0.736105999263894,
|
| 178 |
-
"eval_loss": 2.1914401054382324,
|
| 179 |
-
"eval_runtime": 663.2825,
|
| 180 |
-
"eval_samples_per_second": 8.194,
|
| 181 |
-
"eval_steps_per_second": 1.025,
|
| 182 |
-
"step": 2000
|
| 183 |
-
},
|
| 184 |
-
{
|
| 185 |
-
"epoch": 0.7729112992270887,
|
| 186 |
-
"grad_norm": 4.6113176345825195,
|
| 187 |
-
"learning_rate": 3.696926039499117e-05,
|
| 188 |
-
"loss": 0.5763,
|
| 189 |
-
"step": 2100
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 0.8097165991902834,
|
| 193 |
-
"grad_norm": 4.699350833892822,
|
| 194 |
-
"learning_rate": 3.682180213486995e-05,
|
| 195 |
-
"loss": 0.6007,
|
| 196 |
-
"step": 2200
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 0.8465218991534781,
|
| 200 |
-
"grad_norm": 4.8883233070373535,
|
| 201 |
-
"learning_rate": 3.667434387474873e-05,
|
| 202 |
-
"loss": 0.5741,
|
| 203 |
-
"step": 2300
|
| 204 |
-
},
|
| 205 |
-
{
|
| 206 |
-
"epoch": 0.8833271991166728,
|
| 207 |
-
"grad_norm": 5.460277557373047,
|
| 208 |
-
"learning_rate": 3.652688561462751e-05,
|
| 209 |
-
"loss": 0.5596,
|
| 210 |
-
"step": 2400
|
| 211 |
-
},
|
| 212 |
-
{
|
| 213 |
-
"epoch": 0.9201324990798675,
|
| 214 |
-
"grad_norm": 4.431008338928223,
|
| 215 |
-
"learning_rate": 3.6379427354506294e-05,
|
| 216 |
-
"loss": 0.5831,
|
| 217 |
-
"step": 2500
|
| 218 |
-
},
|
| 219 |
-
{
|
| 220 |
-
"epoch": 0.9201324990798675,
|
| 221 |
-
"eval_loss": 2.2241196632385254,
|
| 222 |
-
"eval_runtime": 662.9801,
|
| 223 |
-
"eval_samples_per_second": 8.198,
|
| 224 |
-
"eval_steps_per_second": 1.026,
|
| 225 |
-
"step": 2500
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.9569377990430622,
|
| 229 |
-
"grad_norm": 4.917581081390381,
|
| 230 |
-
"learning_rate": 3.6231969094385074e-05,
|
| 231 |
-
"loss": 0.5956,
|
| 232 |
-
"step": 2600
|
| 233 |
-
},
|
| 234 |
-
{
|
| 235 |
-
"epoch": 0.9937430990062569,
|
| 236 |
-
"grad_norm": 5.325926780700684,
|
| 237 |
-
"learning_rate": 3.6084510834263855e-05,
|
| 238 |
-
"loss": 0.6783,
|
| 239 |
-
"step": 2700
|
| 240 |
-
},
|
| 241 |
-
{
|
| 242 |
-
"epoch": 1.0305483989694515,
|
| 243 |
-
"grad_norm": 3.779780149459839,
|
| 244 |
-
"learning_rate": 3.593705257414263e-05,
|
| 245 |
-
"loss": 0.356,
|
| 246 |
-
"step": 2800
|
| 247 |
-
},
|
| 248 |
-
{
|
| 249 |
-
"epoch": 1.0673536989326462,
|
| 250 |
-
"grad_norm": 7.602641582489014,
|
| 251 |
-
"learning_rate": 3.5789594314021416e-05,
|
| 252 |
-
"loss": 0.2066,
|
| 253 |
-
"step": 2900
|
| 254 |
-
},
|
| 255 |
-
{
|
| 256 |
-
"epoch": 1.104158998895841,
|
| 257 |
-
"grad_norm": 5.734857082366943,
|
| 258 |
-
"learning_rate": 3.564213605390019e-05,
|
| 259 |
-
"loss": 0.2202,
|
| 260 |
-
"step": 3000
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"epoch": 1.104158998895841,
|
| 264 |
-
"eval_loss": 2.732856512069702,
|
| 265 |
-
"eval_runtime": 663.1541,
|
| 266 |
-
"eval_samples_per_second": 8.196,
|
| 267 |
-
"eval_steps_per_second": 1.025,
|
| 268 |
-
"step": 3000
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 1.1409642988590356,
|
| 272 |
-
"grad_norm": 5.713295936584473,
|
| 273 |
-
"learning_rate": 3.549467779377897e-05,
|
| 274 |
-
"loss": 0.6849,
|
| 275 |
-
"step": 3100
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 1.1777695988222303,
|
| 279 |
-
"grad_norm": 8.02027702331543,
|
| 280 |
-
"learning_rate": 3.534721953365775e-05,
|
| 281 |
-
"loss": 0.6996,
|
| 282 |
-
"step": 3200
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 1.214574898785425,
|
| 286 |
-
"grad_norm": 7.312982082366943,
|
| 287 |
-
"learning_rate": 3.5199761273536525e-05,
|
| 288 |
-
"loss": 0.6721,
|
| 289 |
-
"step": 3300
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 1.2513801987486197,
|
| 293 |
-
"grad_norm": 6.537501811981201,
|
| 294 |
-
"learning_rate": 3.505230301341531e-05,
|
| 295 |
-
"loss": 0.6689,
|
| 296 |
-
"step": 3400
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 1.2881854987118144,
|
| 300 |
-
"grad_norm": 6.13110876083374,
|
| 301 |
-
"learning_rate": 3.4904844753294086e-05,
|
| 302 |
-
"loss": 0.6943,
|
| 303 |
-
"step": 3500
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 1.2881854987118144,
|
| 307 |
-
"eval_loss": 2.2627553939819336,
|
| 308 |
-
"eval_runtime": 662.7328,
|
| 309 |
-
"eval_samples_per_second": 8.201,
|
| 310 |
-
"eval_steps_per_second": 1.026,
|
| 311 |
-
"step": 3500
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 1.3249907986750091,
|
| 315 |
-
"grad_norm": 7.125514984130859,
|
| 316 |
-
"learning_rate": 3.4757386493172874e-05,
|
| 317 |
-
"loss": 0.6925,
|
| 318 |
-
"step": 3600
|
| 319 |
-
}
|
| 320 |
-
],
|
| 321 |
-
"logging_steps": 100,
|
| 322 |
-
"max_steps": 27170,
|
| 323 |
-
"num_input_tokens_seen": 0,
|
| 324 |
-
"num_train_epochs": 10,
|
| 325 |
-
"save_steps": 100,
|
| 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": 1.001052086206464e+17,
|
| 339 |
-
"train_batch_size": 8,
|
| 340 |
-
"trial_name": null,
|
| 341 |
-
"trial_params": null
|
| 342 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|