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": "4.47.0",
|
| 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 sha256:359f03565ffde72ca2ba79491c705fe79c6910e9bd98dc4da3c2d72dd984114c
|
| 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 sha256:9048fafeb2aa676460e8ea091a11b1525206d19f4ef4961093c98af204315634
|
| 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:eb357d1fd873b2e9e783702e673b2d6f1de3e1b1f85efe7ccee99b69a6a7abc9
|
| 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 sha256:2fec4fd6bb533b427a0a9db7b7fd25ac543921c09d5d0a518d007b072a567e94
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,828 +1,108 @@
|
|
| 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": 116.
|
| 22 |
-
"eval_samples_per_second": 41.
|
| 23 |
-
"eval_steps_per_second": 20.
|
| 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": 41.
|
| 38 |
-
"eval_steps_per_second": 20.
|
| 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": 41.
|
| 53 |
-
"eval_steps_per_second": 20.
|
| 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": 116.
|
| 67 |
-
"eval_samples_per_second": 41.
|
| 68 |
-
"eval_steps_per_second": 20.
|
| 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": 41.
|
| 83 |
-
"eval_steps_per_second": 20.
|
| 84 |
"step": 250
|
| 85 |
},
|
| 86 |
{
|
| 87 |
-
"epoch":
|
| 88 |
-
"grad_norm":
|
| 89 |
-
"learning_rate":
|
| 90 |
-
"loss": 1.
|
| 91 |
"step": 300
|
| 92 |
},
|
| 93 |
{
|
| 94 |
-
"epoch":
|
| 95 |
-
"eval_loss": 1.
|
| 96 |
-
"eval_runtime":
|
| 97 |
-
"eval_samples_per_second": 41.
|
| 98 |
-
"eval_steps_per_second": 20.
|
| 99 |
"step": 300
|
| 100 |
-
},
|
| 101 |
-
{
|
| 102 |
-
"epoch": 1.2866168868466399,
|
| 103 |
-
"grad_norm": 104.54215240478516,
|
| 104 |
-
"learning_rate": 8.713235294117648e-05,
|
| 105 |
-
"loss": 1.1591,
|
| 106 |
-
"step": 350
|
| 107 |
-
},
|
| 108 |
-
{
|
| 109 |
-
"epoch": 1.2866168868466399,
|
| 110 |
-
"eval_loss": 1.274283766746521,
|
| 111 |
-
"eval_runtime": 116.0184,
|
| 112 |
-
"eval_samples_per_second": 41.597,
|
| 113 |
-
"eval_steps_per_second": 20.798,
|
| 114 |
-
"step": 350
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 1.4704192992533027,
|
| 118 |
-
"grad_norm": 85.3061294555664,
|
| 119 |
-
"learning_rate": 8.529411764705883e-05,
|
| 120 |
-
"loss": 1.1408,
|
| 121 |
-
"step": 400
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 1.4704192992533027,
|
| 125 |
-
"eval_loss": 1.2692538499832153,
|
| 126 |
-
"eval_runtime": 116.0932,
|
| 127 |
-
"eval_samples_per_second": 41.57,
|
| 128 |
-
"eval_steps_per_second": 20.785,
|
| 129 |
-
"step": 400
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 1.6542217116599656,
|
| 133 |
-
"grad_norm": 72.23489379882812,
|
| 134 |
-
"learning_rate": 8.345588235294118e-05,
|
| 135 |
-
"loss": 1.1256,
|
| 136 |
-
"step": 450
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 1.6542217116599656,
|
| 140 |
-
"eval_loss": 1.2617342472076416,
|
| 141 |
-
"eval_runtime": 116.0634,
|
| 142 |
-
"eval_samples_per_second": 41.581,
|
| 143 |
-
"eval_steps_per_second": 20.79,
|
| 144 |
-
"step": 450
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 1.8380241240666284,
|
| 148 |
-
"grad_norm": 65.63114929199219,
|
| 149 |
-
"learning_rate": 8.161764705882353e-05,
|
| 150 |
-
"loss": 1.1098,
|
| 151 |
-
"step": 500
|
| 152 |
-
},
|
| 153 |
-
{
|
| 154 |
-
"epoch": 1.8380241240666284,
|
| 155 |
-
"eval_loss": 1.2621153593063354,
|
| 156 |
-
"eval_runtime": 115.9613,
|
| 157 |
-
"eval_samples_per_second": 41.617,
|
| 158 |
-
"eval_steps_per_second": 20.809,
|
| 159 |
-
"step": 500
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 2.021826536473291,
|
| 163 |
-
"grad_norm": 45.61464309692383,
|
| 164 |
-
"learning_rate": 7.977941176470589e-05,
|
| 165 |
-
"loss": 1.0919,
|
| 166 |
-
"step": 550
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 2.021826536473291,
|
| 170 |
-
"eval_loss": 1.2552005052566528,
|
| 171 |
-
"eval_runtime": 117.239,
|
| 172 |
-
"eval_samples_per_second": 41.164,
|
| 173 |
-
"eval_steps_per_second": 20.582,
|
| 174 |
-
"step": 550
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 2.205628948879954,
|
| 178 |
-
"grad_norm": 59.019195556640625,
|
| 179 |
-
"learning_rate": 7.794117647058824e-05,
|
| 180 |
-
"loss": 1.0693,
|
| 181 |
-
"step": 600
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 2.205628948879954,
|
| 185 |
-
"eval_loss": 1.2541062831878662,
|
| 186 |
-
"eval_runtime": 115.9595,
|
| 187 |
-
"eval_samples_per_second": 41.618,
|
| 188 |
-
"eval_steps_per_second": 20.809,
|
| 189 |
-
"step": 600
|
| 190 |
-
},
|
| 191 |
-
{
|
| 192 |
-
"epoch": 2.3894313612866167,
|
| 193 |
-
"grad_norm": 43.13253402709961,
|
| 194 |
-
"learning_rate": 7.610294117647059e-05,
|
| 195 |
-
"loss": 1.0562,
|
| 196 |
-
"step": 650
|
| 197 |
-
},
|
| 198 |
-
{
|
| 199 |
-
"epoch": 2.3894313612866167,
|
| 200 |
-
"eval_loss": 1.2531682252883911,
|
| 201 |
-
"eval_runtime": 116.101,
|
| 202 |
-
"eval_samples_per_second": 41.567,
|
| 203 |
-
"eval_steps_per_second": 20.784,
|
| 204 |
-
"step": 650
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 2.5732337736932798,
|
| 208 |
-
"grad_norm": 59.197181701660156,
|
| 209 |
-
"learning_rate": 7.426470588235294e-05,
|
| 210 |
-
"loss": 1.048,
|
| 211 |
-
"step": 700
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 2.5732337736932798,
|
| 215 |
-
"eval_loss": 1.2469751834869385,
|
| 216 |
-
"eval_runtime": 115.9924,
|
| 217 |
-
"eval_samples_per_second": 41.606,
|
| 218 |
-
"eval_steps_per_second": 20.803,
|
| 219 |
-
"step": 700
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 2.757036186099943,
|
| 223 |
-
"grad_norm": 75.58405303955078,
|
| 224 |
-
"learning_rate": 7.242647058823529e-05,
|
| 225 |
-
"loss": 1.0436,
|
| 226 |
-
"step": 750
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 2.757036186099943,
|
| 230 |
-
"eval_loss": 1.2442747354507446,
|
| 231 |
-
"eval_runtime": 116.1229,
|
| 232 |
-
"eval_samples_per_second": 41.559,
|
| 233 |
-
"eval_steps_per_second": 20.78,
|
| 234 |
-
"step": 750
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 2.9408385985066055,
|
| 238 |
-
"grad_norm": 80.94259643554688,
|
| 239 |
-
"learning_rate": 7.058823529411765e-05,
|
| 240 |
-
"loss": 1.0327,
|
| 241 |
-
"step": 800
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 2.9408385985066055,
|
| 245 |
-
"eval_loss": 1.2350915670394897,
|
| 246 |
-
"eval_runtime": 116.1113,
|
| 247 |
-
"eval_samples_per_second": 41.564,
|
| 248 |
-
"eval_steps_per_second": 20.782,
|
| 249 |
-
"step": 800
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 3.124641010913268,
|
| 253 |
-
"grad_norm": 48.83946990966797,
|
| 254 |
-
"learning_rate": 6.875e-05,
|
| 255 |
-
"loss": 1.0081,
|
| 256 |
-
"step": 850
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 3.124641010913268,
|
| 260 |
-
"eval_loss": 1.2360025644302368,
|
| 261 |
-
"eval_runtime": 116.1076,
|
| 262 |
-
"eval_samples_per_second": 41.565,
|
| 263 |
-
"eval_steps_per_second": 20.782,
|
| 264 |
-
"step": 850
|
| 265 |
-
},
|
| 266 |
-
{
|
| 267 |
-
"epoch": 3.308443423319931,
|
| 268 |
-
"grad_norm": 59.597232818603516,
|
| 269 |
-
"learning_rate": 6.691176470588235e-05,
|
| 270 |
-
"loss": 0.9992,
|
| 271 |
-
"step": 900
|
| 272 |
-
},
|
| 273 |
-
{
|
| 274 |
-
"epoch": 3.308443423319931,
|
| 275 |
-
"eval_loss": 1.2394022941589355,
|
| 276 |
-
"eval_runtime": 115.8965,
|
| 277 |
-
"eval_samples_per_second": 41.641,
|
| 278 |
-
"eval_steps_per_second": 20.82,
|
| 279 |
-
"step": 900
|
| 280 |
-
},
|
| 281 |
-
{
|
| 282 |
-
"epoch": 3.4922458357265938,
|
| 283 |
-
"grad_norm": 49.224666595458984,
|
| 284 |
-
"learning_rate": 6.507352941176472e-05,
|
| 285 |
-
"loss": 0.9943,
|
| 286 |
-
"step": 950
|
| 287 |
-
},
|
| 288 |
-
{
|
| 289 |
-
"epoch": 3.4922458357265938,
|
| 290 |
-
"eval_loss": 1.235238790512085,
|
| 291 |
-
"eval_runtime": 116.2683,
|
| 292 |
-
"eval_samples_per_second": 41.507,
|
| 293 |
-
"eval_steps_per_second": 20.754,
|
| 294 |
-
"step": 950
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 3.676048248133257,
|
| 298 |
-
"grad_norm": 65.07023620605469,
|
| 299 |
-
"learning_rate": 6.323529411764705e-05,
|
| 300 |
-
"loss": 0.9868,
|
| 301 |
-
"step": 1000
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 3.676048248133257,
|
| 305 |
-
"eval_loss": 1.23443603515625,
|
| 306 |
-
"eval_runtime": 116.2562,
|
| 307 |
-
"eval_samples_per_second": 41.512,
|
| 308 |
-
"eval_steps_per_second": 20.756,
|
| 309 |
-
"step": 1000
|
| 310 |
-
},
|
| 311 |
-
{
|
| 312 |
-
"epoch": 3.8598506605399194,
|
| 313 |
-
"grad_norm": 49.963409423828125,
|
| 314 |
-
"learning_rate": 6.139705882352942e-05,
|
| 315 |
-
"loss": 0.9781,
|
| 316 |
-
"step": 1050
|
| 317 |
-
},
|
| 318 |
-
{
|
| 319 |
-
"epoch": 3.8598506605399194,
|
| 320 |
-
"eval_loss": 1.232906460762024,
|
| 321 |
-
"eval_runtime": 116.0309,
|
| 322 |
-
"eval_samples_per_second": 41.592,
|
| 323 |
-
"eval_steps_per_second": 20.796,
|
| 324 |
-
"step": 1050
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 4.043653072946582,
|
| 328 |
-
"grad_norm": 57.1251335144043,
|
| 329 |
-
"learning_rate": 5.9558823529411766e-05,
|
| 330 |
-
"loss": 0.9697,
|
| 331 |
-
"step": 1100
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 4.043653072946582,
|
| 335 |
-
"eval_loss": 1.2329473495483398,
|
| 336 |
-
"eval_runtime": 117.8798,
|
| 337 |
-
"eval_samples_per_second": 40.94,
|
| 338 |
-
"eval_steps_per_second": 20.47,
|
| 339 |
-
"step": 1100
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 4.227455485353246,
|
| 343 |
-
"grad_norm": 47.962928771972656,
|
| 344 |
-
"learning_rate": 5.7720588235294116e-05,
|
| 345 |
-
"loss": 0.9499,
|
| 346 |
-
"step": 1150
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 4.227455485353246,
|
| 350 |
-
"eval_loss": 1.2345472574234009,
|
| 351 |
-
"eval_runtime": 115.8584,
|
| 352 |
-
"eval_samples_per_second": 41.654,
|
| 353 |
-
"eval_steps_per_second": 20.827,
|
| 354 |
-
"step": 1150
|
| 355 |
-
},
|
| 356 |
-
{
|
| 357 |
-
"epoch": 4.411257897759908,
|
| 358 |
-
"grad_norm": 43.172767639160156,
|
| 359 |
-
"learning_rate": 5.588235294117647e-05,
|
| 360 |
-
"loss": 0.9447,
|
| 361 |
-
"step": 1200
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 4.411257897759908,
|
| 365 |
-
"eval_loss": 1.2309461832046509,
|
| 366 |
-
"eval_runtime": 116.0943,
|
| 367 |
-
"eval_samples_per_second": 41.57,
|
| 368 |
-
"eval_steps_per_second": 20.785,
|
| 369 |
-
"step": 1200
|
| 370 |
-
},
|
| 371 |
-
{
|
| 372 |
-
"epoch": 4.595060310166571,
|
| 373 |
-
"grad_norm": 89.4483413696289,
|
| 374 |
-
"learning_rate": 5.404411764705882e-05,
|
| 375 |
-
"loss": 0.9384,
|
| 376 |
-
"step": 1250
|
| 377 |
-
},
|
| 378 |
-
{
|
| 379 |
-
"epoch": 4.595060310166571,
|
| 380 |
-
"eval_loss": 1.2274333238601685,
|
| 381 |
-
"eval_runtime": 116.0025,
|
| 382 |
-
"eval_samples_per_second": 41.603,
|
| 383 |
-
"eval_steps_per_second": 20.801,
|
| 384 |
-
"step": 1250
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 4.778862722573233,
|
| 388 |
-
"grad_norm": 61.61293029785156,
|
| 389 |
-
"learning_rate": 5.2205882352941185e-05,
|
| 390 |
-
"loss": 0.9369,
|
| 391 |
-
"step": 1300
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 4.778862722573233,
|
| 395 |
-
"eval_loss": 1.2256300449371338,
|
| 396 |
-
"eval_runtime": 115.8928,
|
| 397 |
-
"eval_samples_per_second": 41.642,
|
| 398 |
-
"eval_steps_per_second": 20.821,
|
| 399 |
-
"step": 1300
|
| 400 |
-
},
|
| 401 |
-
{
|
| 402 |
-
"epoch": 4.962665134979897,
|
| 403 |
-
"grad_norm": 73.70500946044922,
|
| 404 |
-
"learning_rate": 5.036764705882353e-05,
|
| 405 |
-
"loss": 0.9301,
|
| 406 |
-
"step": 1350
|
| 407 |
-
},
|
| 408 |
-
{
|
| 409 |
-
"epoch": 4.962665134979897,
|
| 410 |
-
"eval_loss": 1.2280672788619995,
|
| 411 |
-
"eval_runtime": 116.1193,
|
| 412 |
-
"eval_samples_per_second": 41.561,
|
| 413 |
-
"eval_steps_per_second": 20.78,
|
| 414 |
-
"step": 1350
|
| 415 |
-
},
|
| 416 |
-
{
|
| 417 |
-
"epoch": 5.1464675473865595,
|
| 418 |
-
"grad_norm": 71.07781219482422,
|
| 419 |
-
"learning_rate": 4.8529411764705885e-05,
|
| 420 |
-
"loss": 0.9172,
|
| 421 |
-
"step": 1400
|
| 422 |
-
},
|
| 423 |
-
{
|
| 424 |
-
"epoch": 5.1464675473865595,
|
| 425 |
-
"eval_loss": 1.2280040979385376,
|
| 426 |
-
"eval_runtime": 116.2535,
|
| 427 |
-
"eval_samples_per_second": 41.513,
|
| 428 |
-
"eval_steps_per_second": 20.756,
|
| 429 |
-
"step": 1400
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 5.330269959793222,
|
| 433 |
-
"grad_norm": 59.39795684814453,
|
| 434 |
-
"learning_rate": 4.669117647058824e-05,
|
| 435 |
-
"loss": 0.9067,
|
| 436 |
-
"step": 1450
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 5.330269959793222,
|
| 440 |
-
"eval_loss": 1.2288336753845215,
|
| 441 |
-
"eval_runtime": 115.8973,
|
| 442 |
-
"eval_samples_per_second": 41.64,
|
| 443 |
-
"eval_steps_per_second": 20.82,
|
| 444 |
-
"step": 1450
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 5.514072372199885,
|
| 448 |
-
"grad_norm": 55.501617431640625,
|
| 449 |
-
"learning_rate": 4.485294117647059e-05,
|
| 450 |
-
"loss": 0.9004,
|
| 451 |
-
"step": 1500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 5.514072372199885,
|
| 455 |
-
"eval_loss": 1.2262146472930908,
|
| 456 |
-
"eval_runtime": 115.9478,
|
| 457 |
-
"eval_samples_per_second": 41.622,
|
| 458 |
-
"eval_steps_per_second": 20.811,
|
| 459 |
-
"step": 1500
|
| 460 |
-
},
|
| 461 |
-
{
|
| 462 |
-
"epoch": 5.697874784606548,
|
| 463 |
-
"grad_norm": 69.02213287353516,
|
| 464 |
-
"learning_rate": 4.301470588235295e-05,
|
| 465 |
-
"loss": 0.9035,
|
| 466 |
-
"step": 1550
|
| 467 |
-
},
|
| 468 |
-
{
|
| 469 |
-
"epoch": 5.697874784606548,
|
| 470 |
-
"eval_loss": 1.2236130237579346,
|
| 471 |
-
"eval_runtime": 116.1221,
|
| 472 |
-
"eval_samples_per_second": 41.56,
|
| 473 |
-
"eval_steps_per_second": 20.78,
|
| 474 |
-
"step": 1550
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 5.881677197013211,
|
| 478 |
-
"grad_norm": 45.09730529785156,
|
| 479 |
-
"learning_rate": 4.11764705882353e-05,
|
| 480 |
-
"loss": 0.8962,
|
| 481 |
-
"step": 1600
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 5.881677197013211,
|
| 485 |
-
"eval_loss": 1.2278504371643066,
|
| 486 |
-
"eval_runtime": 115.9916,
|
| 487 |
-
"eval_samples_per_second": 41.606,
|
| 488 |
-
"eval_steps_per_second": 20.803,
|
| 489 |
-
"step": 1600
|
| 490 |
-
},
|
| 491 |
-
{
|
| 492 |
-
"epoch": 6.0654796094198735,
|
| 493 |
-
"grad_norm": 47.59389877319336,
|
| 494 |
-
"learning_rate": 3.933823529411765e-05,
|
| 495 |
-
"loss": 0.8925,
|
| 496 |
-
"step": 1650
|
| 497 |
-
},
|
| 498 |
-
{
|
| 499 |
-
"epoch": 6.0654796094198735,
|
| 500 |
-
"eval_loss": 1.2326780557632446,
|
| 501 |
-
"eval_runtime": 116.7248,
|
| 502 |
-
"eval_samples_per_second": 41.345,
|
| 503 |
-
"eval_steps_per_second": 20.673,
|
| 504 |
-
"step": 1650
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 6.249282021826536,
|
| 508 |
-
"grad_norm": 45.18083190917969,
|
| 509 |
-
"learning_rate": 3.7500000000000003e-05,
|
| 510 |
-
"loss": 0.8771,
|
| 511 |
-
"step": 1700
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 6.249282021826536,
|
| 515 |
-
"eval_loss": 1.2302526235580444,
|
| 516 |
-
"eval_runtime": 115.8769,
|
| 517 |
-
"eval_samples_per_second": 41.648,
|
| 518 |
-
"eval_steps_per_second": 20.824,
|
| 519 |
-
"step": 1700
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 6.4330844342332,
|
| 523 |
-
"grad_norm": 40.455318450927734,
|
| 524 |
-
"learning_rate": 3.566176470588235e-05,
|
| 525 |
-
"loss": 0.8743,
|
| 526 |
-
"step": 1750
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 6.4330844342332,
|
| 530 |
-
"eval_loss": 1.2299398183822632,
|
| 531 |
-
"eval_runtime": 115.9106,
|
| 532 |
-
"eval_samples_per_second": 41.636,
|
| 533 |
-
"eval_steps_per_second": 20.818,
|
| 534 |
-
"step": 1750
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 6.616886846639862,
|
| 538 |
-
"grad_norm": 61.713111877441406,
|
| 539 |
-
"learning_rate": 3.382352941176471e-05,
|
| 540 |
-
"loss": 0.8735,
|
| 541 |
-
"step": 1800
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 6.616886846639862,
|
| 545 |
-
"eval_loss": 1.2240906953811646,
|
| 546 |
-
"eval_runtime": 116.0411,
|
| 547 |
-
"eval_samples_per_second": 41.589,
|
| 548 |
-
"eval_steps_per_second": 20.794,
|
| 549 |
-
"step": 1800
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 6.800689259046525,
|
| 553 |
-
"grad_norm": 69.22649383544922,
|
| 554 |
-
"learning_rate": 3.198529411764706e-05,
|
| 555 |
-
"loss": 0.8648,
|
| 556 |
-
"step": 1850
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 6.800689259046525,
|
| 560 |
-
"eval_loss": 1.2253305912017822,
|
| 561 |
-
"eval_runtime": 115.8996,
|
| 562 |
-
"eval_samples_per_second": 41.639,
|
| 563 |
-
"eval_steps_per_second": 20.82,
|
| 564 |
-
"step": 1850
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 6.9844916714531875,
|
| 568 |
-
"grad_norm": 65.4384994506836,
|
| 569 |
-
"learning_rate": 3.0147058823529413e-05,
|
| 570 |
-
"loss": 0.8649,
|
| 571 |
-
"step": 1900
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 6.9844916714531875,
|
| 575 |
-
"eval_loss": 1.2292358875274658,
|
| 576 |
-
"eval_runtime": 116.0285,
|
| 577 |
-
"eval_samples_per_second": 41.593,
|
| 578 |
-
"eval_steps_per_second": 20.797,
|
| 579 |
-
"step": 1900
|
| 580 |
-
},
|
| 581 |
-
{
|
| 582 |
-
"epoch": 7.168294083859851,
|
| 583 |
-
"grad_norm": 46.392173767089844,
|
| 584 |
-
"learning_rate": 2.8308823529411766e-05,
|
| 585 |
-
"loss": 0.8475,
|
| 586 |
-
"step": 1950
|
| 587 |
-
},
|
| 588 |
-
{
|
| 589 |
-
"epoch": 7.168294083859851,
|
| 590 |
-
"eval_loss": 1.2355010509490967,
|
| 591 |
-
"eval_runtime": 116.0581,
|
| 592 |
-
"eval_samples_per_second": 41.583,
|
| 593 |
-
"eval_steps_per_second": 20.791,
|
| 594 |
-
"step": 1950
|
| 595 |
-
},
|
| 596 |
-
{
|
| 597 |
-
"epoch": 7.352096496266514,
|
| 598 |
-
"grad_norm": 64.82035827636719,
|
| 599 |
-
"learning_rate": 2.647058823529412e-05,
|
| 600 |
-
"loss": 0.8496,
|
| 601 |
-
"step": 2000
|
| 602 |
-
},
|
| 603 |
-
{
|
| 604 |
-
"epoch": 7.352096496266514,
|
| 605 |
-
"eval_loss": 1.2320975065231323,
|
| 606 |
-
"eval_runtime": 115.9151,
|
| 607 |
-
"eval_samples_per_second": 41.634,
|
| 608 |
-
"eval_steps_per_second": 20.817,
|
| 609 |
-
"step": 2000
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 7.535898908673176,
|
| 613 |
-
"grad_norm": 46.27527618408203,
|
| 614 |
-
"learning_rate": 2.4632352941176472e-05,
|
| 615 |
-
"loss": 0.8452,
|
| 616 |
-
"step": 2050
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 7.535898908673176,
|
| 620 |
-
"eval_loss": 1.2323057651519775,
|
| 621 |
-
"eval_runtime": 115.981,
|
| 622 |
-
"eval_samples_per_second": 41.61,
|
| 623 |
-
"eval_steps_per_second": 20.805,
|
| 624 |
-
"step": 2050
|
| 625 |
-
},
|
| 626 |
-
{
|
| 627 |
-
"epoch": 7.719701321079839,
|
| 628 |
-
"grad_norm": 52.665435791015625,
|
| 629 |
-
"learning_rate": 2.2794117647058825e-05,
|
| 630 |
-
"loss": 0.8423,
|
| 631 |
-
"step": 2100
|
| 632 |
-
},
|
| 633 |
-
{
|
| 634 |
-
"epoch": 7.719701321079839,
|
| 635 |
-
"eval_loss": 1.2322068214416504,
|
| 636 |
-
"eval_runtime": 115.9188,
|
| 637 |
-
"eval_samples_per_second": 41.633,
|
| 638 |
-
"eval_steps_per_second": 20.816,
|
| 639 |
-
"step": 2100
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 7.903503733486502,
|
| 643 |
-
"grad_norm": 74.63914489746094,
|
| 644 |
-
"learning_rate": 2.0955882352941178e-05,
|
| 645 |
-
"loss": 0.8421,
|
| 646 |
-
"step": 2150
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 7.903503733486502,
|
| 650 |
-
"eval_loss": 1.2284280061721802,
|
| 651 |
-
"eval_runtime": 115.8808,
|
| 652 |
-
"eval_samples_per_second": 41.646,
|
| 653 |
-
"eval_steps_per_second": 20.823,
|
| 654 |
-
"step": 2150
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 8.087306145893164,
|
| 658 |
-
"grad_norm": 45.93680191040039,
|
| 659 |
-
"learning_rate": 1.9117647058823528e-05,
|
| 660 |
-
"loss": 0.8363,
|
| 661 |
-
"step": 2200
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 8.087306145893164,
|
| 665 |
-
"eval_loss": 1.2366794347763062,
|
| 666 |
-
"eval_runtime": 115.7882,
|
| 667 |
-
"eval_samples_per_second": 41.68,
|
| 668 |
-
"eval_steps_per_second": 20.84,
|
| 669 |
-
"step": 2200
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 8.271108558299828,
|
| 673 |
-
"grad_norm": 73.48126983642578,
|
| 674 |
-
"learning_rate": 1.7279411764705884e-05,
|
| 675 |
-
"loss": 0.8333,
|
| 676 |
-
"step": 2250
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 8.271108558299828,
|
| 680 |
-
"eval_loss": 1.2330245971679688,
|
| 681 |
-
"eval_runtime": 115.848,
|
| 682 |
-
"eval_samples_per_second": 41.658,
|
| 683 |
-
"eval_steps_per_second": 20.829,
|
| 684 |
-
"step": 2250
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 8.454910970706491,
|
| 688 |
-
"grad_norm": 76.98050689697266,
|
| 689 |
-
"learning_rate": 1.5441176470588237e-05,
|
| 690 |
-
"loss": 0.8267,
|
| 691 |
-
"step": 2300
|
| 692 |
-
},
|
| 693 |
-
{
|
| 694 |
-
"epoch": 8.454910970706491,
|
| 695 |
-
"eval_loss": 1.2376160621643066,
|
| 696 |
-
"eval_runtime": 115.9174,
|
| 697 |
-
"eval_samples_per_second": 41.633,
|
| 698 |
-
"eval_steps_per_second": 20.817,
|
| 699 |
-
"step": 2300
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 8.638713383113153,
|
| 703 |
-
"grad_norm": 128.97714233398438,
|
| 704 |
-
"learning_rate": 1.3602941176470587e-05,
|
| 705 |
-
"loss": 0.8217,
|
| 706 |
-
"step": 2350
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 8.638713383113153,
|
| 710 |
-
"eval_loss": 1.2340155839920044,
|
| 711 |
-
"eval_runtime": 115.9099,
|
| 712 |
-
"eval_samples_per_second": 41.636,
|
| 713 |
-
"eval_steps_per_second": 20.818,
|
| 714 |
-
"step": 2350
|
| 715 |
-
},
|
| 716 |
-
{
|
| 717 |
-
"epoch": 8.822515795519816,
|
| 718 |
-
"grad_norm": 55.4945182800293,
|
| 719 |
-
"learning_rate": 1.1764705882352942e-05,
|
| 720 |
-
"loss": 0.8221,
|
| 721 |
-
"step": 2400
|
| 722 |
-
},
|
| 723 |
-
{
|
| 724 |
-
"epoch": 8.822515795519816,
|
| 725 |
-
"eval_loss": 1.2404063940048218,
|
| 726 |
-
"eval_runtime": 116.0058,
|
| 727 |
-
"eval_samples_per_second": 41.601,
|
| 728 |
-
"eval_steps_per_second": 20.801,
|
| 729 |
-
"step": 2400
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 9.00631820792648,
|
| 733 |
-
"grad_norm": 69.84994506835938,
|
| 734 |
-
"learning_rate": 9.926470588235293e-06,
|
| 735 |
-
"loss": 0.8202,
|
| 736 |
-
"step": 2450
|
| 737 |
-
},
|
| 738 |
-
{
|
| 739 |
-
"epoch": 9.00631820792648,
|
| 740 |
-
"eval_loss": 1.2383702993392944,
|
| 741 |
-
"eval_runtime": 116.7345,
|
| 742 |
-
"eval_samples_per_second": 41.342,
|
| 743 |
-
"eval_steps_per_second": 20.671,
|
| 744 |
-
"step": 2450
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 9.190120620333142,
|
| 748 |
-
"grad_norm": 75.23961639404297,
|
| 749 |
-
"learning_rate": 8.088235294117648e-06,
|
| 750 |
-
"loss": 0.8143,
|
| 751 |
-
"step": 2500
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 9.190120620333142,
|
| 755 |
-
"eval_loss": 1.236066460609436,
|
| 756 |
-
"eval_runtime": 115.996,
|
| 757 |
-
"eval_samples_per_second": 41.605,
|
| 758 |
-
"eval_steps_per_second": 20.802,
|
| 759 |
-
"step": 2500
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 9.373923032739805,
|
| 763 |
-
"grad_norm": 62.9267692565918,
|
| 764 |
-
"learning_rate": 6.25e-06,
|
| 765 |
-
"loss": 0.8106,
|
| 766 |
-
"step": 2550
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 9.373923032739805,
|
| 770 |
-
"eval_loss": 1.23640775680542,
|
| 771 |
-
"eval_runtime": 115.848,
|
| 772 |
-
"eval_samples_per_second": 41.658,
|
| 773 |
-
"eval_steps_per_second": 20.829,
|
| 774 |
-
"step": 2550
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 9.557725445146467,
|
| 778 |
-
"grad_norm": 54.76566696166992,
|
| 779 |
-
"learning_rate": 4.411764705882353e-06,
|
| 780 |
-
"loss": 0.8144,
|
| 781 |
-
"step": 2600
|
| 782 |
-
},
|
| 783 |
-
{
|
| 784 |
-
"epoch": 9.557725445146467,
|
| 785 |
-
"eval_loss": 1.2419943809509277,
|
| 786 |
-
"eval_runtime": 116.0232,
|
| 787 |
-
"eval_samples_per_second": 41.595,
|
| 788 |
-
"eval_steps_per_second": 20.798,
|
| 789 |
-
"step": 2600
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 9.74152785755313,
|
| 793 |
-
"grad_norm": 51.20401382446289,
|
| 794 |
-
"learning_rate": 2.573529411764706e-06,
|
| 795 |
-
"loss": 0.8061,
|
| 796 |
-
"step": 2650
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 9.74152785755313,
|
| 800 |
-
"eval_loss": 1.2368206977844238,
|
| 801 |
-
"eval_runtime": 116.069,
|
| 802 |
-
"eval_samples_per_second": 41.579,
|
| 803 |
-
"eval_steps_per_second": 20.789,
|
| 804 |
-
"step": 2650
|
| 805 |
-
},
|
| 806 |
-
{
|
| 807 |
-
"epoch": 9.925330269959794,
|
| 808 |
-
"grad_norm": 52.51292419433594,
|
| 809 |
-
"learning_rate": 7.352941176470589e-07,
|
| 810 |
-
"loss": 0.8122,
|
| 811 |
-
"step": 2700
|
| 812 |
-
},
|
| 813 |
-
{
|
| 814 |
-
"epoch": 9.925330269959794,
|
| 815 |
-
"eval_loss": 1.2355531454086304,
|
| 816 |
-
"eval_runtime": 116.1057,
|
| 817 |
-
"eval_samples_per_second": 41.566,
|
| 818 |
-
"eval_steps_per_second": 20.783,
|
| 819 |
-
"step": 2700
|
| 820 |
}
|
| 821 |
],
|
| 822 |
"logging_steps": 50,
|
| 823 |
-
"max_steps":
|
| 824 |
"num_input_tokens_seen": 0,
|
| 825 |
-
"num_train_epochs":
|
| 826 |
"save_steps": 500,
|
| 827 |
"stateful_callbacks": {
|
| 828 |
"TrainerControl": {
|
|
@@ -831,12 +111,12 @@
|
|
| 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": 0.9979996362975087,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 343,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.14548099654482632,
|
| 13 |
+
"grad_norm": 92.38280487060547,
|
| 14 |
+
"learning_rate": 9.927113702623908e-05,
|
| 15 |
+
"loss": 1.3958,
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.14548099654482632,
|
| 20 |
+
"eval_loss": 1.3268945217132568,
|
| 21 |
+
"eval_runtime": 116.5998,
|
| 22 |
+
"eval_samples_per_second": 41.389,
|
| 23 |
+
"eval_steps_per_second": 20.695,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"epoch": 0.29096199308965265,
|
| 28 |
+
"grad_norm": 75.30126190185547,
|
| 29 |
+
"learning_rate": 9.854227405247813e-05,
|
| 30 |
+
"loss": 1.3228,
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
+
"epoch": 0.29096199308965265,
|
| 35 |
+
"eval_loss": 1.2796554565429688,
|
| 36 |
+
"eval_runtime": 116.9514,
|
| 37 |
+
"eval_samples_per_second": 41.265,
|
| 38 |
+
"eval_steps_per_second": 20.633,
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.436442989634479,
|
| 43 |
+
"grad_norm": 113.90583038330078,
|
| 44 |
+
"learning_rate": 9.781341107871722e-05,
|
| 45 |
+
"loss": 1.2818,
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
+
"epoch": 0.436442989634479,
|
| 50 |
+
"eval_loss": 1.2467392683029175,
|
| 51 |
+
"eval_runtime": 116.7421,
|
| 52 |
+
"eval_samples_per_second": 41.339,
|
| 53 |
+
"eval_steps_per_second": 20.669,
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
+
"epoch": 0.5819239861793053,
|
| 58 |
+
"grad_norm": 75.85360717773438,
|
| 59 |
+
"learning_rate": 9.708454810495627e-05,
|
| 60 |
+
"loss": 1.255,
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
+
"epoch": 0.5819239861793053,
|
| 65 |
+
"eval_loss": 1.2072206735610962,
|
| 66 |
+
"eval_runtime": 116.5054,
|
| 67 |
+
"eval_samples_per_second": 41.423,
|
| 68 |
+
"eval_steps_per_second": 20.711,
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.7274049827241317,
|
| 73 |
+
"grad_norm": 62.863895416259766,
|
| 74 |
+
"learning_rate": 9.635568513119534e-05,
|
| 75 |
+
"loss": 1.2329,
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
+
"epoch": 0.7274049827241317,
|
| 80 |
+
"eval_loss": 1.1830443143844604,
|
| 81 |
+
"eval_runtime": 116.5694,
|
| 82 |
+
"eval_samples_per_second": 41.4,
|
| 83 |
+
"eval_steps_per_second": 20.7,
|
| 84 |
"step": 250
|
| 85 |
},
|
| 86 |
{
|
| 87 |
+
"epoch": 0.872885979268958,
|
| 88 |
+
"grad_norm": 67.02438354492188,
|
| 89 |
+
"learning_rate": 9.56268221574344e-05,
|
| 90 |
+
"loss": 1.2113,
|
| 91 |
"step": 300
|
| 92 |
},
|
| 93 |
{
|
| 94 |
+
"epoch": 0.872885979268958,
|
| 95 |
+
"eval_loss": 1.1557544469833374,
|
| 96 |
+
"eval_runtime": 116.6632,
|
| 97 |
+
"eval_samples_per_second": 41.367,
|
| 98 |
+
"eval_steps_per_second": 20.683,
|
| 99 |
"step": 300
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 100 |
}
|
| 101 |
],
|
| 102 |
"logging_steps": 50,
|
| 103 |
+
"max_steps": 6860,
|
| 104 |
"num_input_tokens_seen": 0,
|
| 105 |
+
"num_train_epochs": 20,
|
| 106 |
"save_steps": 500,
|
| 107 |
"stateful_callbacks": {
|
| 108 |
"TrainerControl": {
|
|
|
|
| 111 |
"should_evaluate": false,
|
| 112 |
"should_log": false,
|
| 113 |
"should_save": true,
|
| 114 |
+
"should_training_stop": false
|
| 115 |
},
|
| 116 |
"attributes": {}
|
| 117 |
}
|
| 118 |
},
|
| 119 |
+
"total_flos": 8.9337717422293e+16,
|
| 120 |
"train_batch_size": 2,
|
| 121 |
"trial_name": null,
|
| 122 |
"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
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e7fd3757bfae30126ccb3f467f9722223f1ea2ad8678404c8edb0c2dfc443523
|
| 3 |
+
size 5368
|