Training in progress, epoch 0, checkpoint
Browse files
last-checkpoint/config.json
CHANGED
|
@@ -16,6 +16,6 @@
|
|
| 16 |
"norm_eps": 1e-05,
|
| 17 |
"swiglu_multiple_of": 256,
|
| 18 |
"torch_dtype": "float32",
|
| 19 |
-
"transformers_version": "4.
|
| 20 |
"vocab_size": 37
|
| 21 |
}
|
|
|
|
| 16 |
"norm_eps": 1e-05,
|
| 17 |
"swiglu_multiple_of": 256,
|
| 18 |
"torch_dtype": "float32",
|
| 19 |
+
"transformers_version": "4.49.0",
|
| 20 |
"vocab_size": 37
|
| 21 |
}
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2682482800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d8e1016a9b1bb6828135669bed415800d12ec729a55e57ed78cb1ee7ab1bbc2
|
| 3 |
size 2682482800
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5365108834
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd973e6941dbfc054a82a8fc125c7daff458020cf56c2f24a45ed7a61ec17d74
|
| 3 |
size 5365108834
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15006
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:08baa7817becd0350efcb73cd20230a3510c2a4c44eabb35442e644589d91b4e
|
| 3 |
size 15006
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8196dfb311939b1d2d31afb1d2e2c773958196863e2aef7e228bd8c79258297b
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,826 +1,91 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 50,
|
| 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": 9.
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"eval_loss": 1.
|
| 21 |
-
"eval_runtime":
|
| 22 |
-
"eval_samples_per_second":
|
| 23 |
-
"eval_steps_per_second":
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"epoch": 0.
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 9.
|
| 30 |
-
"loss": 1.
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
-
"epoch": 0.
|
| 35 |
-
"eval_loss": 1.
|
| 36 |
-
"eval_runtime":
|
| 37 |
-
"eval_samples_per_second":
|
| 38 |
-
"eval_steps_per_second":
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"grad_norm":
|
| 44 |
-
"learning_rate": 9.
|
| 45 |
-
"loss": 1.
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
-
"epoch": 0.
|
| 50 |
-
"eval_loss": 1.
|
| 51 |
-
"eval_runtime":
|
| 52 |
-
"eval_samples_per_second":
|
| 53 |
-
"eval_steps_per_second":
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
-
"epoch": 0.
|
| 58 |
-
"grad_norm":
|
| 59 |
-
"learning_rate": 9.
|
| 60 |
-
"loss": 1.
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
-
"epoch": 0.
|
| 65 |
-
"eval_loss": 1.
|
| 66 |
-
"eval_runtime":
|
| 67 |
-
"eval_samples_per_second":
|
| 68 |
-
"eval_steps_per_second":
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch": 0.
|
| 73 |
-
"grad_norm":
|
| 74 |
-
"learning_rate": 9.
|
| 75 |
-
"loss": 1.
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
-
"epoch": 0.
|
| 80 |
-
"eval_loss": 1.
|
| 81 |
-
"eval_runtime":
|
| 82 |
-
"eval_samples_per_second":
|
| 83 |
-
"eval_steps_per_second":
|
| 84 |
"step": 250
|
| 85 |
-
},
|
| 86 |
-
{
|
| 87 |
-
"epoch": 1.0992533026995979,
|
| 88 |
-
"grad_norm": 54.570987701416016,
|
| 89 |
-
"learning_rate": 8.897058823529412e-05,
|
| 90 |
-
"loss": 1.1632,
|
| 91 |
-
"step": 300
|
| 92 |
-
},
|
| 93 |
-
{
|
| 94 |
-
"epoch": 1.0992533026995979,
|
| 95 |
-
"eval_loss": 1.2854100465774536,
|
| 96 |
-
"eval_runtime": 116.058,
|
| 97 |
-
"eval_samples_per_second": 41.583,
|
| 98 |
-
"eval_steps_per_second": 20.791,
|
| 99 |
-
"step": 300
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 1.2830557151062607,
|
| 103 |
-
"grad_norm": 62.69035720825195,
|
| 104 |
-
"learning_rate": 8.713235294117648e-05,
|
| 105 |
-
"loss": 1.1598,
|
| 106 |
-
"step": 350
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 1.2830557151062607,
|
| 110 |
-
"eval_loss": 1.2804261445999146,
|
| 111 |
-
"eval_runtime": 116.1289,
|
| 112 |
-
"eval_samples_per_second": 41.557,
|
| 113 |
-
"eval_steps_per_second": 20.779,
|
| 114 |
-
"step": 350
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 1.4668581275129235,
|
| 118 |
-
"grad_norm": 66.37086486816406,
|
| 119 |
-
"learning_rate": 8.529411764705883e-05,
|
| 120 |
-
"loss": 1.1419,
|
| 121 |
-
"step": 400
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 1.4668581275129235,
|
| 125 |
-
"eval_loss": 1.272080898284912,
|
| 126 |
-
"eval_runtime": 116.1097,
|
| 127 |
-
"eval_samples_per_second": 41.564,
|
| 128 |
-
"eval_steps_per_second": 20.782,
|
| 129 |
-
"step": 400
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 1.6506605399195864,
|
| 133 |
-
"grad_norm": 55.69127655029297,
|
| 134 |
-
"learning_rate": 8.345588235294118e-05,
|
| 135 |
-
"loss": 1.1256,
|
| 136 |
-
"step": 450
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 1.6506605399195864,
|
| 140 |
-
"eval_loss": 1.2628560066223145,
|
| 141 |
-
"eval_runtime": 116.0668,
|
| 142 |
-
"eval_samples_per_second": 41.58,
|
| 143 |
-
"eval_steps_per_second": 20.79,
|
| 144 |
-
"step": 450
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 1.8344629523262492,
|
| 148 |
-
"grad_norm": 84.45687103271484,
|
| 149 |
-
"learning_rate": 8.161764705882353e-05,
|
| 150 |
-
"loss": 1.1115,
|
| 151 |
-
"step": 500
|
| 152 |
-
},
|
| 153 |
-
{
|
| 154 |
-
"epoch": 1.8344629523262492,
|
| 155 |
-
"eval_loss": 1.2613238096237183,
|
| 156 |
-
"eval_runtime": 116.2349,
|
| 157 |
-
"eval_samples_per_second": 41.519,
|
| 158 |
-
"eval_steps_per_second": 20.76,
|
| 159 |
-
"step": 500
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 2.014704192992533,
|
| 163 |
-
"grad_norm": 66.61148834228516,
|
| 164 |
-
"learning_rate": 7.977941176470589e-05,
|
| 165 |
-
"loss": 1.0706,
|
| 166 |
-
"step": 550
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 2.014704192992533,
|
| 170 |
-
"eval_loss": 1.2579104900360107,
|
| 171 |
-
"eval_runtime": 117.3176,
|
| 172 |
-
"eval_samples_per_second": 41.136,
|
| 173 |
-
"eval_steps_per_second": 20.568,
|
| 174 |
-
"step": 550
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 2.1985066053991957,
|
| 178 |
-
"grad_norm": 83.05467987060547,
|
| 179 |
-
"learning_rate": 7.794117647058824e-05,
|
| 180 |
-
"loss": 1.0691,
|
| 181 |
-
"step": 600
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 2.1985066053991957,
|
| 185 |
-
"eval_loss": 1.2582261562347412,
|
| 186 |
-
"eval_runtime": 116.1323,
|
| 187 |
-
"eval_samples_per_second": 41.556,
|
| 188 |
-
"eval_steps_per_second": 20.778,
|
| 189 |
-
"step": 600
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 2.382309017805859,
|
| 193 |
-
"grad_norm": 71.0223617553711,
|
| 194 |
-
"learning_rate": 7.610294117647059e-05,
|
| 195 |
-
"loss": 1.0573,
|
| 196 |
-
"step": 650
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 2.382309017805859,
|
| 200 |
-
"eval_loss": 1.2502797842025757,
|
| 201 |
-
"eval_runtime": 116.1247,
|
| 202 |
-
"eval_samples_per_second": 41.559,
|
| 203 |
-
"eval_steps_per_second": 20.779,
|
| 204 |
-
"step": 650
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 2.5661114302125214,
|
| 208 |
-
"grad_norm": 64.11682891845703,
|
| 209 |
-
"learning_rate": 7.426470588235294e-05,
|
| 210 |
-
"loss": 1.0459,
|
| 211 |
-
"step": 700
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 2.5661114302125214,
|
| 215 |
-
"eval_loss": 1.2464367151260376,
|
| 216 |
-
"eval_runtime": 116.0733,
|
| 217 |
-
"eval_samples_per_second": 41.577,
|
| 218 |
-
"eval_steps_per_second": 20.789,
|
| 219 |
-
"step": 700
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 2.7499138426191845,
|
| 223 |
-
"grad_norm": 90.718505859375,
|
| 224 |
-
"learning_rate": 7.242647058823529e-05,
|
| 225 |
-
"loss": 1.0444,
|
| 226 |
-
"step": 750
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 2.7499138426191845,
|
| 230 |
-
"eval_loss": 1.238742470741272,
|
| 231 |
-
"eval_runtime": 116.1005,
|
| 232 |
-
"eval_samples_per_second": 41.567,
|
| 233 |
-
"eval_steps_per_second": 20.784,
|
| 234 |
-
"step": 750
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 2.933716255025847,
|
| 238 |
-
"grad_norm": 52.93131637573242,
|
| 239 |
-
"learning_rate": 7.058823529411765e-05,
|
| 240 |
-
"loss": 1.033,
|
| 241 |
-
"step": 800
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 2.933716255025847,
|
| 245 |
-
"eval_loss": 1.2367525100708008,
|
| 246 |
-
"eval_runtime": 115.9357,
|
| 247 |
-
"eval_samples_per_second": 41.627,
|
| 248 |
-
"eval_steps_per_second": 20.813,
|
| 249 |
-
"step": 800
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 3.113957495692131,
|
| 253 |
-
"grad_norm": 47.737281799316406,
|
| 254 |
-
"learning_rate": 6.875e-05,
|
| 255 |
-
"loss": 0.9902,
|
| 256 |
-
"step": 850
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 3.113957495692131,
|
| 260 |
-
"eval_loss": 1.2378348112106323,
|
| 261 |
-
"eval_runtime": 115.9587,
|
| 262 |
-
"eval_samples_per_second": 41.618,
|
| 263 |
-
"eval_steps_per_second": 20.809,
|
| 264 |
-
"step": 850
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 3.2977599080987936,
|
| 268 |
-
"grad_norm": 64.8248519897461,
|
| 269 |
-
"learning_rate": 6.691176470588235e-05,
|
| 270 |
-
"loss": 0.9991,
|
| 271 |
-
"step": 900
|
| 272 |
-
},
|
| 273 |
-
{
|
| 274 |
-
"epoch": 3.2977599080987936,
|
| 275 |
-
"eval_loss": 1.2380481958389282,
|
| 276 |
-
"eval_runtime": 116.1261,
|
| 277 |
-
"eval_samples_per_second": 41.558,
|
| 278 |
-
"eval_steps_per_second": 20.779,
|
| 279 |
-
"step": 900
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 3.4815623205054567,
|
| 283 |
-
"grad_norm": 44.85393524169922,
|
| 284 |
-
"learning_rate": 6.507352941176472e-05,
|
| 285 |
-
"loss": 0.9954,
|
| 286 |
-
"step": 950
|
| 287 |
-
},
|
| 288 |
-
{
|
| 289 |
-
"epoch": 3.4815623205054567,
|
| 290 |
-
"eval_loss": 1.2317270040512085,
|
| 291 |
-
"eval_runtime": 116.1731,
|
| 292 |
-
"eval_samples_per_second": 41.541,
|
| 293 |
-
"eval_steps_per_second": 20.771,
|
| 294 |
-
"step": 950
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 3.6653647329121197,
|
| 298 |
-
"grad_norm": 95.18327331542969,
|
| 299 |
-
"learning_rate": 6.323529411764705e-05,
|
| 300 |
-
"loss": 0.9896,
|
| 301 |
-
"step": 1000
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 3.6653647329121197,
|
| 305 |
-
"eval_loss": 1.2345850467681885,
|
| 306 |
-
"eval_runtime": 116.2127,
|
| 307 |
-
"eval_samples_per_second": 41.527,
|
| 308 |
-
"eval_steps_per_second": 20.764,
|
| 309 |
-
"step": 1000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 3.8491671453187823,
|
| 313 |
-
"grad_norm": 53.003849029541016,
|
| 314 |
-
"learning_rate": 6.139705882352942e-05,
|
| 315 |
-
"loss": 0.9786,
|
| 316 |
-
"step": 1050
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"epoch": 3.8491671453187823,
|
| 320 |
-
"eval_loss": 1.235013723373413,
|
| 321 |
-
"eval_runtime": 116.1383,
|
| 322 |
-
"eval_samples_per_second": 41.554,
|
| 323 |
-
"eval_steps_per_second": 20.777,
|
| 324 |
-
"step": 1050
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 4.029408385985066,
|
| 328 |
-
"grad_norm": 60.95920181274414,
|
| 329 |
-
"learning_rate": 5.9558823529411766e-05,
|
| 330 |
-
"loss": 0.953,
|
| 331 |
-
"step": 1100
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 4.029408385985066,
|
| 335 |
-
"eval_loss": 1.2303814888000488,
|
| 336 |
-
"eval_runtime": 116.9283,
|
| 337 |
-
"eval_samples_per_second": 41.273,
|
| 338 |
-
"eval_steps_per_second": 20.637,
|
| 339 |
-
"step": 1100
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 4.213210798391729,
|
| 343 |
-
"grad_norm": 94.7210922241211,
|
| 344 |
-
"learning_rate": 5.7720588235294116e-05,
|
| 345 |
-
"loss": 0.9526,
|
| 346 |
-
"step": 1150
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 4.213210798391729,
|
| 350 |
-
"eval_loss": 1.2337546348571777,
|
| 351 |
-
"eval_runtime": 116.1668,
|
| 352 |
-
"eval_samples_per_second": 41.544,
|
| 353 |
-
"eval_steps_per_second": 20.772,
|
| 354 |
-
"step": 1150
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 4.3970132107983915,
|
| 358 |
-
"grad_norm": 42.170989990234375,
|
| 359 |
-
"learning_rate": 5.588235294117647e-05,
|
| 360 |
-
"loss": 0.9448,
|
| 361 |
-
"step": 1200
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 4.3970132107983915,
|
| 365 |
-
"eval_loss": 1.228541612625122,
|
| 366 |
-
"eval_runtime": 116.2415,
|
| 367 |
-
"eval_samples_per_second": 41.517,
|
| 368 |
-
"eval_steps_per_second": 20.759,
|
| 369 |
-
"step": 1200
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 4.580815623205055,
|
| 373 |
-
"grad_norm": 60.85408401489258,
|
| 374 |
-
"learning_rate": 5.404411764705882e-05,
|
| 375 |
-
"loss": 0.9408,
|
| 376 |
-
"step": 1250
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 4.580815623205055,
|
| 380 |
-
"eval_loss": 1.2270257472991943,
|
| 381 |
-
"eval_runtime": 116.2096,
|
| 382 |
-
"eval_samples_per_second": 41.528,
|
| 383 |
-
"eval_steps_per_second": 20.764,
|
| 384 |
-
"step": 1250
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 4.764618035611718,
|
| 388 |
-
"grad_norm": 59.16058349609375,
|
| 389 |
-
"learning_rate": 5.2205882352941185e-05,
|
| 390 |
-
"loss": 0.9364,
|
| 391 |
-
"step": 1300
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 4.764618035611718,
|
| 395 |
-
"eval_loss": 1.2274161577224731,
|
| 396 |
-
"eval_runtime": 116.2519,
|
| 397 |
-
"eval_samples_per_second": 41.513,
|
| 398 |
-
"eval_steps_per_second": 20.757,
|
| 399 |
-
"step": 1300
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 4.94842044801838,
|
| 403 |
-
"grad_norm": 94.78569030761719,
|
| 404 |
-
"learning_rate": 5.036764705882353e-05,
|
| 405 |
-
"loss": 0.9319,
|
| 406 |
-
"step": 1350
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 4.94842044801838,
|
| 410 |
-
"eval_loss": 1.228610873222351,
|
| 411 |
-
"eval_runtime": 116.0677,
|
| 412 |
-
"eval_samples_per_second": 41.579,
|
| 413 |
-
"eval_steps_per_second": 20.79,
|
| 414 |
-
"step": 1350
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 5.128661688684664,
|
| 418 |
-
"grad_norm": 56.210121154785156,
|
| 419 |
-
"learning_rate": 4.8529411764705885e-05,
|
| 420 |
-
"loss": 0.9023,
|
| 421 |
-
"step": 1400
|
| 422 |
-
},
|
| 423 |
-
{
|
| 424 |
-
"epoch": 5.128661688684664,
|
| 425 |
-
"eval_loss": 1.2338696718215942,
|
| 426 |
-
"eval_runtime": 116.0945,
|
| 427 |
-
"eval_samples_per_second": 41.57,
|
| 428 |
-
"eval_steps_per_second": 20.785,
|
| 429 |
-
"step": 1400
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 5.312464101091327,
|
| 433 |
-
"grad_norm": 54.96003723144531,
|
| 434 |
-
"learning_rate": 4.669117647058824e-05,
|
| 435 |
-
"loss": 0.9067,
|
| 436 |
-
"step": 1450
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 5.312464101091327,
|
| 440 |
-
"eval_loss": 1.231054663658142,
|
| 441 |
-
"eval_runtime": 115.9656,
|
| 442 |
-
"eval_samples_per_second": 41.616,
|
| 443 |
-
"eval_steps_per_second": 20.808,
|
| 444 |
-
"step": 1450
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 5.496266513497989,
|
| 448 |
-
"grad_norm": 45.37810516357422,
|
| 449 |
-
"learning_rate": 4.485294117647059e-05,
|
| 450 |
-
"loss": 0.9032,
|
| 451 |
-
"step": 1500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 5.496266513497989,
|
| 455 |
-
"eval_loss": 1.2278664112091064,
|
| 456 |
-
"eval_runtime": 116.0475,
|
| 457 |
-
"eval_samples_per_second": 41.586,
|
| 458 |
-
"eval_steps_per_second": 20.793,
|
| 459 |
-
"step": 1500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 5.680068925904653,
|
| 463 |
-
"grad_norm": 51.236480712890625,
|
| 464 |
-
"learning_rate": 4.301470588235295e-05,
|
| 465 |
-
"loss": 0.9032,
|
| 466 |
-
"step": 1550
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 5.680068925904653,
|
| 470 |
-
"eval_loss": 1.2236429452896118,
|
| 471 |
-
"eval_runtime": 116.3342,
|
| 472 |
-
"eval_samples_per_second": 41.484,
|
| 473 |
-
"eval_steps_per_second": 20.742,
|
| 474 |
-
"step": 1550
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 5.8638713383113155,
|
| 478 |
-
"grad_norm": 40.550933837890625,
|
| 479 |
-
"learning_rate": 4.11764705882353e-05,
|
| 480 |
-
"loss": 0.896,
|
| 481 |
-
"step": 1600
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 5.8638713383113155,
|
| 485 |
-
"eval_loss": 1.2268708944320679,
|
| 486 |
-
"eval_runtime": 116.1038,
|
| 487 |
-
"eval_samples_per_second": 41.566,
|
| 488 |
-
"eval_steps_per_second": 20.783,
|
| 489 |
-
"step": 1600
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 6.044112578977599,
|
| 493 |
-
"grad_norm": 59.42768859863281,
|
| 494 |
-
"learning_rate": 3.933823529411765e-05,
|
| 495 |
-
"loss": 0.8781,
|
| 496 |
-
"step": 1650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 6.044112578977599,
|
| 500 |
-
"eval_loss": 1.233597993850708,
|
| 501 |
-
"eval_runtime": 117.1571,
|
| 502 |
-
"eval_samples_per_second": 41.193,
|
| 503 |
-
"eval_steps_per_second": 20.596,
|
| 504 |
-
"step": 1650
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 6.227914991384262,
|
| 508 |
-
"grad_norm": 68.26610565185547,
|
| 509 |
-
"learning_rate": 3.7500000000000003e-05,
|
| 510 |
-
"loss": 0.8804,
|
| 511 |
-
"step": 1700
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 6.227914991384262,
|
| 515 |
-
"eval_loss": 1.2279460430145264,
|
| 516 |
-
"eval_runtime": 116.1011,
|
| 517 |
-
"eval_samples_per_second": 41.567,
|
| 518 |
-
"eval_steps_per_second": 20.784,
|
| 519 |
-
"step": 1700
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 6.411717403790925,
|
| 523 |
-
"grad_norm": 77.21823120117188,
|
| 524 |
-
"learning_rate": 3.566176470588235e-05,
|
| 525 |
-
"loss": 0.8733,
|
| 526 |
-
"step": 1750
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 6.411717403790925,
|
| 530 |
-
"eval_loss": 1.2353451251983643,
|
| 531 |
-
"eval_runtime": 116.0518,
|
| 532 |
-
"eval_samples_per_second": 41.585,
|
| 533 |
-
"eval_steps_per_second": 20.792,
|
| 534 |
-
"step": 1750
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 6.595519816197587,
|
| 538 |
-
"grad_norm": 49.22051239013672,
|
| 539 |
-
"learning_rate": 3.382352941176471e-05,
|
| 540 |
-
"loss": 0.875,
|
| 541 |
-
"step": 1800
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 6.595519816197587,
|
| 545 |
-
"eval_loss": 1.2324572801589966,
|
| 546 |
-
"eval_runtime": 116.0982,
|
| 547 |
-
"eval_samples_per_second": 41.568,
|
| 548 |
-
"eval_steps_per_second": 20.784,
|
| 549 |
-
"step": 1800
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 6.779322228604251,
|
| 553 |
-
"grad_norm": 61.27114486694336,
|
| 554 |
-
"learning_rate": 3.198529411764706e-05,
|
| 555 |
-
"loss": 0.8634,
|
| 556 |
-
"step": 1850
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 6.779322228604251,
|
| 560 |
-
"eval_loss": 1.2263100147247314,
|
| 561 |
-
"eval_runtime": 116.0582,
|
| 562 |
-
"eval_samples_per_second": 41.583,
|
| 563 |
-
"eval_steps_per_second": 20.791,
|
| 564 |
-
"step": 1850
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 6.963124641010913,
|
| 568 |
-
"grad_norm": 53.27342224121094,
|
| 569 |
-
"learning_rate": 3.0147058823529413e-05,
|
| 570 |
-
"loss": 0.8647,
|
| 571 |
-
"step": 1900
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 6.963124641010913,
|
| 575 |
-
"eval_loss": 1.2306259870529175,
|
| 576 |
-
"eval_runtime": 116.21,
|
| 577 |
-
"eval_samples_per_second": 41.528,
|
| 578 |
-
"eval_steps_per_second": 20.764,
|
| 579 |
-
"step": 1900
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 7.143365881677197,
|
| 583 |
-
"grad_norm": 56.99700927734375,
|
| 584 |
-
"learning_rate": 2.8308823529411766e-05,
|
| 585 |
-
"loss": 0.8335,
|
| 586 |
-
"step": 1950
|
| 587 |
-
},
|
| 588 |
-
{
|
| 589 |
-
"epoch": 7.143365881677197,
|
| 590 |
-
"eval_loss": 1.2323832511901855,
|
| 591 |
-
"eval_runtime": 116.0282,
|
| 592 |
-
"eval_samples_per_second": 41.593,
|
| 593 |
-
"eval_steps_per_second": 20.797,
|
| 594 |
-
"step": 1950
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 7.32716829408386,
|
| 598 |
-
"grad_norm": 111.48177337646484,
|
| 599 |
-
"learning_rate": 2.647058823529412e-05,
|
| 600 |
-
"loss": 0.8489,
|
| 601 |
-
"step": 2000
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 7.32716829408386,
|
| 605 |
-
"eval_loss": 1.2314597368240356,
|
| 606 |
-
"eval_runtime": 116.1391,
|
| 607 |
-
"eval_samples_per_second": 41.554,
|
| 608 |
-
"eval_steps_per_second": 20.777,
|
| 609 |
-
"step": 2000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 7.5109707064905225,
|
| 613 |
-
"grad_norm": 44.07224655151367,
|
| 614 |
-
"learning_rate": 2.4632352941176472e-05,
|
| 615 |
-
"loss": 0.8473,
|
| 616 |
-
"step": 2050
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 7.5109707064905225,
|
| 620 |
-
"eval_loss": 1.2360129356384277,
|
| 621 |
-
"eval_runtime": 116.3906,
|
| 622 |
-
"eval_samples_per_second": 41.464,
|
| 623 |
-
"eval_steps_per_second": 20.732,
|
| 624 |
-
"step": 2050
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 7.694773118897185,
|
| 628 |
-
"grad_norm": 58.74856948852539,
|
| 629 |
-
"learning_rate": 2.2794117647058825e-05,
|
| 630 |
-
"loss": 0.8422,
|
| 631 |
-
"step": 2100
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 7.694773118897185,
|
| 635 |
-
"eval_loss": 1.23045015335083,
|
| 636 |
-
"eval_runtime": 116.4238,
|
| 637 |
-
"eval_samples_per_second": 41.452,
|
| 638 |
-
"eval_steps_per_second": 20.726,
|
| 639 |
-
"step": 2100
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 7.878575531303849,
|
| 643 |
-
"grad_norm": 42.6165771484375,
|
| 644 |
-
"learning_rate": 2.0955882352941178e-05,
|
| 645 |
-
"loss": 0.8414,
|
| 646 |
-
"step": 2150
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 7.878575531303849,
|
| 650 |
-
"eval_loss": 1.232067346572876,
|
| 651 |
-
"eval_runtime": 115.9906,
|
| 652 |
-
"eval_samples_per_second": 41.607,
|
| 653 |
-
"eval_steps_per_second": 20.803,
|
| 654 |
-
"step": 2150
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 8.058816771970132,
|
| 658 |
-
"grad_norm": 80.5433349609375,
|
| 659 |
-
"learning_rate": 1.9117647058823528e-05,
|
| 660 |
-
"loss": 0.8227,
|
| 661 |
-
"step": 2200
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 8.058816771970132,
|
| 665 |
-
"eval_loss": 1.2389429807662964,
|
| 666 |
-
"eval_runtime": 117.23,
|
| 667 |
-
"eval_samples_per_second": 41.167,
|
| 668 |
-
"eval_steps_per_second": 20.583,
|
| 669 |
-
"step": 2200
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 8.242619184376794,
|
| 673 |
-
"grad_norm": 45.97893142700195,
|
| 674 |
-
"learning_rate": 1.7279411764705884e-05,
|
| 675 |
-
"loss": 0.8316,
|
| 676 |
-
"step": 2250
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 8.242619184376794,
|
| 680 |
-
"eval_loss": 1.2351105213165283,
|
| 681 |
-
"eval_runtime": 116.2345,
|
| 682 |
-
"eval_samples_per_second": 41.52,
|
| 683 |
-
"eval_steps_per_second": 20.76,
|
| 684 |
-
"step": 2250
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 8.426421596783458,
|
| 688 |
-
"grad_norm": 68.8030014038086,
|
| 689 |
-
"learning_rate": 1.5441176470588237e-05,
|
| 690 |
-
"loss": 0.8299,
|
| 691 |
-
"step": 2300
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 8.426421596783458,
|
| 695 |
-
"eval_loss": 1.2383313179016113,
|
| 696 |
-
"eval_runtime": 115.9203,
|
| 697 |
-
"eval_samples_per_second": 41.632,
|
| 698 |
-
"eval_steps_per_second": 20.816,
|
| 699 |
-
"step": 2300
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 8.610224009190121,
|
| 703 |
-
"grad_norm": 48.16875076293945,
|
| 704 |
-
"learning_rate": 1.3602941176470587e-05,
|
| 705 |
-
"loss": 0.822,
|
| 706 |
-
"step": 2350
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 8.610224009190121,
|
| 710 |
-
"eval_loss": 1.2370705604553223,
|
| 711 |
-
"eval_runtime": 116.2999,
|
| 712 |
-
"eval_samples_per_second": 41.496,
|
| 713 |
-
"eval_steps_per_second": 20.748,
|
| 714 |
-
"step": 2350
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 8.794026421596783,
|
| 718 |
-
"grad_norm": 49.53213119506836,
|
| 719 |
-
"learning_rate": 1.1764705882352942e-05,
|
| 720 |
-
"loss": 0.8251,
|
| 721 |
-
"step": 2400
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 8.794026421596783,
|
| 725 |
-
"eval_loss": 1.2367668151855469,
|
| 726 |
-
"eval_runtime": 116.3102,
|
| 727 |
-
"eval_samples_per_second": 41.493,
|
| 728 |
-
"eval_steps_per_second": 20.746,
|
| 729 |
-
"step": 2400
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 8.977828834003446,
|
| 733 |
-
"grad_norm": 42.136714935302734,
|
| 734 |
-
"learning_rate": 9.926470588235293e-06,
|
| 735 |
-
"loss": 0.8225,
|
| 736 |
-
"step": 2450
|
| 737 |
-
},
|
| 738 |
-
{
|
| 739 |
-
"epoch": 8.977828834003446,
|
| 740 |
-
"eval_loss": 1.2319527864456177,
|
| 741 |
-
"eval_runtime": 116.086,
|
| 742 |
-
"eval_samples_per_second": 41.573,
|
| 743 |
-
"eval_steps_per_second": 20.786,
|
| 744 |
-
"step": 2450
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 9.15807007466973,
|
| 748 |
-
"grad_norm": 55.69232940673828,
|
| 749 |
-
"learning_rate": 8.088235294117648e-06,
|
| 750 |
-
"loss": 0.7995,
|
| 751 |
-
"step": 2500
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 9.15807007466973,
|
| 755 |
-
"eval_loss": 1.2360199689865112,
|
| 756 |
-
"eval_runtime": 116.0847,
|
| 757 |
-
"eval_samples_per_second": 41.573,
|
| 758 |
-
"eval_steps_per_second": 20.787,
|
| 759 |
-
"step": 2500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 9.341872487076392,
|
| 763 |
-
"grad_norm": 62.24937438964844,
|
| 764 |
-
"learning_rate": 6.25e-06,
|
| 765 |
-
"loss": 0.8149,
|
| 766 |
-
"step": 2550
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 9.341872487076392,
|
| 770 |
-
"eval_loss": 1.2363650798797607,
|
| 771 |
-
"eval_runtime": 116.0508,
|
| 772 |
-
"eval_samples_per_second": 41.585,
|
| 773 |
-
"eval_steps_per_second": 20.793,
|
| 774 |
-
"step": 2550
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 9.525674899483056,
|
| 778 |
-
"grad_norm": 50.01460266113281,
|
| 779 |
-
"learning_rate": 4.411764705882353e-06,
|
| 780 |
-
"loss": 0.8146,
|
| 781 |
-
"step": 2600
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 9.525674899483056,
|
| 785 |
-
"eval_loss": 1.2402119636535645,
|
| 786 |
-
"eval_runtime": 116.1442,
|
| 787 |
-
"eval_samples_per_second": 41.552,
|
| 788 |
-
"eval_steps_per_second": 20.776,
|
| 789 |
-
"step": 2600
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 9.709477311889719,
|
| 793 |
-
"grad_norm": 60.61581802368164,
|
| 794 |
-
"learning_rate": 2.573529411764706e-06,
|
| 795 |
-
"loss": 0.8075,
|
| 796 |
-
"step": 2650
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 9.709477311889719,
|
| 800 |
-
"eval_loss": 1.2341493368148804,
|
| 801 |
-
"eval_runtime": 116.3546,
|
| 802 |
-
"eval_samples_per_second": 41.477,
|
| 803 |
-
"eval_steps_per_second": 20.738,
|
| 804 |
-
"step": 2650
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 9.89327972429638,
|
| 808 |
-
"grad_norm": 54.73764419555664,
|
| 809 |
-
"learning_rate": 7.352941176470589e-07,
|
| 810 |
-
"loss": 0.8108,
|
| 811 |
-
"step": 2700
|
| 812 |
-
},
|
| 813 |
-
{
|
| 814 |
-
"epoch": 9.89327972429638,
|
| 815 |
-
"eval_loss": 1.2355538606643677,
|
| 816 |
-
"eval_runtime": 116.1282,
|
| 817 |
-
"eval_samples_per_second": 41.558,
|
| 818 |
-
"eval_steps_per_second": 20.779,
|
| 819 |
-
"step": 2700
|
| 820 |
}
|
| 821 |
],
|
| 822 |
"logging_steps": 50,
|
| 823 |
-
"max_steps":
|
| 824 |
"num_input_tokens_seen": 0,
|
| 825 |
"num_train_epochs": 10,
|
| 826 |
"save_steps": 500,
|
|
@@ -831,12 +96,12 @@
|
|
| 831 |
"should_evaluate": false,
|
| 832 |
"should_log": false,
|
| 833 |
"should_save": true,
|
| 834 |
-
"should_training_stop":
|
| 835 |
},
|
| 836 |
"attributes": {}
|
| 837 |
}
|
| 838 |
},
|
| 839 |
-
"total_flos":
|
| 840 |
"train_batch_size": 2,
|
| 841 |
"trial_name": null,
|
| 842 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9972065305108946,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 251,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.1986467192252778,
|
| 13 |
+
"grad_norm": 1.6958907842636108,
|
| 14 |
+
"learning_rate": 9.800796812749005e-05,
|
| 15 |
+
"loss": 1.6362,
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.1986467192252778,
|
| 20 |
+
"eval_loss": 1.5567175149917603,
|
| 21 |
+
"eval_runtime": 14.8041,
|
| 22 |
+
"eval_samples_per_second": 57.281,
|
| 23 |
+
"eval_steps_per_second": 28.641,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.3972934384505556,
|
| 28 |
+
"grad_norm": 1.6160608530044556,
|
| 29 |
+
"learning_rate": 9.601593625498009e-05,
|
| 30 |
+
"loss": 1.5283,
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.3972934384505556,
|
| 35 |
+
"eval_loss": 1.5002830028533936,
|
| 36 |
+
"eval_runtime": 14.7266,
|
| 37 |
+
"eval_samples_per_second": 57.583,
|
| 38 |
+
"eval_steps_per_second": 28.792,
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.5959401576758334,
|
| 43 |
+
"grad_norm": 1.500954270362854,
|
| 44 |
+
"learning_rate": 9.402390438247013e-05,
|
| 45 |
+
"loss": 1.4825,
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
+
"epoch": 0.5959401576758334,
|
| 50 |
+
"eval_loss": 1.4542008638381958,
|
| 51 |
+
"eval_runtime": 14.5669,
|
| 52 |
+
"eval_samples_per_second": 58.214,
|
| 53 |
+
"eval_steps_per_second": 29.107,
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
+
"epoch": 0.7945868769011112,
|
| 58 |
+
"grad_norm": 0.8321912884712219,
|
| 59 |
+
"learning_rate": 9.203187250996016e-05,
|
| 60 |
+
"loss": 1.4431,
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 0.7945868769011112,
|
| 65 |
+
"eval_loss": 1.4306951761245728,
|
| 66 |
+
"eval_runtime": 14.563,
|
| 67 |
+
"eval_samples_per_second": 58.23,
|
| 68 |
+
"eval_steps_per_second": 29.115,
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.993233596126389,
|
| 73 |
+
"grad_norm": 1.2672511339187622,
|
| 74 |
+
"learning_rate": 9.00398406374502e-05,
|
| 75 |
+
"loss": 1.4083,
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
+
"epoch": 0.993233596126389,
|
| 80 |
+
"eval_loss": 1.3854182958602905,
|
| 81 |
+
"eval_runtime": 14.6247,
|
| 82 |
+
"eval_samples_per_second": 57.984,
|
| 83 |
+
"eval_steps_per_second": 28.992,
|
| 84 |
"step": 250
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 85 |
}
|
| 86 |
],
|
| 87 |
"logging_steps": 50,
|
| 88 |
+
"max_steps": 2510,
|
| 89 |
"num_input_tokens_seen": 0,
|
| 90 |
"num_train_epochs": 10,
|
| 91 |
"save_steps": 500,
|
|
|
|
| 96 |
"should_evaluate": false,
|
| 97 |
"should_log": false,
|
| 98 |
"should_save": true,
|
| 99 |
+
"should_training_stop": false
|
| 100 |
},
|
| 101 |
"attributes": {}
|
| 102 |
}
|
| 103 |
},
|
| 104 |
+
"total_flos": 4.004589805030605e+16,
|
| 105 |
"train_batch_size": 2,
|
| 106 |
"trial_name": null,
|
| 107 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5368
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d60cd4fa81843b4806dea3364d37d3df9835095733d168d051c0b135e77b91aa
|
| 3 |
size 5368
|