End of training
Browse files- all_results.json +6 -6
- train_results.json +6 -6
- trainer_state.json +1942 -255
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 1.3877265500181135e+19,
|
| 4 |
+
"train_loss": 0.3087472263830049,
|
| 5 |
+
"train_runtime": 6534.1253,
|
| 6 |
+
"train_samples_per_second": 27.407,
|
| 7 |
+
"train_steps_per_second": 0.429
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 1.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 1.3877265500181135e+19,
|
| 4 |
+
"train_loss": 0.3087472263830049,
|
| 5 |
+
"train_runtime": 6534.1253,
|
| 6 |
+
"train_samples_per_second": 27.407,
|
| 7 |
+
"train_steps_per_second": 0.429
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,388 +1,2075 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "tsec_vit_model/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 1.
|
| 15 |
-
"loss": 0.
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate":
|
| 22 |
-
"loss": 0.
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate":
|
| 29 |
-
"loss": 0.
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"
|
| 35 |
-
"
|
| 36 |
-
"
|
| 37 |
-
"eval_samples_per_second": 147.012,
|
| 38 |
-
"eval_steps_per_second": 9.305,
|
| 39 |
-
"step": 39
|
| 40 |
-
},
|
| 41 |
-
{
|
| 42 |
-
"epoch": 1.0126582278481013,
|
| 43 |
-
"grad_norm": 2.1996357440948486,
|
| 44 |
-
"learning_rate": 4.985754985754986e-05,
|
| 45 |
-
"loss": 0.5522,
|
| 46 |
"step": 40
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch":
|
| 50 |
-
"grad_norm": 1.
|
| 51 |
-
"learning_rate":
|
| 52 |
-
"loss": 0.
|
| 53 |
"step": 50
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch":
|
| 57 |
-
"grad_norm":
|
| 58 |
-
"learning_rate":
|
| 59 |
-
"loss": 0.
|
| 60 |
"step": 60
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch":
|
| 64 |
-
"grad_norm":
|
| 65 |
-
"learning_rate":
|
| 66 |
-
"loss": 0.
|
| 67 |
"step": 70
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"
|
| 72 |
-
"
|
| 73 |
-
"
|
| 74 |
-
"eval_samples_per_second": 151.489,
|
| 75 |
-
"eval_steps_per_second": 9.588,
|
| 76 |
-
"step": 79
|
| 77 |
-
},
|
| 78 |
-
{
|
| 79 |
-
"epoch": 2.0253164556962027,
|
| 80 |
-
"grad_norm": 1.6714565753936768,
|
| 81 |
-
"learning_rate": 4.415954415954416e-05,
|
| 82 |
-
"loss": 0.4969,
|
| 83 |
"step": 80
|
| 84 |
},
|
| 85 |
{
|
| 86 |
-
"epoch":
|
| 87 |
-
"grad_norm": 1.
|
| 88 |
-
"learning_rate":
|
| 89 |
-
"loss": 0.
|
| 90 |
"step": 90
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch":
|
| 94 |
-
"grad_norm":
|
| 95 |
-
"learning_rate":
|
| 96 |
-
"loss": 0.
|
| 97 |
"step": 100
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch":
|
| 101 |
-
"grad_norm":
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 0.
|
| 104 |
"step": 110
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"
|
| 109 |
-
"
|
| 110 |
-
"
|
| 111 |
-
"eval_samples_per_second": 146.201,
|
| 112 |
-
"eval_steps_per_second": 9.253,
|
| 113 |
-
"step": 118
|
| 114 |
-
},
|
| 115 |
-
{
|
| 116 |
-
"epoch": 3.037974683544304,
|
| 117 |
-
"grad_norm": 1.0525858402252197,
|
| 118 |
-
"learning_rate": 3.846153846153846e-05,
|
| 119 |
-
"loss": 0.4946,
|
| 120 |
"step": 120
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"epoch":
|
| 124 |
-
"grad_norm": 2.
|
| 125 |
-
"learning_rate":
|
| 126 |
-
"loss": 0.
|
| 127 |
"step": 130
|
| 128 |
},
|
| 129 |
{
|
| 130 |
-
"epoch":
|
| 131 |
-
"grad_norm": 1.
|
| 132 |
-
"learning_rate":
|
| 133 |
-
"loss": 0.
|
| 134 |
"step": 140
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch":
|
| 138 |
-
"grad_norm": 1.
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 150
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"
|
| 146 |
-
"
|
| 147 |
-
"
|
| 148 |
-
"eval_samples_per_second": 144.335,
|
| 149 |
-
"eval_steps_per_second": 9.135,
|
| 150 |
-
"step": 158
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 4.050632911392405,
|
| 154 |
-
"grad_norm": 2.1460752487182617,
|
| 155 |
-
"learning_rate": 3.2763532763532764e-05,
|
| 156 |
-
"loss": 0.4983,
|
| 157 |
"step": 160
|
| 158 |
},
|
| 159 |
{
|
| 160 |
-
"epoch":
|
| 161 |
-
"grad_norm": 1.
|
| 162 |
-
"learning_rate": 3.
|
| 163 |
-
"loss": 0.
|
| 164 |
"step": 170
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"grad_norm":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 180
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"grad_norm":
|
| 176 |
-
"learning_rate":
|
| 177 |
-
"loss": 0.
|
| 178 |
"step": 190
|
| 179 |
},
|
| 180 |
{
|
| 181 |
-
"epoch":
|
| 182 |
-
"
|
| 183 |
-
"
|
| 184 |
-
"
|
| 185 |
-
"eval_samples_per_second": 144.763,
|
| 186 |
-
"eval_steps_per_second": 9.162,
|
| 187 |
-
"step": 197
|
| 188 |
-
},
|
| 189 |
-
{
|
| 190 |
-
"epoch": 5.063291139240507,
|
| 191 |
-
"grad_norm": 1.212802529335022,
|
| 192 |
-
"learning_rate": 2.706552706552707e-05,
|
| 193 |
-
"loss": 0.4828,
|
| 194 |
"step": 200
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch":
|
| 198 |
-
"grad_norm": 1.
|
| 199 |
-
"learning_rate":
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 210
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"grad_norm": 1.
|
| 206 |
-
"learning_rate":
|
| 207 |
-
"loss": 0.
|
| 208 |
"step": 220
|
| 209 |
},
|
| 210 |
{
|
| 211 |
-
"epoch":
|
| 212 |
-
"grad_norm":
|
| 213 |
-
"learning_rate":
|
| 214 |
-
"loss": 0.
|
| 215 |
"step": 230
|
| 216 |
},
|
| 217 |
{
|
| 218 |
-
"epoch":
|
| 219 |
-
"
|
| 220 |
-
"
|
| 221 |
-
"
|
| 222 |
-
"eval_samples_per_second": 148.382,
|
| 223 |
-
"eval_steps_per_second": 9.391,
|
| 224 |
-
"step": 237
|
| 225 |
-
},
|
| 226 |
-
{
|
| 227 |
-
"epoch": 6.075949367088608,
|
| 228 |
-
"grad_norm": 1.6727008819580078,
|
| 229 |
-
"learning_rate": 2.1367521367521368e-05,
|
| 230 |
-
"loss": 0.4878,
|
| 231 |
"step": 240
|
| 232 |
},
|
| 233 |
{
|
| 234 |
-
"epoch":
|
| 235 |
-
"grad_norm":
|
| 236 |
-
"learning_rate":
|
| 237 |
-
"loss": 0.
|
| 238 |
"step": 250
|
| 239 |
},
|
| 240 |
{
|
| 241 |
-
"epoch":
|
| 242 |
-
"grad_norm":
|
| 243 |
-
"learning_rate":
|
| 244 |
-
"loss": 0.
|
| 245 |
"step": 260
|
| 246 |
},
|
| 247 |
{
|
| 248 |
-
"epoch":
|
| 249 |
-
"grad_norm":
|
| 250 |
-
"learning_rate":
|
| 251 |
-
"loss": 0.
|
| 252 |
"step": 270
|
| 253 |
},
|
| 254 |
{
|
| 255 |
-
"epoch":
|
| 256 |
-
"
|
| 257 |
-
"
|
| 258 |
-
"
|
| 259 |
-
"
|
| 260 |
-
"eval_steps_per_second": 9.551,
|
| 261 |
-
"step": 276
|
| 262 |
},
|
| 263 |
{
|
| 264 |
-
"epoch":
|
| 265 |
-
"
|
| 266 |
-
"
|
| 267 |
-
"
|
|
|
|
|
|
|
| 268 |
"step": 280
|
| 269 |
},
|
| 270 |
{
|
| 271 |
-
"epoch":
|
| 272 |
-
"grad_norm":
|
| 273 |
-
"learning_rate":
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 290
|
| 276 |
},
|
| 277 |
{
|
| 278 |
-
"epoch":
|
| 279 |
-
"grad_norm":
|
| 280 |
-
"learning_rate":
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 300
|
| 283 |
},
|
| 284 |
{
|
| 285 |
-
"epoch":
|
| 286 |
-
"grad_norm": 1.
|
| 287 |
-
"learning_rate":
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 310
|
| 290 |
},
|
| 291 |
{
|
| 292 |
-
"epoch":
|
| 293 |
-
"
|
| 294 |
-
"
|
| 295 |
-
"
|
| 296 |
-
"eval_samples_per_second": 145.004,
|
| 297 |
-
"eval_steps_per_second": 9.177,
|
| 298 |
-
"step": 316
|
| 299 |
-
},
|
| 300 |
-
{
|
| 301 |
-
"epoch": 8.10126582278481,
|
| 302 |
-
"grad_norm": 1.7846542596817017,
|
| 303 |
-
"learning_rate": 9.971509971509972e-06,
|
| 304 |
-
"loss": 0.4552,
|
| 305 |
"step": 320
|
| 306 |
},
|
| 307 |
{
|
| 308 |
-
"epoch":
|
| 309 |
-
"grad_norm": 1.
|
| 310 |
-
"learning_rate":
|
| 311 |
-
"loss": 0.
|
| 312 |
"step": 330
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"epoch":
|
| 316 |
-
"grad_norm": 1.
|
| 317 |
-
"learning_rate":
|
| 318 |
-
"loss": 0.
|
| 319 |
"step": 340
|
| 320 |
},
|
| 321 |
{
|
| 322 |
-
"epoch":
|
| 323 |
-
"grad_norm": 1.
|
| 324 |
-
"learning_rate":
|
| 325 |
-
"loss": 0.
|
| 326 |
"step": 350
|
| 327 |
},
|
| 328 |
{
|
| 329 |
-
"epoch":
|
| 330 |
-
"
|
| 331 |
-
"
|
| 332 |
-
"
|
| 333 |
-
"eval_samples_per_second": 148.154,
|
| 334 |
-
"eval_steps_per_second": 9.377,
|
| 335 |
-
"step": 355
|
| 336 |
-
},
|
| 337 |
-
{
|
| 338 |
-
"epoch": 9.113924050632912,
|
| 339 |
-
"grad_norm": 1.3692480325698853,
|
| 340 |
-
"learning_rate": 4.273504273504274e-06,
|
| 341 |
-
"loss": 0.437,
|
| 342 |
"step": 360
|
| 343 |
},
|
| 344 |
{
|
| 345 |
-
"epoch":
|
| 346 |
-
"grad_norm":
|
| 347 |
-
"learning_rate":
|
| 348 |
-
"loss": 0.
|
| 349 |
"step": 370
|
| 350 |
},
|
| 351 |
{
|
| 352 |
-
"epoch":
|
| 353 |
-
"grad_norm": 1.
|
| 354 |
-
"learning_rate":
|
| 355 |
-
"loss": 0.
|
| 356 |
"step": 380
|
| 357 |
},
|
| 358 |
{
|
| 359 |
-
"epoch":
|
| 360 |
-
"grad_norm": 1.
|
| 361 |
-
"learning_rate":
|
| 362 |
-
"loss": 0.
|
| 363 |
"step": 390
|
| 364 |
},
|
| 365 |
{
|
| 366 |
-
"epoch":
|
| 367 |
-
"
|
| 368 |
-
"
|
| 369 |
-
"
|
| 370 |
-
"
|
| 371 |
-
|
| 372 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"epoch":
|
| 376 |
-
"step":
|
| 377 |
-
"total_flos": 1.
|
| 378 |
-
"train_loss": 0.
|
| 379 |
-
"train_runtime":
|
| 380 |
-
"train_samples_per_second":
|
| 381 |
-
"train_steps_per_second":
|
| 382 |
}
|
| 383 |
],
|
| 384 |
"logging_steps": 10,
|
| 385 |
-
"max_steps":
|
| 386 |
"num_input_tokens_seen": 0,
|
| 387 |
"num_train_epochs": 10,
|
| 388 |
"save_steps": 500,
|
|
@@ -398,7 +2085,7 @@
|
|
| 398 |
"attributes": {}
|
| 399 |
}
|
| 400 |
},
|
| 401 |
-
"total_flos": 1.
|
| 402 |
"train_batch_size": 16,
|
| 403 |
"trial_name": null,
|
| 404 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.8894595801697186,
|
| 3 |
+
"best_model_checkpoint": "tsec_vit_model/checkpoint-2520",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 2800,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.03571428571428571,
|
| 13 |
+
"grad_norm": 0.632759690284729,
|
| 14 |
+
"learning_rate": 1.7857142857142857e-06,
|
| 15 |
+
"loss": 0.7005,
|
| 16 |
"step": 10
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.07142857142857142,
|
| 20 |
+
"grad_norm": 0.6090130805969238,
|
| 21 |
+
"learning_rate": 3.5714285714285714e-06,
|
| 22 |
+
"loss": 0.6888,
|
| 23 |
"step": 20
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.10714285714285714,
|
| 27 |
+
"grad_norm": 0.5027034878730774,
|
| 28 |
+
"learning_rate": 5.357142857142857e-06,
|
| 29 |
+
"loss": 0.6841,
|
| 30 |
"step": 30
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.14285714285714285,
|
| 34 |
+
"grad_norm": 0.7158761620521545,
|
| 35 |
+
"learning_rate": 7.142857142857143e-06,
|
| 36 |
+
"loss": 0.6711,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
"step": 40
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.17857142857142858,
|
| 41 |
+
"grad_norm": 1.0677605867385864,
|
| 42 |
+
"learning_rate": 8.92857142857143e-06,
|
| 43 |
+
"loss": 0.6439,
|
| 44 |
"step": 50
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.21428571428571427,
|
| 48 |
+
"grad_norm": 0.716444194316864,
|
| 49 |
+
"learning_rate": 1.0714285714285714e-05,
|
| 50 |
+
"loss": 0.6388,
|
| 51 |
"step": 60
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.25,
|
| 55 |
+
"grad_norm": 0.9646293520927429,
|
| 56 |
+
"learning_rate": 1.25e-05,
|
| 57 |
+
"loss": 0.5942,
|
| 58 |
"step": 70
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.2857142857142857,
|
| 62 |
+
"grad_norm": 1.5171552896499634,
|
| 63 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 64 |
+
"loss": 0.587,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
"step": 80
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.32142857142857145,
|
| 69 |
+
"grad_norm": 1.2230862379074097,
|
| 70 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 71 |
+
"loss": 0.5882,
|
| 72 |
"step": 90
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"epoch": 0.35714285714285715,
|
| 76 |
+
"grad_norm": 1.0726110935211182,
|
| 77 |
+
"learning_rate": 1.785714285714286e-05,
|
| 78 |
+
"loss": 0.5506,
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
+
"epoch": 0.39285714285714285,
|
| 83 |
+
"grad_norm": 2.488288640975952,
|
| 84 |
+
"learning_rate": 1.9642857142857145e-05,
|
| 85 |
+
"loss": 0.5487,
|
| 86 |
"step": 110
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.42857142857142855,
|
| 90 |
+
"grad_norm": 2.186500310897827,
|
| 91 |
+
"learning_rate": 2.1428571428571428e-05,
|
| 92 |
+
"loss": 0.5138,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 93 |
"step": 120
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.4642857142857143,
|
| 97 |
+
"grad_norm": 2.18605637550354,
|
| 98 |
+
"learning_rate": 2.3214285714285715e-05,
|
| 99 |
+
"loss": 0.5203,
|
| 100 |
"step": 130
|
| 101 |
},
|
| 102 |
{
|
| 103 |
+
"epoch": 0.5,
|
| 104 |
+
"grad_norm": 1.5495938062667847,
|
| 105 |
+
"learning_rate": 2.5e-05,
|
| 106 |
+
"loss": 0.5006,
|
| 107 |
"step": 140
|
| 108 |
},
|
| 109 |
{
|
| 110 |
+
"epoch": 0.5357142857142857,
|
| 111 |
+
"grad_norm": 1.3939014673233032,
|
| 112 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 113 |
+
"loss": 0.5171,
|
| 114 |
"step": 150
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 0.5714285714285714,
|
| 118 |
+
"grad_norm": 1.4516574144363403,
|
| 119 |
+
"learning_rate": 2.857142857142857e-05,
|
| 120 |
+
"loss": 0.4882,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 121 |
"step": 160
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 0.6071428571428571,
|
| 125 |
+
"grad_norm": 1.5352245569229126,
|
| 126 |
+
"learning_rate": 3.0357142857142857e-05,
|
| 127 |
+
"loss": 0.4809,
|
| 128 |
"step": 170
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.6428571428571429,
|
| 132 |
+
"grad_norm": 2.6514480113983154,
|
| 133 |
+
"learning_rate": 3.2142857142857144e-05,
|
| 134 |
+
"loss": 0.458,
|
| 135 |
"step": 180
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.6785714285714286,
|
| 139 |
+
"grad_norm": 2.1452901363372803,
|
| 140 |
+
"learning_rate": 3.392857142857143e-05,
|
| 141 |
+
"loss": 0.478,
|
| 142 |
"step": 190
|
| 143 |
},
|
| 144 |
{
|
| 145 |
+
"epoch": 0.7142857142857143,
|
| 146 |
+
"grad_norm": 1.5536608695983887,
|
| 147 |
+
"learning_rate": 3.571428571428572e-05,
|
| 148 |
+
"loss": 0.4851,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 149 |
"step": 200
|
| 150 |
},
|
| 151 |
{
|
| 152 |
+
"epoch": 0.75,
|
| 153 |
+
"grad_norm": 1.7489731311798096,
|
| 154 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 155 |
+
"loss": 0.4608,
|
| 156 |
"step": 210
|
| 157 |
},
|
| 158 |
{
|
| 159 |
+
"epoch": 0.7857142857142857,
|
| 160 |
+
"grad_norm": 1.6109980344772339,
|
| 161 |
+
"learning_rate": 3.928571428571429e-05,
|
| 162 |
+
"loss": 0.4611,
|
| 163 |
"step": 220
|
| 164 |
},
|
| 165 |
{
|
| 166 |
+
"epoch": 0.8214285714285714,
|
| 167 |
+
"grad_norm": 2.391719102859497,
|
| 168 |
+
"learning_rate": 4.107142857142857e-05,
|
| 169 |
+
"loss": 0.3998,
|
| 170 |
"step": 230
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.8571428571428571,
|
| 174 |
+
"grad_norm": 3.6933391094207764,
|
| 175 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 176 |
+
"loss": 0.4624,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 177 |
"step": 240
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 0.8928571428571429,
|
| 181 |
+
"grad_norm": 2.0588369369506836,
|
| 182 |
+
"learning_rate": 4.464285714285715e-05,
|
| 183 |
+
"loss": 0.4437,
|
| 184 |
"step": 250
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"epoch": 0.9285714285714286,
|
| 188 |
+
"grad_norm": 2.831156015396118,
|
| 189 |
+
"learning_rate": 4.642857142857143e-05,
|
| 190 |
+
"loss": 0.4652,
|
| 191 |
"step": 260
|
| 192 |
},
|
| 193 |
{
|
| 194 |
+
"epoch": 0.9642857142857143,
|
| 195 |
+
"grad_norm": 2.0229790210723877,
|
| 196 |
+
"learning_rate": 4.8214285714285716e-05,
|
| 197 |
+
"loss": 0.4037,
|
| 198 |
"step": 270
|
| 199 |
},
|
| 200 |
{
|
| 201 |
+
"epoch": 1.0,
|
| 202 |
+
"grad_norm": 2.1834700107574463,
|
| 203 |
+
"learning_rate": 5e-05,
|
| 204 |
+
"loss": 0.4387,
|
| 205 |
+
"step": 280
|
|
|
|
|
|
|
| 206 |
},
|
| 207 |
{
|
| 208 |
+
"epoch": 1.0,
|
| 209 |
+
"eval_accuracy": 0.8150960250111657,
|
| 210 |
+
"eval_loss": 0.4179099500179291,
|
| 211 |
+
"eval_runtime": 110.4006,
|
| 212 |
+
"eval_samples_per_second": 40.561,
|
| 213 |
+
"eval_steps_per_second": 2.536,
|
| 214 |
"step": 280
|
| 215 |
},
|
| 216 |
{
|
| 217 |
+
"epoch": 1.0357142857142858,
|
| 218 |
+
"grad_norm": 1.7746831178665161,
|
| 219 |
+
"learning_rate": 4.9801587301587306e-05,
|
| 220 |
+
"loss": 0.4069,
|
| 221 |
"step": 290
|
| 222 |
},
|
| 223 |
{
|
| 224 |
+
"epoch": 1.0714285714285714,
|
| 225 |
+
"grad_norm": 2.6947922706604004,
|
| 226 |
+
"learning_rate": 4.960317460317461e-05,
|
| 227 |
+
"loss": 0.4334,
|
| 228 |
"step": 300
|
| 229 |
},
|
| 230 |
{
|
| 231 |
+
"epoch": 1.1071428571428572,
|
| 232 |
+
"grad_norm": 1.4933383464813232,
|
| 233 |
+
"learning_rate": 4.940476190476191e-05,
|
| 234 |
+
"loss": 0.4409,
|
| 235 |
"step": 310
|
| 236 |
},
|
| 237 |
{
|
| 238 |
+
"epoch": 1.1428571428571428,
|
| 239 |
+
"grad_norm": 1.8830089569091797,
|
| 240 |
+
"learning_rate": 4.9206349206349204e-05,
|
| 241 |
+
"loss": 0.4065,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 242 |
"step": 320
|
| 243 |
},
|
| 244 |
{
|
| 245 |
+
"epoch": 1.1785714285714286,
|
| 246 |
+
"grad_norm": 1.978843092918396,
|
| 247 |
+
"learning_rate": 4.900793650793651e-05,
|
| 248 |
+
"loss": 0.4273,
|
| 249 |
"step": 330
|
| 250 |
},
|
| 251 |
{
|
| 252 |
+
"epoch": 1.2142857142857142,
|
| 253 |
+
"grad_norm": 1.2928149700164795,
|
| 254 |
+
"learning_rate": 4.880952380952381e-05,
|
| 255 |
+
"loss": 0.4376,
|
| 256 |
"step": 340
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"epoch": 1.25,
|
| 260 |
+
"grad_norm": 1.5401108264923096,
|
| 261 |
+
"learning_rate": 4.8611111111111115e-05,
|
| 262 |
+
"loss": 0.4061,
|
| 263 |
"step": 350
|
| 264 |
},
|
| 265 |
{
|
| 266 |
+
"epoch": 1.2857142857142856,
|
| 267 |
+
"grad_norm": 2.296520709991455,
|
| 268 |
+
"learning_rate": 4.841269841269841e-05,
|
| 269 |
+
"loss": 0.433,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 270 |
"step": 360
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 1.3214285714285714,
|
| 274 |
+
"grad_norm": 1.2640879154205322,
|
| 275 |
+
"learning_rate": 4.8214285714285716e-05,
|
| 276 |
+
"loss": 0.4142,
|
| 277 |
"step": 370
|
| 278 |
},
|
| 279 |
{
|
| 280 |
+
"epoch": 1.3571428571428572,
|
| 281 |
+
"grad_norm": 1.3309462070465088,
|
| 282 |
+
"learning_rate": 4.801587301587302e-05,
|
| 283 |
+
"loss": 0.4222,
|
| 284 |
"step": 380
|
| 285 |
},
|
| 286 |
{
|
| 287 |
+
"epoch": 1.3928571428571428,
|
| 288 |
+
"grad_norm": 1.8462568521499634,
|
| 289 |
+
"learning_rate": 4.781746031746032e-05,
|
| 290 |
+
"loss": 0.4483,
|
| 291 |
"step": 390
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 1.4285714285714286,
|
| 295 |
+
"grad_norm": 1.074401617050171,
|
| 296 |
+
"learning_rate": 4.761904761904762e-05,
|
| 297 |
+
"loss": 0.3762,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.4642857142857144,
|
| 302 |
+
"grad_norm": 1.5691473484039307,
|
| 303 |
+
"learning_rate": 4.7420634920634924e-05,
|
| 304 |
+
"loss": 0.415,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.5,
|
| 309 |
+
"grad_norm": 1.2222189903259277,
|
| 310 |
+
"learning_rate": 4.722222222222222e-05,
|
| 311 |
+
"loss": 0.3862,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.5357142857142856,
|
| 316 |
+
"grad_norm": 1.4210410118103027,
|
| 317 |
+
"learning_rate": 4.7023809523809525e-05,
|
| 318 |
+
"loss": 0.352,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.5714285714285714,
|
| 323 |
+
"grad_norm": 1.632934331893921,
|
| 324 |
+
"learning_rate": 4.682539682539683e-05,
|
| 325 |
+
"loss": 0.3873,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.6071428571428572,
|
| 330 |
+
"grad_norm": 1.3977652788162231,
|
| 331 |
+
"learning_rate": 4.662698412698413e-05,
|
| 332 |
+
"loss": 0.3763,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.6428571428571428,
|
| 337 |
+
"grad_norm": 2.74967622756958,
|
| 338 |
+
"learning_rate": 4.642857142857143e-05,
|
| 339 |
+
"loss": 0.3823,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.6785714285714286,
|
| 344 |
+
"grad_norm": 1.1579408645629883,
|
| 345 |
+
"learning_rate": 4.623015873015873e-05,
|
| 346 |
+
"loss": 0.3967,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.7142857142857144,
|
| 351 |
+
"grad_norm": 2.7291440963745117,
|
| 352 |
+
"learning_rate": 4.603174603174603e-05,
|
| 353 |
+
"loss": 0.4201,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.75,
|
| 358 |
+
"grad_norm": 1.0126214027404785,
|
| 359 |
+
"learning_rate": 4.5833333333333334e-05,
|
| 360 |
+
"loss": 0.3598,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.7857142857142856,
|
| 365 |
+
"grad_norm": 1.6705009937286377,
|
| 366 |
+
"learning_rate": 4.563492063492064e-05,
|
| 367 |
+
"loss": 0.366,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.8214285714285714,
|
| 372 |
+
"grad_norm": 1.6749546527862549,
|
| 373 |
+
"learning_rate": 4.543650793650794e-05,
|
| 374 |
+
"loss": 0.3854,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.8571428571428572,
|
| 379 |
+
"grad_norm": 1.0892480611801147,
|
| 380 |
+
"learning_rate": 4.523809523809524e-05,
|
| 381 |
+
"loss": 0.3563,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.8928571428571428,
|
| 386 |
+
"grad_norm": 1.4169703722000122,
|
| 387 |
+
"learning_rate": 4.503968253968254e-05,
|
| 388 |
+
"loss": 0.3903,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.9285714285714286,
|
| 393 |
+
"grad_norm": 1.393864631652832,
|
| 394 |
+
"learning_rate": 4.4841269841269846e-05,
|
| 395 |
+
"loss": 0.3915,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.9642857142857144,
|
| 400 |
+
"grad_norm": 1.6008644104003906,
|
| 401 |
+
"learning_rate": 4.464285714285715e-05,
|
| 402 |
+
"loss": 0.3665,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.0,
|
| 407 |
+
"grad_norm": 1.0977710485458374,
|
| 408 |
+
"learning_rate": 4.4444444444444447e-05,
|
| 409 |
+
"loss": 0.4239,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.0,
|
| 414 |
+
"eval_accuracy": 0.8398838767306833,
|
| 415 |
+
"eval_loss": 0.3610800504684448,
|
| 416 |
+
"eval_runtime": 114.7581,
|
| 417 |
+
"eval_samples_per_second": 39.021,
|
| 418 |
+
"eval_steps_per_second": 2.44,
|
| 419 |
+
"step": 560
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 2.0357142857142856,
|
| 423 |
+
"grad_norm": 1.4717609882354736,
|
| 424 |
+
"learning_rate": 4.4246031746031744e-05,
|
| 425 |
+
"loss": 0.3609,
|
| 426 |
+
"step": 570
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 2.0714285714285716,
|
| 430 |
+
"grad_norm": 1.0943901538848877,
|
| 431 |
+
"learning_rate": 4.404761904761905e-05,
|
| 432 |
+
"loss": 0.3379,
|
| 433 |
+
"step": 580
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 2.107142857142857,
|
| 437 |
+
"grad_norm": 1.8704146146774292,
|
| 438 |
+
"learning_rate": 4.384920634920635e-05,
|
| 439 |
+
"loss": 0.364,
|
| 440 |
+
"step": 590
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.142857142857143,
|
| 444 |
+
"grad_norm": 1.2977815866470337,
|
| 445 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 446 |
+
"loss": 0.3834,
|
| 447 |
+
"step": 600
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.1785714285714284,
|
| 451 |
+
"grad_norm": 1.6529649496078491,
|
| 452 |
+
"learning_rate": 4.345238095238096e-05,
|
| 453 |
+
"loss": 0.3688,
|
| 454 |
+
"step": 610
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 2.2142857142857144,
|
| 458 |
+
"grad_norm": 1.1914763450622559,
|
| 459 |
+
"learning_rate": 4.3253968253968256e-05,
|
| 460 |
+
"loss": 0.3548,
|
| 461 |
+
"step": 620
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.25,
|
| 465 |
+
"grad_norm": 1.70374596118927,
|
| 466 |
+
"learning_rate": 4.305555555555556e-05,
|
| 467 |
+
"loss": 0.3323,
|
| 468 |
+
"step": 630
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.2857142857142856,
|
| 472 |
+
"grad_norm": 1.1331793069839478,
|
| 473 |
+
"learning_rate": 4.2857142857142856e-05,
|
| 474 |
+
"loss": 0.3494,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.3214285714285716,
|
| 479 |
+
"grad_norm": 1.3151099681854248,
|
| 480 |
+
"learning_rate": 4.265873015873016e-05,
|
| 481 |
+
"loss": 0.3177,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.357142857142857,
|
| 486 |
+
"grad_norm": 1.7670910358428955,
|
| 487 |
+
"learning_rate": 4.2460317460317464e-05,
|
| 488 |
+
"loss": 0.3145,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.392857142857143,
|
| 493 |
+
"grad_norm": 1.441829800605774,
|
| 494 |
+
"learning_rate": 4.226190476190476e-05,
|
| 495 |
+
"loss": 0.2968,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.4285714285714284,
|
| 500 |
+
"grad_norm": 1.5992521047592163,
|
| 501 |
+
"learning_rate": 4.2063492063492065e-05,
|
| 502 |
+
"loss": 0.353,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.4642857142857144,
|
| 507 |
+
"grad_norm": 1.076711654663086,
|
| 508 |
+
"learning_rate": 4.186507936507937e-05,
|
| 509 |
+
"loss": 0.3461,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.5,
|
| 514 |
+
"grad_norm": 1.0764169692993164,
|
| 515 |
+
"learning_rate": 4.166666666666667e-05,
|
| 516 |
+
"loss": 0.3345,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.5357142857142856,
|
| 521 |
+
"grad_norm": 1.3735431432724,
|
| 522 |
+
"learning_rate": 4.1468253968253976e-05,
|
| 523 |
+
"loss": 0.3612,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.571428571428571,
|
| 528 |
+
"grad_norm": 1.4703452587127686,
|
| 529 |
+
"learning_rate": 4.126984126984127e-05,
|
| 530 |
+
"loss": 0.3444,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.607142857142857,
|
| 535 |
+
"grad_norm": 1.451257348060608,
|
| 536 |
+
"learning_rate": 4.107142857142857e-05,
|
| 537 |
+
"loss": 0.3287,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.642857142857143,
|
| 542 |
+
"grad_norm": 1.8615336418151855,
|
| 543 |
+
"learning_rate": 4.0873015873015874e-05,
|
| 544 |
+
"loss": 0.3419,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.678571428571429,
|
| 549 |
+
"grad_norm": 2.925323963165283,
|
| 550 |
+
"learning_rate": 4.067460317460318e-05,
|
| 551 |
+
"loss": 0.3217,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.7142857142857144,
|
| 556 |
+
"grad_norm": 1.7260726690292358,
|
| 557 |
+
"learning_rate": 4.047619047619048e-05,
|
| 558 |
+
"loss": 0.3584,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.75,
|
| 563 |
+
"grad_norm": 1.4659258127212524,
|
| 564 |
+
"learning_rate": 4.027777777777778e-05,
|
| 565 |
+
"loss": 0.3821,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.7857142857142856,
|
| 570 |
+
"grad_norm": 1.7859466075897217,
|
| 571 |
+
"learning_rate": 4.007936507936508e-05,
|
| 572 |
+
"loss": 0.3595,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.821428571428571,
|
| 577 |
+
"grad_norm": 1.2320784330368042,
|
| 578 |
+
"learning_rate": 3.9880952380952386e-05,
|
| 579 |
+
"loss": 0.3764,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.857142857142857,
|
| 584 |
+
"grad_norm": 1.4097161293029785,
|
| 585 |
+
"learning_rate": 3.968253968253968e-05,
|
| 586 |
+
"loss": 0.3282,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.892857142857143,
|
| 591 |
+
"grad_norm": 1.185285210609436,
|
| 592 |
+
"learning_rate": 3.9484126984126986e-05,
|
| 593 |
+
"loss": 0.3214,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.928571428571429,
|
| 598 |
+
"grad_norm": 1.3706032037734985,
|
| 599 |
+
"learning_rate": 3.928571428571429e-05,
|
| 600 |
+
"loss": 0.3092,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.9642857142857144,
|
| 605 |
+
"grad_norm": 2.1342084407806396,
|
| 606 |
+
"learning_rate": 3.908730158730159e-05,
|
| 607 |
+
"loss": 0.3274,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.0,
|
| 612 |
+
"grad_norm": 1.438025712966919,
|
| 613 |
+
"learning_rate": 3.888888888888889e-05,
|
| 614 |
+
"loss": 0.3148,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 3.0,
|
| 619 |
+
"eval_accuracy": 0.8599821348816435,
|
| 620 |
+
"eval_loss": 0.31560027599334717,
|
| 621 |
+
"eval_runtime": 116.0276,
|
| 622 |
+
"eval_samples_per_second": 38.594,
|
| 623 |
+
"eval_steps_per_second": 2.413,
|
| 624 |
+
"step": 840
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 3.0357142857142856,
|
| 628 |
+
"grad_norm": 1.6471174955368042,
|
| 629 |
+
"learning_rate": 3.8690476190476195e-05,
|
| 630 |
+
"loss": 0.3198,
|
| 631 |
+
"step": 850
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 3.0714285714285716,
|
| 635 |
+
"grad_norm": 1.5575629472732544,
|
| 636 |
+
"learning_rate": 3.84920634920635e-05,
|
| 637 |
+
"loss": 0.3127,
|
| 638 |
+
"step": 860
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 3.107142857142857,
|
| 642 |
+
"grad_norm": 1.1578933000564575,
|
| 643 |
+
"learning_rate": 3.8293650793650795e-05,
|
| 644 |
+
"loss": 0.3416,
|
| 645 |
+
"step": 870
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 3.142857142857143,
|
| 649 |
+
"grad_norm": 1.9167553186416626,
|
| 650 |
+
"learning_rate": 3.809523809523809e-05,
|
| 651 |
+
"loss": 0.3626,
|
| 652 |
+
"step": 880
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"epoch": 3.1785714285714284,
|
| 656 |
+
"grad_norm": 1.038451910018921,
|
| 657 |
+
"learning_rate": 3.7896825396825396e-05,
|
| 658 |
+
"loss": 0.336,
|
| 659 |
+
"step": 890
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.2142857142857144,
|
| 663 |
+
"grad_norm": 1.2739081382751465,
|
| 664 |
+
"learning_rate": 3.76984126984127e-05,
|
| 665 |
+
"loss": 0.2838,
|
| 666 |
+
"step": 900
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 3.25,
|
| 670 |
+
"grad_norm": 1.2061238288879395,
|
| 671 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 672 |
+
"loss": 0.3099,
|
| 673 |
+
"step": 910
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 3.2857142857142856,
|
| 677 |
+
"grad_norm": 1.5365681648254395,
|
| 678 |
+
"learning_rate": 3.730158730158731e-05,
|
| 679 |
+
"loss": 0.3163,
|
| 680 |
+
"step": 920
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 3.3214285714285716,
|
| 684 |
+
"grad_norm": 2.33838152885437,
|
| 685 |
+
"learning_rate": 3.7103174603174604e-05,
|
| 686 |
+
"loss": 0.3109,
|
| 687 |
+
"step": 930
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 3.357142857142857,
|
| 691 |
+
"grad_norm": 1.2280935049057007,
|
| 692 |
+
"learning_rate": 3.690476190476191e-05,
|
| 693 |
+
"loss": 0.3104,
|
| 694 |
+
"step": 940
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.392857142857143,
|
| 698 |
+
"grad_norm": 1.2506486177444458,
|
| 699 |
+
"learning_rate": 3.6706349206349205e-05,
|
| 700 |
+
"loss": 0.339,
|
| 701 |
+
"step": 950
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.4285714285714284,
|
| 705 |
+
"grad_norm": 1.0430885553359985,
|
| 706 |
+
"learning_rate": 3.650793650793651e-05,
|
| 707 |
+
"loss": 0.3015,
|
| 708 |
+
"step": 960
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.4642857142857144,
|
| 712 |
+
"grad_norm": 1.5323641300201416,
|
| 713 |
+
"learning_rate": 3.630952380952381e-05,
|
| 714 |
+
"loss": 0.3408,
|
| 715 |
+
"step": 970
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.5,
|
| 719 |
+
"grad_norm": 1.0826184749603271,
|
| 720 |
+
"learning_rate": 3.611111111111111e-05,
|
| 721 |
+
"loss": 0.3061,
|
| 722 |
+
"step": 980
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.5357142857142856,
|
| 726 |
+
"grad_norm": 2.019437313079834,
|
| 727 |
+
"learning_rate": 3.591269841269841e-05,
|
| 728 |
+
"loss": 0.3465,
|
| 729 |
+
"step": 990
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.571428571428571,
|
| 733 |
+
"grad_norm": 1.5877951383590698,
|
| 734 |
+
"learning_rate": 3.571428571428572e-05,
|
| 735 |
+
"loss": 0.3023,
|
| 736 |
+
"step": 1000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.607142857142857,
|
| 740 |
+
"grad_norm": 1.5832830667495728,
|
| 741 |
+
"learning_rate": 3.551587301587302e-05,
|
| 742 |
+
"loss": 0.262,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.642857142857143,
|
| 747 |
+
"grad_norm": 1.5758150815963745,
|
| 748 |
+
"learning_rate": 3.5317460317460324e-05,
|
| 749 |
+
"loss": 0.3337,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.678571428571429,
|
| 754 |
+
"grad_norm": 1.255638837814331,
|
| 755 |
+
"learning_rate": 3.511904761904762e-05,
|
| 756 |
+
"loss": 0.2857,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.7142857142857144,
|
| 761 |
+
"grad_norm": 1.5360593795776367,
|
| 762 |
+
"learning_rate": 3.492063492063492e-05,
|
| 763 |
+
"loss": 0.2964,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.75,
|
| 768 |
+
"grad_norm": 1.0826270580291748,
|
| 769 |
+
"learning_rate": 3.472222222222222e-05,
|
| 770 |
+
"loss": 0.2872,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.7857142857142856,
|
| 775 |
+
"grad_norm": 1.6883878707885742,
|
| 776 |
+
"learning_rate": 3.4523809523809526e-05,
|
| 777 |
+
"loss": 0.3016,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.821428571428571,
|
| 782 |
+
"grad_norm": 1.8018616437911987,
|
| 783 |
+
"learning_rate": 3.432539682539683e-05,
|
| 784 |
+
"loss": 0.3115,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.857142857142857,
|
| 789 |
+
"grad_norm": 1.9364327192306519,
|
| 790 |
+
"learning_rate": 3.412698412698413e-05,
|
| 791 |
+
"loss": 0.2986,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.892857142857143,
|
| 796 |
+
"grad_norm": 1.6606508493423462,
|
| 797 |
+
"learning_rate": 3.392857142857143e-05,
|
| 798 |
+
"loss": 0.3017,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.928571428571429,
|
| 803 |
+
"grad_norm": 2.1849498748779297,
|
| 804 |
+
"learning_rate": 3.3730158730158734e-05,
|
| 805 |
+
"loss": 0.2779,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.9642857142857144,
|
| 810 |
+
"grad_norm": 1.5579859018325806,
|
| 811 |
+
"learning_rate": 3.353174603174603e-05,
|
| 812 |
+
"loss": 0.3199,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 4.0,
|
| 817 |
+
"grad_norm": 2.6287128925323486,
|
| 818 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 819 |
+
"loss": 0.2988,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 4.0,
|
| 824 |
+
"eval_accuracy": 0.8729343456900402,
|
| 825 |
+
"eval_loss": 0.30018478631973267,
|
| 826 |
+
"eval_runtime": 115.4434,
|
| 827 |
+
"eval_samples_per_second": 38.79,
|
| 828 |
+
"eval_steps_per_second": 2.425,
|
| 829 |
+
"step": 1120
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 4.035714285714286,
|
| 833 |
+
"grad_norm": 1.5949360132217407,
|
| 834 |
+
"learning_rate": 3.313492063492064e-05,
|
| 835 |
+
"loss": 0.305,
|
| 836 |
+
"step": 1130
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 4.071428571428571,
|
| 840 |
+
"grad_norm": 1.678801417350769,
|
| 841 |
+
"learning_rate": 3.2936507936507936e-05,
|
| 842 |
+
"loss": 0.2677,
|
| 843 |
+
"step": 1140
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 4.107142857142857,
|
| 847 |
+
"grad_norm": 1.6015031337738037,
|
| 848 |
+
"learning_rate": 3.273809523809524e-05,
|
| 849 |
+
"loss": 0.3167,
|
| 850 |
+
"step": 1150
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 4.142857142857143,
|
| 854 |
+
"grad_norm": 1.525894045829773,
|
| 855 |
+
"learning_rate": 3.253968253968254e-05,
|
| 856 |
+
"loss": 0.2745,
|
| 857 |
+
"step": 1160
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 4.178571428571429,
|
| 861 |
+
"grad_norm": 1.2955286502838135,
|
| 862 |
+
"learning_rate": 3.234126984126985e-05,
|
| 863 |
+
"loss": 0.2925,
|
| 864 |
+
"step": 1170
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 4.214285714285714,
|
| 868 |
+
"grad_norm": 2.1041815280914307,
|
| 869 |
+
"learning_rate": 3.2142857142857144e-05,
|
| 870 |
+
"loss": 0.2782,
|
| 871 |
+
"step": 1180
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 4.25,
|
| 875 |
+
"grad_norm": 1.7325941324234009,
|
| 876 |
+
"learning_rate": 3.194444444444444e-05,
|
| 877 |
+
"loss": 0.302,
|
| 878 |
+
"step": 1190
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 4.285714285714286,
|
| 882 |
+
"grad_norm": 1.9786794185638428,
|
| 883 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 884 |
+
"loss": 0.2864,
|
| 885 |
+
"step": 1200
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 4.321428571428571,
|
| 889 |
+
"grad_norm": 1.3869291543960571,
|
| 890 |
+
"learning_rate": 3.154761904761905e-05,
|
| 891 |
+
"loss": 0.2834,
|
| 892 |
+
"step": 1210
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 4.357142857142857,
|
| 896 |
+
"grad_norm": 2.1556620597839355,
|
| 897 |
+
"learning_rate": 3.134920634920635e-05,
|
| 898 |
+
"loss": 0.3011,
|
| 899 |
+
"step": 1220
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 4.392857142857143,
|
| 903 |
+
"grad_norm": 1.6058976650238037,
|
| 904 |
+
"learning_rate": 3.1150793650793656e-05,
|
| 905 |
+
"loss": 0.2893,
|
| 906 |
+
"step": 1230
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 4.428571428571429,
|
| 910 |
+
"grad_norm": 1.9528658390045166,
|
| 911 |
+
"learning_rate": 3.095238095238095e-05,
|
| 912 |
+
"loss": 0.2725,
|
| 913 |
+
"step": 1240
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 4.464285714285714,
|
| 917 |
+
"grad_norm": 1.5343618392944336,
|
| 918 |
+
"learning_rate": 3.075396825396826e-05,
|
| 919 |
+
"loss": 0.2578,
|
| 920 |
+
"step": 1250
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 4.5,
|
| 924 |
+
"grad_norm": 1.0154541730880737,
|
| 925 |
+
"learning_rate": 3.055555555555556e-05,
|
| 926 |
+
"loss": 0.2897,
|
| 927 |
+
"step": 1260
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.535714285714286,
|
| 931 |
+
"grad_norm": 2.355865001678467,
|
| 932 |
+
"learning_rate": 3.0357142857142857e-05,
|
| 933 |
+
"loss": 0.2658,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.571428571428571,
|
| 938 |
+
"grad_norm": 1.24091374874115,
|
| 939 |
+
"learning_rate": 3.0158730158730158e-05,
|
| 940 |
+
"loss": 0.3198,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.607142857142857,
|
| 945 |
+
"grad_norm": 1.0385922193527222,
|
| 946 |
+
"learning_rate": 2.996031746031746e-05,
|
| 947 |
+
"loss": 0.285,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.642857142857143,
|
| 952 |
+
"grad_norm": 1.032619833946228,
|
| 953 |
+
"learning_rate": 2.9761904761904762e-05,
|
| 954 |
+
"loss": 0.2741,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.678571428571429,
|
| 959 |
+
"grad_norm": 1.587849736213684,
|
| 960 |
+
"learning_rate": 2.9563492063492066e-05,
|
| 961 |
+
"loss": 0.3093,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.714285714285714,
|
| 966 |
+
"grad_norm": 1.2873854637145996,
|
| 967 |
+
"learning_rate": 2.9365079365079366e-05,
|
| 968 |
+
"loss": 0.299,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.75,
|
| 973 |
+
"grad_norm": 1.6315929889678955,
|
| 974 |
+
"learning_rate": 2.916666666666667e-05,
|
| 975 |
+
"loss": 0.2776,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.785714285714286,
|
| 980 |
+
"grad_norm": 1.8490331172943115,
|
| 981 |
+
"learning_rate": 2.8968253968253974e-05,
|
| 982 |
+
"loss": 0.2613,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.821428571428571,
|
| 987 |
+
"grad_norm": 1.8352344036102295,
|
| 988 |
+
"learning_rate": 2.876984126984127e-05,
|
| 989 |
+
"loss": 0.281,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.857142857142857,
|
| 994 |
+
"grad_norm": 1.0744656324386597,
|
| 995 |
+
"learning_rate": 2.857142857142857e-05,
|
| 996 |
+
"loss": 0.2622,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.892857142857143,
|
| 1001 |
+
"grad_norm": 1.2477718591690063,
|
| 1002 |
+
"learning_rate": 2.8373015873015875e-05,
|
| 1003 |
+
"loss": 0.2997,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.928571428571429,
|
| 1008 |
+
"grad_norm": 1.5114529132843018,
|
| 1009 |
+
"learning_rate": 2.8174603174603175e-05,
|
| 1010 |
+
"loss": 0.2827,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.964285714285714,
|
| 1015 |
+
"grad_norm": 1.3266760110855103,
|
| 1016 |
+
"learning_rate": 2.797619047619048e-05,
|
| 1017 |
+
"loss": 0.283,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 5.0,
|
| 1022 |
+
"grad_norm": 2.2647006511688232,
|
| 1023 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1024 |
+
"loss": 0.2498,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 5.0,
|
| 1029 |
+
"eval_accuracy": 0.8693613220187584,
|
| 1030 |
+
"eval_loss": 0.30874186754226685,
|
| 1031 |
+
"eval_runtime": 116.4377,
|
| 1032 |
+
"eval_samples_per_second": 38.458,
|
| 1033 |
+
"eval_steps_per_second": 2.405,
|
| 1034 |
+
"step": 1400
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 5.035714285714286,
|
| 1038 |
+
"grad_norm": 1.6704837083816528,
|
| 1039 |
+
"learning_rate": 2.7579365079365083e-05,
|
| 1040 |
+
"loss": 0.2973,
|
| 1041 |
+
"step": 1410
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 5.071428571428571,
|
| 1045 |
+
"grad_norm": 1.9291003942489624,
|
| 1046 |
+
"learning_rate": 2.7380952380952383e-05,
|
| 1047 |
+
"loss": 0.2756,
|
| 1048 |
+
"step": 1420
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 5.107142857142857,
|
| 1052 |
+
"grad_norm": 2.44765043258667,
|
| 1053 |
+
"learning_rate": 2.718253968253968e-05,
|
| 1054 |
+
"loss": 0.2575,
|
| 1055 |
+
"step": 1430
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 5.142857142857143,
|
| 1059 |
+
"grad_norm": 2.0583574771881104,
|
| 1060 |
+
"learning_rate": 2.6984126984126984e-05,
|
| 1061 |
+
"loss": 0.256,
|
| 1062 |
+
"step": 1440
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 5.178571428571429,
|
| 1066 |
+
"grad_norm": 1.8882899284362793,
|
| 1067 |
+
"learning_rate": 2.6785714285714288e-05,
|
| 1068 |
+
"loss": 0.299,
|
| 1069 |
+
"step": 1450
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 5.214285714285714,
|
| 1073 |
+
"grad_norm": 1.5178027153015137,
|
| 1074 |
+
"learning_rate": 2.6587301587301588e-05,
|
| 1075 |
+
"loss": 0.3209,
|
| 1076 |
+
"step": 1460
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 5.25,
|
| 1080 |
+
"grad_norm": 1.4342873096466064,
|
| 1081 |
+
"learning_rate": 2.6388888888888892e-05,
|
| 1082 |
+
"loss": 0.2828,
|
| 1083 |
+
"step": 1470
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 5.285714285714286,
|
| 1087 |
+
"grad_norm": 1.3157316446304321,
|
| 1088 |
+
"learning_rate": 2.6190476190476192e-05,
|
| 1089 |
+
"loss": 0.2655,
|
| 1090 |
+
"step": 1480
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 5.321428571428571,
|
| 1094 |
+
"grad_norm": 1.0939053297042847,
|
| 1095 |
+
"learning_rate": 2.5992063492063496e-05,
|
| 1096 |
+
"loss": 0.278,
|
| 1097 |
+
"step": 1490
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 5.357142857142857,
|
| 1101 |
+
"grad_norm": 1.9465833902359009,
|
| 1102 |
+
"learning_rate": 2.5793650793650796e-05,
|
| 1103 |
+
"loss": 0.2652,
|
| 1104 |
+
"step": 1500
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 5.392857142857143,
|
| 1108 |
+
"grad_norm": 1.5688917636871338,
|
| 1109 |
+
"learning_rate": 2.5595238095238093e-05,
|
| 1110 |
+
"loss": 0.2622,
|
| 1111 |
+
"step": 1510
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 5.428571428571429,
|
| 1115 |
+
"grad_norm": 1.3724976778030396,
|
| 1116 |
+
"learning_rate": 2.5396825396825397e-05,
|
| 1117 |
+
"loss": 0.2524,
|
| 1118 |
+
"step": 1520
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 5.464285714285714,
|
| 1122 |
+
"grad_norm": 1.160733699798584,
|
| 1123 |
+
"learning_rate": 2.5198412698412697e-05,
|
| 1124 |
+
"loss": 0.2868,
|
| 1125 |
+
"step": 1530
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 5.5,
|
| 1129 |
+
"grad_norm": 1.2799668312072754,
|
| 1130 |
+
"learning_rate": 2.5e-05,
|
| 1131 |
+
"loss": 0.2713,
|
| 1132 |
+
"step": 1540
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 5.535714285714286,
|
| 1136 |
+
"grad_norm": 1.9062435626983643,
|
| 1137 |
+
"learning_rate": 2.4801587301587305e-05,
|
| 1138 |
+
"loss": 0.2527,
|
| 1139 |
+
"step": 1550
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 5.571428571428571,
|
| 1143 |
+
"grad_norm": 2.5998220443725586,
|
| 1144 |
+
"learning_rate": 2.4603174603174602e-05,
|
| 1145 |
+
"loss": 0.2668,
|
| 1146 |
+
"step": 1560
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 5.607142857142857,
|
| 1150 |
+
"grad_norm": 2.346576452255249,
|
| 1151 |
+
"learning_rate": 2.4404761904761906e-05,
|
| 1152 |
+
"loss": 0.2557,
|
| 1153 |
+
"step": 1570
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 5.642857142857143,
|
| 1157 |
+
"grad_norm": 1.9609774351119995,
|
| 1158 |
+
"learning_rate": 2.4206349206349206e-05,
|
| 1159 |
+
"loss": 0.2706,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.678571428571429,
|
| 1164 |
+
"grad_norm": 1.8994617462158203,
|
| 1165 |
+
"learning_rate": 2.400793650793651e-05,
|
| 1166 |
+
"loss": 0.2878,
|
| 1167 |
+
"step": 1590
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 5.714285714285714,
|
| 1171 |
+
"grad_norm": 1.677849531173706,
|
| 1172 |
+
"learning_rate": 2.380952380952381e-05,
|
| 1173 |
+
"loss": 0.2761,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 5.75,
|
| 1178 |
+
"grad_norm": 1.6345635652542114,
|
| 1179 |
+
"learning_rate": 2.361111111111111e-05,
|
| 1180 |
+
"loss": 0.2231,
|
| 1181 |
+
"step": 1610
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 5.785714285714286,
|
| 1185 |
+
"grad_norm": 1.6198554039001465,
|
| 1186 |
+
"learning_rate": 2.3412698412698414e-05,
|
| 1187 |
+
"loss": 0.2612,
|
| 1188 |
+
"step": 1620
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 5.821428571428571,
|
| 1192 |
+
"grad_norm": 2.052764415740967,
|
| 1193 |
+
"learning_rate": 2.3214285714285715e-05,
|
| 1194 |
+
"loss": 0.2738,
|
| 1195 |
+
"step": 1630
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 5.857142857142857,
|
| 1199 |
+
"grad_norm": 1.0596450567245483,
|
| 1200 |
+
"learning_rate": 2.3015873015873015e-05,
|
| 1201 |
+
"loss": 0.2597,
|
| 1202 |
+
"step": 1640
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 5.892857142857143,
|
| 1206 |
+
"grad_norm": 1.9113609790802002,
|
| 1207 |
+
"learning_rate": 2.281746031746032e-05,
|
| 1208 |
+
"loss": 0.2894,
|
| 1209 |
+
"step": 1650
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 5.928571428571429,
|
| 1213 |
+
"grad_norm": 1.7055829763412476,
|
| 1214 |
+
"learning_rate": 2.261904761904762e-05,
|
| 1215 |
+
"loss": 0.226,
|
| 1216 |
+
"step": 1660
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 5.964285714285714,
|
| 1220 |
+
"grad_norm": 0.969464898109436,
|
| 1221 |
+
"learning_rate": 2.2420634920634923e-05,
|
| 1222 |
+
"loss": 0.2977,
|
| 1223 |
+
"step": 1670
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 6.0,
|
| 1227 |
+
"grad_norm": 3.1485939025878906,
|
| 1228 |
+
"learning_rate": 2.2222222222222223e-05,
|
| 1229 |
+
"loss": 0.3028,
|
| 1230 |
+
"step": 1680
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 6.0,
|
| 1234 |
+
"eval_accuracy": 0.8715944618133095,
|
| 1235 |
+
"eval_loss": 0.29658079147338867,
|
| 1236 |
+
"eval_runtime": 113.6842,
|
| 1237 |
+
"eval_samples_per_second": 39.39,
|
| 1238 |
+
"eval_steps_per_second": 2.463,
|
| 1239 |
+
"step": 1680
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 6.035714285714286,
|
| 1243 |
+
"grad_norm": 1.4179354906082153,
|
| 1244 |
+
"learning_rate": 2.2023809523809524e-05,
|
| 1245 |
+
"loss": 0.2681,
|
| 1246 |
+
"step": 1690
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 6.071428571428571,
|
| 1250 |
+
"grad_norm": 2.099681854248047,
|
| 1251 |
+
"learning_rate": 2.1825396825396827e-05,
|
| 1252 |
+
"loss": 0.2441,
|
| 1253 |
+
"step": 1700
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 6.107142857142857,
|
| 1257 |
+
"grad_norm": 1.5406619310379028,
|
| 1258 |
+
"learning_rate": 2.1626984126984128e-05,
|
| 1259 |
+
"loss": 0.2707,
|
| 1260 |
+
"step": 1710
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 6.142857142857143,
|
| 1264 |
+
"grad_norm": 2.4518625736236572,
|
| 1265 |
+
"learning_rate": 2.1428571428571428e-05,
|
| 1266 |
+
"loss": 0.2942,
|
| 1267 |
+
"step": 1720
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 6.178571428571429,
|
| 1271 |
+
"grad_norm": 1.1450200080871582,
|
| 1272 |
+
"learning_rate": 2.1230158730158732e-05,
|
| 1273 |
+
"loss": 0.2336,
|
| 1274 |
+
"step": 1730
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 6.214285714285714,
|
| 1278 |
+
"grad_norm": 1.4170438051223755,
|
| 1279 |
+
"learning_rate": 2.1031746031746032e-05,
|
| 1280 |
+
"loss": 0.2857,
|
| 1281 |
+
"step": 1740
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 6.25,
|
| 1285 |
+
"grad_norm": 1.2231560945510864,
|
| 1286 |
+
"learning_rate": 2.0833333333333336e-05,
|
| 1287 |
+
"loss": 0.2507,
|
| 1288 |
+
"step": 1750
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 6.285714285714286,
|
| 1292 |
+
"grad_norm": 1.4597039222717285,
|
| 1293 |
+
"learning_rate": 2.0634920634920636e-05,
|
| 1294 |
+
"loss": 0.2688,
|
| 1295 |
+
"step": 1760
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 6.321428571428571,
|
| 1299 |
+
"grad_norm": 1.9395854473114014,
|
| 1300 |
+
"learning_rate": 2.0436507936507937e-05,
|
| 1301 |
+
"loss": 0.2715,
|
| 1302 |
+
"step": 1770
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 6.357142857142857,
|
| 1306 |
+
"grad_norm": 0.9649907946586609,
|
| 1307 |
+
"learning_rate": 2.023809523809524e-05,
|
| 1308 |
+
"loss": 0.2508,
|
| 1309 |
+
"step": 1780
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 6.392857142857143,
|
| 1313 |
+
"grad_norm": 1.3842933177947998,
|
| 1314 |
+
"learning_rate": 2.003968253968254e-05,
|
| 1315 |
+
"loss": 0.2818,
|
| 1316 |
+
"step": 1790
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 6.428571428571429,
|
| 1320 |
+
"grad_norm": 1.7550239562988281,
|
| 1321 |
+
"learning_rate": 1.984126984126984e-05,
|
| 1322 |
+
"loss": 0.2553,
|
| 1323 |
+
"step": 1800
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 6.464285714285714,
|
| 1327 |
+
"grad_norm": 1.04232919216156,
|
| 1328 |
+
"learning_rate": 1.9642857142857145e-05,
|
| 1329 |
+
"loss": 0.2303,
|
| 1330 |
+
"step": 1810
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 6.5,
|
| 1334 |
+
"grad_norm": 2.306325674057007,
|
| 1335 |
+
"learning_rate": 1.9444444444444445e-05,
|
| 1336 |
+
"loss": 0.257,
|
| 1337 |
+
"step": 1820
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 6.535714285714286,
|
| 1341 |
+
"grad_norm": 1.704186201095581,
|
| 1342 |
+
"learning_rate": 1.924603174603175e-05,
|
| 1343 |
+
"loss": 0.2844,
|
| 1344 |
+
"step": 1830
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 6.571428571428571,
|
| 1348 |
+
"grad_norm": 1.3930236101150513,
|
| 1349 |
+
"learning_rate": 1.9047619047619046e-05,
|
| 1350 |
+
"loss": 0.2784,
|
| 1351 |
+
"step": 1840
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 6.607142857142857,
|
| 1355 |
+
"grad_norm": 1.5494309663772583,
|
| 1356 |
+
"learning_rate": 1.884920634920635e-05,
|
| 1357 |
+
"loss": 0.2735,
|
| 1358 |
+
"step": 1850
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 6.642857142857143,
|
| 1362 |
+
"grad_norm": 1.5971505641937256,
|
| 1363 |
+
"learning_rate": 1.8650793650793654e-05,
|
| 1364 |
+
"loss": 0.2466,
|
| 1365 |
+
"step": 1860
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 6.678571428571429,
|
| 1369 |
+
"grad_norm": 1.8419290781021118,
|
| 1370 |
+
"learning_rate": 1.8452380952380954e-05,
|
| 1371 |
+
"loss": 0.2476,
|
| 1372 |
+
"step": 1870
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 6.714285714285714,
|
| 1376 |
+
"grad_norm": 1.3679064512252808,
|
| 1377 |
+
"learning_rate": 1.8253968253968254e-05,
|
| 1378 |
+
"loss": 0.2277,
|
| 1379 |
+
"step": 1880
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 6.75,
|
| 1383 |
+
"grad_norm": 1.1586477756500244,
|
| 1384 |
+
"learning_rate": 1.8055555555555555e-05,
|
| 1385 |
+
"loss": 0.2266,
|
| 1386 |
+
"step": 1890
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 6.785714285714286,
|
| 1390 |
+
"grad_norm": 1.9979654550552368,
|
| 1391 |
+
"learning_rate": 1.785714285714286e-05,
|
| 1392 |
+
"loss": 0.2747,
|
| 1393 |
+
"step": 1900
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 6.821428571428571,
|
| 1397 |
+
"grad_norm": 1.4875764846801758,
|
| 1398 |
+
"learning_rate": 1.7658730158730162e-05,
|
| 1399 |
+
"loss": 0.2666,
|
| 1400 |
+
"step": 1910
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 6.857142857142857,
|
| 1404 |
+
"grad_norm": 1.6196482181549072,
|
| 1405 |
+
"learning_rate": 1.746031746031746e-05,
|
| 1406 |
+
"loss": 0.2532,
|
| 1407 |
+
"step": 1920
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 6.892857142857143,
|
| 1411 |
+
"grad_norm": 2.7135918140411377,
|
| 1412 |
+
"learning_rate": 1.7261904761904763e-05,
|
| 1413 |
+
"loss": 0.2265,
|
| 1414 |
+
"step": 1930
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 6.928571428571429,
|
| 1418 |
+
"grad_norm": 1.423257827758789,
|
| 1419 |
+
"learning_rate": 1.7063492063492063e-05,
|
| 1420 |
+
"loss": 0.2621,
|
| 1421 |
+
"step": 1940
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 6.964285714285714,
|
| 1425 |
+
"grad_norm": 2.1509084701538086,
|
| 1426 |
+
"learning_rate": 1.6865079365079367e-05,
|
| 1427 |
+
"loss": 0.2636,
|
| 1428 |
+
"step": 1950
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 7.0,
|
| 1432 |
+
"grad_norm": 1.562657117843628,
|
| 1433 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 1434 |
+
"loss": 0.2179,
|
| 1435 |
+
"step": 1960
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 7.0,
|
| 1439 |
+
"eval_accuracy": 0.8807503349709692,
|
| 1440 |
+
"eval_loss": 0.2742190361022949,
|
| 1441 |
+
"eval_runtime": 113.8381,
|
| 1442 |
+
"eval_samples_per_second": 39.337,
|
| 1443 |
+
"eval_steps_per_second": 2.46,
|
| 1444 |
+
"step": 1960
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 7.035714285714286,
|
| 1448 |
+
"grad_norm": 1.4138425588607788,
|
| 1449 |
+
"learning_rate": 1.6468253968253968e-05,
|
| 1450 |
+
"loss": 0.2251,
|
| 1451 |
+
"step": 1970
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 7.071428571428571,
|
| 1455 |
+
"grad_norm": 2.6958255767822266,
|
| 1456 |
+
"learning_rate": 1.626984126984127e-05,
|
| 1457 |
+
"loss": 0.2521,
|
| 1458 |
+
"step": 1980
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 7.107142857142857,
|
| 1462 |
+
"grad_norm": 2.014803647994995,
|
| 1463 |
+
"learning_rate": 1.6071428571428572e-05,
|
| 1464 |
+
"loss": 0.216,
|
| 1465 |
+
"step": 1990
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 7.142857142857143,
|
| 1469 |
+
"grad_norm": 2.1656410694122314,
|
| 1470 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 1471 |
+
"loss": 0.2085,
|
| 1472 |
+
"step": 2000
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 7.178571428571429,
|
| 1476 |
+
"grad_norm": 2.207980155944824,
|
| 1477 |
+
"learning_rate": 1.5674603174603176e-05,
|
| 1478 |
+
"loss": 0.263,
|
| 1479 |
+
"step": 2010
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 7.214285714285714,
|
| 1483 |
+
"grad_norm": 1.978162169456482,
|
| 1484 |
+
"learning_rate": 1.5476190476190476e-05,
|
| 1485 |
+
"loss": 0.2478,
|
| 1486 |
+
"step": 2020
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 7.25,
|
| 1490 |
+
"grad_norm": 1.2705848217010498,
|
| 1491 |
+
"learning_rate": 1.527777777777778e-05,
|
| 1492 |
+
"loss": 0.2338,
|
| 1493 |
+
"step": 2030
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 7.285714285714286,
|
| 1497 |
+
"grad_norm": 1.6287261247634888,
|
| 1498 |
+
"learning_rate": 1.5079365079365079e-05,
|
| 1499 |
+
"loss": 0.2063,
|
| 1500 |
+
"step": 2040
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 7.321428571428571,
|
| 1504 |
+
"grad_norm": 2.0518875122070312,
|
| 1505 |
+
"learning_rate": 1.4880952380952381e-05,
|
| 1506 |
+
"loss": 0.3032,
|
| 1507 |
+
"step": 2050
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 7.357142857142857,
|
| 1511 |
+
"grad_norm": 2.7869536876678467,
|
| 1512 |
+
"learning_rate": 1.4682539682539683e-05,
|
| 1513 |
+
"loss": 0.2242,
|
| 1514 |
+
"step": 2060
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 7.392857142857143,
|
| 1518 |
+
"grad_norm": 1.007391095161438,
|
| 1519 |
+
"learning_rate": 1.4484126984126987e-05,
|
| 1520 |
+
"loss": 0.2589,
|
| 1521 |
+
"step": 2070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 7.428571428571429,
|
| 1525 |
+
"grad_norm": 1.2812851667404175,
|
| 1526 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 1527 |
+
"loss": 0.2383,
|
| 1528 |
+
"step": 2080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 7.464285714285714,
|
| 1532 |
+
"grad_norm": 1.1465330123901367,
|
| 1533 |
+
"learning_rate": 1.4087301587301587e-05,
|
| 1534 |
+
"loss": 0.2284,
|
| 1535 |
+
"step": 2090
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 7.5,
|
| 1539 |
+
"grad_norm": 2.2567813396453857,
|
| 1540 |
+
"learning_rate": 1.388888888888889e-05,
|
| 1541 |
+
"loss": 0.213,
|
| 1542 |
+
"step": 2100
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 7.535714285714286,
|
| 1546 |
+
"grad_norm": 1.7949641942977905,
|
| 1547 |
+
"learning_rate": 1.3690476190476192e-05,
|
| 1548 |
+
"loss": 0.2368,
|
| 1549 |
+
"step": 2110
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 7.571428571428571,
|
| 1553 |
+
"grad_norm": 2.443598985671997,
|
| 1554 |
+
"learning_rate": 1.3492063492063492e-05,
|
| 1555 |
+
"loss": 0.2665,
|
| 1556 |
+
"step": 2120
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 7.607142857142857,
|
| 1560 |
+
"grad_norm": 1.9699336290359497,
|
| 1561 |
+
"learning_rate": 1.3293650793650794e-05,
|
| 1562 |
+
"loss": 0.2346,
|
| 1563 |
+
"step": 2130
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 7.642857142857143,
|
| 1567 |
+
"grad_norm": 1.543039083480835,
|
| 1568 |
+
"learning_rate": 1.3095238095238096e-05,
|
| 1569 |
+
"loss": 0.2432,
|
| 1570 |
+
"step": 2140
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 7.678571428571429,
|
| 1574 |
+
"grad_norm": 1.9814691543579102,
|
| 1575 |
+
"learning_rate": 1.2896825396825398e-05,
|
| 1576 |
+
"loss": 0.2575,
|
| 1577 |
+
"step": 2150
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 7.714285714285714,
|
| 1581 |
+
"grad_norm": 2.1088602542877197,
|
| 1582 |
+
"learning_rate": 1.2698412698412699e-05,
|
| 1583 |
+
"loss": 0.2346,
|
| 1584 |
+
"step": 2160
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 7.75,
|
| 1588 |
+
"grad_norm": 1.5648256540298462,
|
| 1589 |
+
"learning_rate": 1.25e-05,
|
| 1590 |
+
"loss": 0.2403,
|
| 1591 |
+
"step": 2170
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 7.785714285714286,
|
| 1595 |
+
"grad_norm": 1.6079583168029785,
|
| 1596 |
+
"learning_rate": 1.2301587301587301e-05,
|
| 1597 |
+
"loss": 0.2414,
|
| 1598 |
+
"step": 2180
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 7.821428571428571,
|
| 1602 |
+
"grad_norm": 1.2859593629837036,
|
| 1603 |
+
"learning_rate": 1.2103174603174603e-05,
|
| 1604 |
+
"loss": 0.2756,
|
| 1605 |
+
"step": 2190
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 7.857142857142857,
|
| 1609 |
+
"grad_norm": 2.072089672088623,
|
| 1610 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 1611 |
+
"loss": 0.2484,
|
| 1612 |
+
"step": 2200
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 7.892857142857143,
|
| 1616 |
+
"grad_norm": 1.623353362083435,
|
| 1617 |
+
"learning_rate": 1.1706349206349207e-05,
|
| 1618 |
+
"loss": 0.274,
|
| 1619 |
+
"step": 2210
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 7.928571428571429,
|
| 1623 |
+
"grad_norm": 1.70241379737854,
|
| 1624 |
+
"learning_rate": 1.1507936507936508e-05,
|
| 1625 |
+
"loss": 0.2307,
|
| 1626 |
+
"step": 2220
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 7.964285714285714,
|
| 1630 |
+
"grad_norm": 1.7186700105667114,
|
| 1631 |
+
"learning_rate": 1.130952380952381e-05,
|
| 1632 |
+
"loss": 0.2223,
|
| 1633 |
+
"step": 2230
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 8.0,
|
| 1637 |
+
"grad_norm": 2.0665862560272217,
|
| 1638 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1639 |
+
"loss": 0.2274,
|
| 1640 |
+
"step": 2240
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 8.0,
|
| 1644 |
+
"eval_accuracy": 0.8814202769093346,
|
| 1645 |
+
"eval_loss": 0.2860513925552368,
|
| 1646 |
+
"eval_runtime": 115.9278,
|
| 1647 |
+
"eval_samples_per_second": 38.627,
|
| 1648 |
+
"eval_steps_per_second": 2.415,
|
| 1649 |
+
"step": 2240
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 8.035714285714286,
|
| 1653 |
+
"grad_norm": 1.3822689056396484,
|
| 1654 |
+
"learning_rate": 1.0912698412698414e-05,
|
| 1655 |
+
"loss": 0.238,
|
| 1656 |
+
"step": 2250
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 8.071428571428571,
|
| 1660 |
+
"grad_norm": 1.4226853847503662,
|
| 1661 |
+
"learning_rate": 1.0714285714285714e-05,
|
| 1662 |
+
"loss": 0.266,
|
| 1663 |
+
"step": 2260
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 8.107142857142858,
|
| 1667 |
+
"grad_norm": 1.9675803184509277,
|
| 1668 |
+
"learning_rate": 1.0515873015873016e-05,
|
| 1669 |
+
"loss": 0.2814,
|
| 1670 |
+
"step": 2270
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 8.142857142857142,
|
| 1674 |
+
"grad_norm": 1.8133440017700195,
|
| 1675 |
+
"learning_rate": 1.0317460317460318e-05,
|
| 1676 |
+
"loss": 0.2486,
|
| 1677 |
+
"step": 2280
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 8.178571428571429,
|
| 1681 |
+
"grad_norm": 1.2323780059814453,
|
| 1682 |
+
"learning_rate": 1.011904761904762e-05,
|
| 1683 |
+
"loss": 0.2175,
|
| 1684 |
+
"step": 2290
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 8.214285714285714,
|
| 1688 |
+
"grad_norm": 1.7063086032867432,
|
| 1689 |
+
"learning_rate": 9.92063492063492e-06,
|
| 1690 |
+
"loss": 0.2125,
|
| 1691 |
+
"step": 2300
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 8.25,
|
| 1695 |
+
"grad_norm": 1.532769799232483,
|
| 1696 |
+
"learning_rate": 9.722222222222223e-06,
|
| 1697 |
+
"loss": 0.1813,
|
| 1698 |
+
"step": 2310
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 8.285714285714286,
|
| 1702 |
+
"grad_norm": 1.5620160102844238,
|
| 1703 |
+
"learning_rate": 9.523809523809523e-06,
|
| 1704 |
+
"loss": 0.2355,
|
| 1705 |
+
"step": 2320
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 8.321428571428571,
|
| 1709 |
+
"grad_norm": 1.6230847835540771,
|
| 1710 |
+
"learning_rate": 9.325396825396827e-06,
|
| 1711 |
+
"loss": 0.2199,
|
| 1712 |
+
"step": 2330
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 8.357142857142858,
|
| 1716 |
+
"grad_norm": 1.552085041999817,
|
| 1717 |
+
"learning_rate": 9.126984126984127e-06,
|
| 1718 |
+
"loss": 0.2479,
|
| 1719 |
+
"step": 2340
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 8.392857142857142,
|
| 1723 |
+
"grad_norm": 1.8006685972213745,
|
| 1724 |
+
"learning_rate": 8.92857142857143e-06,
|
| 1725 |
+
"loss": 0.2677,
|
| 1726 |
+
"step": 2350
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 8.428571428571429,
|
| 1730 |
+
"grad_norm": 1.2057029008865356,
|
| 1731 |
+
"learning_rate": 8.73015873015873e-06,
|
| 1732 |
+
"loss": 0.2022,
|
| 1733 |
+
"step": 2360
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 8.464285714285714,
|
| 1737 |
+
"grad_norm": 1.4805638790130615,
|
| 1738 |
+
"learning_rate": 8.531746031746032e-06,
|
| 1739 |
+
"loss": 0.2158,
|
| 1740 |
+
"step": 2370
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 8.5,
|
| 1744 |
+
"grad_norm": 1.441603422164917,
|
| 1745 |
+
"learning_rate": 8.333333333333334e-06,
|
| 1746 |
+
"loss": 0.2166,
|
| 1747 |
+
"step": 2380
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 8.535714285714286,
|
| 1751 |
+
"grad_norm": 2.171687602996826,
|
| 1752 |
+
"learning_rate": 8.134920634920636e-06,
|
| 1753 |
+
"loss": 0.2332,
|
| 1754 |
+
"step": 2390
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 8.571428571428571,
|
| 1758 |
+
"grad_norm": 1.5523641109466553,
|
| 1759 |
+
"learning_rate": 7.936507936507936e-06,
|
| 1760 |
+
"loss": 0.2099,
|
| 1761 |
+
"step": 2400
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 8.607142857142858,
|
| 1765 |
+
"grad_norm": 1.429527759552002,
|
| 1766 |
+
"learning_rate": 7.738095238095238e-06,
|
| 1767 |
+
"loss": 0.1934,
|
| 1768 |
+
"step": 2410
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 8.642857142857142,
|
| 1772 |
+
"grad_norm": 3.8555209636688232,
|
| 1773 |
+
"learning_rate": 7.5396825396825394e-06,
|
| 1774 |
+
"loss": 0.2813,
|
| 1775 |
+
"step": 2420
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 8.678571428571429,
|
| 1779 |
+
"grad_norm": 1.118416666984558,
|
| 1780 |
+
"learning_rate": 7.3412698412698415e-06,
|
| 1781 |
+
"loss": 0.2358,
|
| 1782 |
+
"step": 2430
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 8.714285714285714,
|
| 1786 |
+
"grad_norm": 1.9037988185882568,
|
| 1787 |
+
"learning_rate": 7.142857142857143e-06,
|
| 1788 |
+
"loss": 0.2091,
|
| 1789 |
+
"step": 2440
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 8.75,
|
| 1793 |
+
"grad_norm": 1.8768919706344604,
|
| 1794 |
+
"learning_rate": 6.944444444444445e-06,
|
| 1795 |
+
"loss": 0.2176,
|
| 1796 |
+
"step": 2450
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 8.785714285714286,
|
| 1800 |
+
"grad_norm": 1.606123685836792,
|
| 1801 |
+
"learning_rate": 6.746031746031746e-06,
|
| 1802 |
+
"loss": 0.2388,
|
| 1803 |
+
"step": 2460
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 8.821428571428571,
|
| 1807 |
+
"grad_norm": 2.6942107677459717,
|
| 1808 |
+
"learning_rate": 6.547619047619048e-06,
|
| 1809 |
+
"loss": 0.2182,
|
| 1810 |
+
"step": 2470
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 8.857142857142858,
|
| 1814 |
+
"grad_norm": 1.449601173400879,
|
| 1815 |
+
"learning_rate": 6.349206349206349e-06,
|
| 1816 |
+
"loss": 0.2102,
|
| 1817 |
+
"step": 2480
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 8.892857142857142,
|
| 1821 |
+
"grad_norm": 2.651686668395996,
|
| 1822 |
+
"learning_rate": 6.1507936507936505e-06,
|
| 1823 |
+
"loss": 0.2166,
|
| 1824 |
+
"step": 2490
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 8.928571428571429,
|
| 1828 |
+
"grad_norm": 2.6743762493133545,
|
| 1829 |
+
"learning_rate": 5.9523809523809525e-06,
|
| 1830 |
+
"loss": 0.2468,
|
| 1831 |
+
"step": 2500
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 8.964285714285714,
|
| 1835 |
+
"grad_norm": 2.29903244972229,
|
| 1836 |
+
"learning_rate": 5.753968253968254e-06,
|
| 1837 |
+
"loss": 0.2111,
|
| 1838 |
+
"step": 2510
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 9.0,
|
| 1842 |
+
"grad_norm": 1.8269622325897217,
|
| 1843 |
+
"learning_rate": 5.555555555555556e-06,
|
| 1844 |
+
"loss": 0.2195,
|
| 1845 |
+
"step": 2520
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 9.0,
|
| 1849 |
+
"eval_accuracy": 0.8894595801697186,
|
| 1850 |
+
"eval_loss": 0.26261791586875916,
|
| 1851 |
+
"eval_runtime": 114.2956,
|
| 1852 |
+
"eval_samples_per_second": 39.179,
|
| 1853 |
+
"eval_steps_per_second": 2.45,
|
| 1854 |
+
"step": 2520
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 9.035714285714286,
|
| 1858 |
+
"grad_norm": 3.011329174041748,
|
| 1859 |
+
"learning_rate": 5.357142857142857e-06,
|
| 1860 |
+
"loss": 0.1827,
|
| 1861 |
+
"step": 2530
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 9.071428571428571,
|
| 1865 |
+
"grad_norm": 2.3060262203216553,
|
| 1866 |
+
"learning_rate": 5.158730158730159e-06,
|
| 1867 |
+
"loss": 0.1862,
|
| 1868 |
+
"step": 2540
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 9.107142857142858,
|
| 1872 |
+
"grad_norm": 1.7220100164413452,
|
| 1873 |
+
"learning_rate": 4.96031746031746e-06,
|
| 1874 |
+
"loss": 0.2215,
|
| 1875 |
+
"step": 2550
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 9.142857142857142,
|
| 1879 |
+
"grad_norm": 2.463092803955078,
|
| 1880 |
+
"learning_rate": 4.7619047619047615e-06,
|
| 1881 |
+
"loss": 0.2228,
|
| 1882 |
+
"step": 2560
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 9.178571428571429,
|
| 1886 |
+
"grad_norm": 1.204136848449707,
|
| 1887 |
+
"learning_rate": 4.563492063492064e-06,
|
| 1888 |
+
"loss": 0.2439,
|
| 1889 |
+
"step": 2570
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 9.214285714285714,
|
| 1893 |
+
"grad_norm": 2.263396978378296,
|
| 1894 |
+
"learning_rate": 4.365079365079365e-06,
|
| 1895 |
+
"loss": 0.2266,
|
| 1896 |
+
"step": 2580
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 9.25,
|
| 1900 |
+
"grad_norm": 2.7832555770874023,
|
| 1901 |
+
"learning_rate": 4.166666666666667e-06,
|
| 1902 |
+
"loss": 0.2089,
|
| 1903 |
+
"step": 2590
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 9.285714285714286,
|
| 1907 |
+
"grad_norm": 2.0564024448394775,
|
| 1908 |
+
"learning_rate": 3.968253968253968e-06,
|
| 1909 |
+
"loss": 0.1839,
|
| 1910 |
+
"step": 2600
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 9.321428571428571,
|
| 1914 |
+
"grad_norm": 2.0316998958587646,
|
| 1915 |
+
"learning_rate": 3.7698412698412697e-06,
|
| 1916 |
+
"loss": 0.2424,
|
| 1917 |
+
"step": 2610
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 9.357142857142858,
|
| 1921 |
+
"grad_norm": 2.229687213897705,
|
| 1922 |
+
"learning_rate": 3.5714285714285714e-06,
|
| 1923 |
+
"loss": 0.2489,
|
| 1924 |
+
"step": 2620
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 9.392857142857142,
|
| 1928 |
+
"grad_norm": 1.7529199123382568,
|
| 1929 |
+
"learning_rate": 3.373015873015873e-06,
|
| 1930 |
+
"loss": 0.2273,
|
| 1931 |
+
"step": 2630
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 9.428571428571429,
|
| 1935 |
+
"grad_norm": 1.5242239236831665,
|
| 1936 |
+
"learning_rate": 3.1746031746031746e-06,
|
| 1937 |
+
"loss": 0.2224,
|
| 1938 |
+
"step": 2640
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 9.464285714285714,
|
| 1942 |
+
"grad_norm": 1.5499508380889893,
|
| 1943 |
+
"learning_rate": 2.9761904761904763e-06,
|
| 1944 |
+
"loss": 0.2338,
|
| 1945 |
+
"step": 2650
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 9.5,
|
| 1949 |
+
"grad_norm": 2.5902230739593506,
|
| 1950 |
+
"learning_rate": 2.777777777777778e-06,
|
| 1951 |
+
"loss": 0.1909,
|
| 1952 |
+
"step": 2660
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 9.535714285714286,
|
| 1956 |
+
"grad_norm": 1.3243242502212524,
|
| 1957 |
+
"learning_rate": 2.5793650793650795e-06,
|
| 1958 |
+
"loss": 0.2199,
|
| 1959 |
+
"step": 2670
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 9.571428571428571,
|
| 1963 |
+
"grad_norm": 1.9745112657546997,
|
| 1964 |
+
"learning_rate": 2.3809523809523808e-06,
|
| 1965 |
+
"loss": 0.2172,
|
| 1966 |
+
"step": 2680
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 9.607142857142858,
|
| 1970 |
+
"grad_norm": 2.10951828956604,
|
| 1971 |
+
"learning_rate": 2.1825396825396824e-06,
|
| 1972 |
+
"loss": 0.2331,
|
| 1973 |
+
"step": 2690
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 9.642857142857142,
|
| 1977 |
+
"grad_norm": 2.113539457321167,
|
| 1978 |
+
"learning_rate": 1.984126984126984e-06,
|
| 1979 |
+
"loss": 0.2373,
|
| 1980 |
+
"step": 2700
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 9.678571428571429,
|
| 1984 |
+
"grad_norm": 1.548854112625122,
|
| 1985 |
+
"learning_rate": 1.7857142857142857e-06,
|
| 1986 |
+
"loss": 0.2035,
|
| 1987 |
+
"step": 2710
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 9.714285714285714,
|
| 1991 |
+
"grad_norm": 1.797196626663208,
|
| 1992 |
+
"learning_rate": 1.5873015873015873e-06,
|
| 1993 |
+
"loss": 0.2015,
|
| 1994 |
+
"step": 2720
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 9.75,
|
| 1998 |
+
"grad_norm": 1.9279841184616089,
|
| 1999 |
+
"learning_rate": 1.388888888888889e-06,
|
| 2000 |
+
"loss": 0.2208,
|
| 2001 |
+
"step": 2730
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 9.785714285714286,
|
| 2005 |
+
"grad_norm": 0.950290858745575,
|
| 2006 |
+
"learning_rate": 1.1904761904761904e-06,
|
| 2007 |
+
"loss": 0.2394,
|
| 2008 |
+
"step": 2740
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 9.821428571428571,
|
| 2012 |
+
"grad_norm": 1.2573094367980957,
|
| 2013 |
+
"learning_rate": 9.92063492063492e-07,
|
| 2014 |
+
"loss": 0.2491,
|
| 2015 |
+
"step": 2750
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 9.857142857142858,
|
| 2019 |
+
"grad_norm": 1.8162927627563477,
|
| 2020 |
+
"learning_rate": 7.936507936507937e-07,
|
| 2021 |
+
"loss": 0.2229,
|
| 2022 |
+
"step": 2760
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 9.892857142857142,
|
| 2026 |
+
"grad_norm": 2.0845260620117188,
|
| 2027 |
+
"learning_rate": 5.952380952380952e-07,
|
| 2028 |
+
"loss": 0.1984,
|
| 2029 |
+
"step": 2770
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 9.928571428571429,
|
| 2033 |
+
"grad_norm": 1.8501282930374146,
|
| 2034 |
+
"learning_rate": 3.9682539682539683e-07,
|
| 2035 |
+
"loss": 0.2287,
|
| 2036 |
+
"step": 2780
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 9.964285714285714,
|
| 2040 |
+
"grad_norm": 2.941807746887207,
|
| 2041 |
+
"learning_rate": 1.9841269841269841e-07,
|
| 2042 |
+
"loss": 0.2233,
|
| 2043 |
+
"step": 2790
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 10.0,
|
| 2047 |
+
"grad_norm": 1.2508630752563477,
|
| 2048 |
+
"learning_rate": 0.0,
|
| 2049 |
+
"loss": 0.1886,
|
| 2050 |
+
"step": 2800
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 10.0,
|
| 2054 |
+
"eval_accuracy": 0.8865564984368022,
|
| 2055 |
+
"eval_loss": 0.2717145085334778,
|
| 2056 |
+
"eval_runtime": 113.959,
|
| 2057 |
+
"eval_samples_per_second": 39.295,
|
| 2058 |
+
"eval_steps_per_second": 2.457,
|
| 2059 |
+
"step": 2800
|
| 2060 |
},
|
| 2061 |
{
|
| 2062 |
+
"epoch": 10.0,
|
| 2063 |
+
"step": 2800,
|
| 2064 |
+
"total_flos": 1.3877265500181135e+19,
|
| 2065 |
+
"train_loss": 0.3087472263830049,
|
| 2066 |
+
"train_runtime": 6534.1253,
|
| 2067 |
+
"train_samples_per_second": 27.407,
|
| 2068 |
+
"train_steps_per_second": 0.429
|
| 2069 |
}
|
| 2070 |
],
|
| 2071 |
"logging_steps": 10,
|
| 2072 |
+
"max_steps": 2800,
|
| 2073 |
"num_input_tokens_seen": 0,
|
| 2074 |
"num_train_epochs": 10,
|
| 2075 |
"save_steps": 500,
|
|
|
|
| 2085 |
"attributes": {}
|
| 2086 |
}
|
| 2087 |
},
|
| 2088 |
+
"total_flos": 1.3877265500181135e+19,
|
| 2089 |
"train_batch_size": 16,
|
| 2090 |
"trial_name": null,
|
| 2091 |
"trial_params": null
|