Remove unnecessary files to save space
Browse files
checkpoint-3052/trainer_state.json
DELETED
|
@@ -1,719 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": null,
|
| 3 |
-
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.10143578835416113,
|
| 5 |
-
"eval_steps": 500,
|
| 6 |
-
"global_step": 3052,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.001030311087476735,
|
| 13 |
-
"grad_norm": 60.25547409057617,
|
| 14 |
-
"learning_rate": 1.0157273918741808e-06,
|
| 15 |
-
"loss": 8.8455,
|
| 16 |
-
"step": 31
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.00206062217495347,
|
| 20 |
-
"grad_norm": 15.669363975524902,
|
| 21 |
-
"learning_rate": 2.0314547837483616e-06,
|
| 22 |
-
"loss": 7.1553,
|
| 23 |
-
"step": 62
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.003090933262430205,
|
| 27 |
-
"grad_norm": 15.366345405578613,
|
| 28 |
-
"learning_rate": 3.0471821756225426e-06,
|
| 29 |
-
"loss": 5.8784,
|
| 30 |
-
"step": 93
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.00412124434990694,
|
| 34 |
-
"grad_norm": 36.30561828613281,
|
| 35 |
-
"learning_rate": 4.062909567496723e-06,
|
| 36 |
-
"loss": 4.7708,
|
| 37 |
-
"step": 124
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.005151555437383675,
|
| 41 |
-
"grad_norm": 27.202678680419922,
|
| 42 |
-
"learning_rate": 5.078636959370905e-06,
|
| 43 |
-
"loss": 4.1629,
|
| 44 |
-
"step": 155
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.00618186652486041,
|
| 48 |
-
"grad_norm": 24.30484962463379,
|
| 49 |
-
"learning_rate": 6.094364351245085e-06,
|
| 50 |
-
"loss": 3.867,
|
| 51 |
-
"step": 186
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.007212177612337145,
|
| 55 |
-
"grad_norm": 19.916366577148438,
|
| 56 |
-
"learning_rate": 7.110091743119267e-06,
|
| 57 |
-
"loss": 3.6131,
|
| 58 |
-
"step": 217
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.00824248869981388,
|
| 62 |
-
"grad_norm": 17.577274322509766,
|
| 63 |
-
"learning_rate": 8.125819134993446e-06,
|
| 64 |
-
"loss": 3.4772,
|
| 65 |
-
"step": 248
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.009272799787290615,
|
| 69 |
-
"grad_norm": 12.133190155029297,
|
| 70 |
-
"learning_rate": 9.141546526867629e-06,
|
| 71 |
-
"loss": 3.3218,
|
| 72 |
-
"step": 279
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.01030311087476735,
|
| 76 |
-
"grad_norm": 19.79263687133789,
|
| 77 |
-
"learning_rate": 1.015727391874181e-05,
|
| 78 |
-
"loss": 3.2055,
|
| 79 |
-
"step": 310
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.011333421962244085,
|
| 83 |
-
"grad_norm": 16.38133430480957,
|
| 84 |
-
"learning_rate": 1.117300131061599e-05,
|
| 85 |
-
"loss": 3.1062,
|
| 86 |
-
"step": 341
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.01236373304972082,
|
| 90 |
-
"grad_norm": 12.638299942016602,
|
| 91 |
-
"learning_rate": 1.218872870249017e-05,
|
| 92 |
-
"loss": 3.0106,
|
| 93 |
-
"step": 372
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.013394044137197554,
|
| 97 |
-
"grad_norm": 9.46596908569336,
|
| 98 |
-
"learning_rate": 1.3204456094364351e-05,
|
| 99 |
-
"loss": 2.924,
|
| 100 |
-
"step": 403
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.01442435522467429,
|
| 104 |
-
"grad_norm": 10.945392608642578,
|
| 105 |
-
"learning_rate": 1.4220183486238533e-05,
|
| 106 |
-
"loss": 2.844,
|
| 107 |
-
"step": 434
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.015454666312151024,
|
| 111 |
-
"grad_norm": 8.474015235900879,
|
| 112 |
-
"learning_rate": 1.5235910878112714e-05,
|
| 113 |
-
"loss": 2.7892,
|
| 114 |
-
"step": 465
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.01648497739962776,
|
| 118 |
-
"grad_norm": 9.370804786682129,
|
| 119 |
-
"learning_rate": 1.6251638269986893e-05,
|
| 120 |
-
"loss": 2.7509,
|
| 121 |
-
"step": 496
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.017515288487104493,
|
| 125 |
-
"grad_norm": 11.63398551940918,
|
| 126 |
-
"learning_rate": 1.7267365661861077e-05,
|
| 127 |
-
"loss": 2.6999,
|
| 128 |
-
"step": 527
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.01854559957458123,
|
| 132 |
-
"grad_norm": 9.17713451385498,
|
| 133 |
-
"learning_rate": 1.8283093053735257e-05,
|
| 134 |
-
"loss": 2.6459,
|
| 135 |
-
"step": 558
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.019575910662057962,
|
| 139 |
-
"grad_norm": 7.119054794311523,
|
| 140 |
-
"learning_rate": 1.9298820445609438e-05,
|
| 141 |
-
"loss": 2.603,
|
| 142 |
-
"step": 589
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.0206062217495347,
|
| 146 |
-
"grad_norm": 6.653646945953369,
|
| 147 |
-
"learning_rate": 2.031454783748362e-05,
|
| 148 |
-
"loss": 2.5588,
|
| 149 |
-
"step": 620
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.021636532837011432,
|
| 153 |
-
"grad_norm": 8.332653045654297,
|
| 154 |
-
"learning_rate": 2.13302752293578e-05,
|
| 155 |
-
"loss": 2.5357,
|
| 156 |
-
"step": 651
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.02266684392448817,
|
| 160 |
-
"grad_norm": 6.4949116706848145,
|
| 161 |
-
"learning_rate": 2.234600262123198e-05,
|
| 162 |
-
"loss": 2.4967,
|
| 163 |
-
"step": 682
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.023697155011964902,
|
| 167 |
-
"grad_norm": 9.41009521484375,
|
| 168 |
-
"learning_rate": 2.336173001310616e-05,
|
| 169 |
-
"loss": 2.4563,
|
| 170 |
-
"step": 713
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.02472746609944164,
|
| 174 |
-
"grad_norm": 7.840345859527588,
|
| 175 |
-
"learning_rate": 2.437745740498034e-05,
|
| 176 |
-
"loss": 2.4383,
|
| 177 |
-
"step": 744
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.025757777186918372,
|
| 181 |
-
"grad_norm": 6.116458415985107,
|
| 182 |
-
"learning_rate": 2.5393184796854525e-05,
|
| 183 |
-
"loss": 2.3817,
|
| 184 |
-
"step": 775
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.02678808827439511,
|
| 188 |
-
"grad_norm": 5.938300609588623,
|
| 189 |
-
"learning_rate": 2.6408912188728702e-05,
|
| 190 |
-
"loss": 2.3508,
|
| 191 |
-
"step": 806
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.027818399361871842,
|
| 195 |
-
"grad_norm": 5.4408345222473145,
|
| 196 |
-
"learning_rate": 2.7424639580602886e-05,
|
| 197 |
-
"loss": 2.3325,
|
| 198 |
-
"step": 837
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.02884871044934858,
|
| 202 |
-
"grad_norm": 5.375136375427246,
|
| 203 |
-
"learning_rate": 2.8440366972477066e-05,
|
| 204 |
-
"loss": 2.3101,
|
| 205 |
-
"step": 868
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.029879021536825312,
|
| 209 |
-
"grad_norm": 5.149726867675781,
|
| 210 |
-
"learning_rate": 2.9456094364351244e-05,
|
| 211 |
-
"loss": 2.282,
|
| 212 |
-
"step": 899
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.03090933262430205,
|
| 216 |
-
"grad_norm": 4.591221332550049,
|
| 217 |
-
"learning_rate": 3.0471821756225428e-05,
|
| 218 |
-
"loss": 2.2427,
|
| 219 |
-
"step": 930
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.031939643711778785,
|
| 223 |
-
"grad_norm": 4.977034091949463,
|
| 224 |
-
"learning_rate": 3.148754914809961e-05,
|
| 225 |
-
"loss": 2.2218,
|
| 226 |
-
"step": 961
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.03296995479925552,
|
| 230 |
-
"grad_norm": 5.038781642913818,
|
| 231 |
-
"learning_rate": 3.2503276539973785e-05,
|
| 232 |
-
"loss": 2.2044,
|
| 233 |
-
"step": 992
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.03400026588673225,
|
| 237 |
-
"grad_norm": 4.872281551361084,
|
| 238 |
-
"learning_rate": 3.351900393184797e-05,
|
| 239 |
-
"loss": 2.1657,
|
| 240 |
-
"step": 1023
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.035030576974208985,
|
| 244 |
-
"grad_norm": 4.370841979980469,
|
| 245 |
-
"learning_rate": 3.453473132372215e-05,
|
| 246 |
-
"loss": 2.1365,
|
| 247 |
-
"step": 1054
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.036060888061685725,
|
| 251 |
-
"grad_norm": 4.087072849273682,
|
| 252 |
-
"learning_rate": 3.555045871559633e-05,
|
| 253 |
-
"loss": 2.1253,
|
| 254 |
-
"step": 1085
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.03709119914916246,
|
| 258 |
-
"grad_norm": 4.113957405090332,
|
| 259 |
-
"learning_rate": 3.6566186107470514e-05,
|
| 260 |
-
"loss": 2.0973,
|
| 261 |
-
"step": 1116
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.03812151023663919,
|
| 265 |
-
"grad_norm": 4.0119733810424805,
|
| 266 |
-
"learning_rate": 3.7581913499344695e-05,
|
| 267 |
-
"loss": 2.1024,
|
| 268 |
-
"step": 1147
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.039151821324115925,
|
| 272 |
-
"grad_norm": 4.247573375701904,
|
| 273 |
-
"learning_rate": 3.8597640891218876e-05,
|
| 274 |
-
"loss": 2.0722,
|
| 275 |
-
"step": 1178
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.04018213241159266,
|
| 279 |
-
"grad_norm": 3.5575129985809326,
|
| 280 |
-
"learning_rate": 3.9613368283093056e-05,
|
| 281 |
-
"loss": 2.056,
|
| 282 |
-
"step": 1209
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.0412124434990694,
|
| 286 |
-
"grad_norm": 3.8885862827301025,
|
| 287 |
-
"learning_rate": 4.062909567496724e-05,
|
| 288 |
-
"loss": 2.0389,
|
| 289 |
-
"step": 1240
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.04224275458654613,
|
| 293 |
-
"grad_norm": 3.680628538131714,
|
| 294 |
-
"learning_rate": 4.164482306684142e-05,
|
| 295 |
-
"loss": 2.0385,
|
| 296 |
-
"step": 1271
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.043273065674022865,
|
| 300 |
-
"grad_norm": 3.780876874923706,
|
| 301 |
-
"learning_rate": 4.26605504587156e-05,
|
| 302 |
-
"loss": 2.0097,
|
| 303 |
-
"step": 1302
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.0443033767614996,
|
| 307 |
-
"grad_norm": 4.235328674316406,
|
| 308 |
-
"learning_rate": 4.367627785058978e-05,
|
| 309 |
-
"loss": 2.0024,
|
| 310 |
-
"step": 1333
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.04533368784897634,
|
| 314 |
-
"grad_norm": 3.326941967010498,
|
| 315 |
-
"learning_rate": 4.469200524246396e-05,
|
| 316 |
-
"loss": 1.9953,
|
| 317 |
-
"step": 1364
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 0.04636399893645307,
|
| 321 |
-
"grad_norm": 3.28456449508667,
|
| 322 |
-
"learning_rate": 4.570773263433814e-05,
|
| 323 |
-
"loss": 1.9579,
|
| 324 |
-
"step": 1395
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.047394310023929805,
|
| 328 |
-
"grad_norm": 16.107433319091797,
|
| 329 |
-
"learning_rate": 4.672346002621232e-05,
|
| 330 |
-
"loss": 1.9701,
|
| 331 |
-
"step": 1426
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.04842462111140654,
|
| 335 |
-
"grad_norm": 3.5708224773406982,
|
| 336 |
-
"learning_rate": 4.77391874180865e-05,
|
| 337 |
-
"loss": 1.9621,
|
| 338 |
-
"step": 1457
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.04945493219888328,
|
| 342 |
-
"grad_norm": 2.9053499698638916,
|
| 343 |
-
"learning_rate": 4.875491480996068e-05,
|
| 344 |
-
"loss": 1.9458,
|
| 345 |
-
"step": 1488
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.05048524328636001,
|
| 349 |
-
"grad_norm": 3.0863258838653564,
|
| 350 |
-
"learning_rate": 4.977064220183487e-05,
|
| 351 |
-
"loss": 1.9483,
|
| 352 |
-
"step": 1519
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 0.051515554373836744,
|
| 356 |
-
"grad_norm": 2.9012269973754883,
|
| 357 |
-
"learning_rate": 4.9999915451558777e-05,
|
| 358 |
-
"loss": 1.928,
|
| 359 |
-
"step": 1550
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.05254586546131348,
|
| 363 |
-
"grad_norm": 3.0949041843414307,
|
| 364 |
-
"learning_rate": 4.999955597496219e-05,
|
| 365 |
-
"loss": 1.9229,
|
| 366 |
-
"step": 1581
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.05357617654879022,
|
| 370 |
-
"grad_norm": 2.8687901496887207,
|
| 371 |
-
"learning_rate": 4.9998914381774255e-05,
|
| 372 |
-
"loss": 1.915,
|
| 373 |
-
"step": 1612
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.05460648763626695,
|
| 377 |
-
"grad_norm": 3.2136878967285156,
|
| 378 |
-
"learning_rate": 4.999799067923527e-05,
|
| 379 |
-
"loss": 1.9197,
|
| 380 |
-
"step": 1643
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.055636798723743684,
|
| 384 |
-
"grad_norm": 2.590843677520752,
|
| 385 |
-
"learning_rate": 4.999678487776908e-05,
|
| 386 |
-
"loss": 1.8756,
|
| 387 |
-
"step": 1674
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.05666710981122042,
|
| 391 |
-
"grad_norm": 2.64634108543396,
|
| 392 |
-
"learning_rate": 4.9995296990983006e-05,
|
| 393 |
-
"loss": 1.9033,
|
| 394 |
-
"step": 1705
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.05769742089869716,
|
| 398 |
-
"grad_norm": 3.0151331424713135,
|
| 399 |
-
"learning_rate": 4.999352703566763e-05,
|
| 400 |
-
"loss": 1.8883,
|
| 401 |
-
"step": 1736
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.05872773198617389,
|
| 405 |
-
"grad_norm": 2.526806354522705,
|
| 406 |
-
"learning_rate": 4.999147503179668e-05,
|
| 407 |
-
"loss": 1.8666,
|
| 408 |
-
"step": 1767
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.059758043073650624,
|
| 412 |
-
"grad_norm": 2.510300397872925,
|
| 413 |
-
"learning_rate": 4.998914100252672e-05,
|
| 414 |
-
"loss": 1.854,
|
| 415 |
-
"step": 1798
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.06078835416112736,
|
| 419 |
-
"grad_norm": 2.4867682456970215,
|
| 420 |
-
"learning_rate": 4.998652497419696e-05,
|
| 421 |
-
"loss": 1.8548,
|
| 422 |
-
"step": 1829
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.0618186652486041,
|
| 426 |
-
"grad_norm": 2.3920586109161377,
|
| 427 |
-
"learning_rate": 4.9983626976328927e-05,
|
| 428 |
-
"loss": 1.8495,
|
| 429 |
-
"step": 1860
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.06284897633608083,
|
| 433 |
-
"grad_norm": 2.714177370071411,
|
| 434 |
-
"learning_rate": 4.998044704162613e-05,
|
| 435 |
-
"loss": 1.8433,
|
| 436 |
-
"step": 1891
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.06387928742355757,
|
| 440 |
-
"grad_norm": 2.3094465732574463,
|
| 441 |
-
"learning_rate": 4.9976985205973705e-05,
|
| 442 |
-
"loss": 1.8382,
|
| 443 |
-
"step": 1922
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.0649095985110343,
|
| 447 |
-
"grad_norm": 2.47184419631958,
|
| 448 |
-
"learning_rate": 4.997324150843799e-05,
|
| 449 |
-
"loss": 1.8464,
|
| 450 |
-
"step": 1953
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.06593990959851104,
|
| 454 |
-
"grad_norm": 2.391841411590576,
|
| 455 |
-
"learning_rate": 4.99692159912661e-05,
|
| 456 |
-
"loss": 1.8179,
|
| 457 |
-
"step": 1984
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.06697022068598776,
|
| 461 |
-
"grad_norm": 2.2471864223480225,
|
| 462 |
-
"learning_rate": 4.996490869988546e-05,
|
| 463 |
-
"loss": 1.8149,
|
| 464 |
-
"step": 2015
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.0680005317734645,
|
| 468 |
-
"grad_norm": 2.5497376918792725,
|
| 469 |
-
"learning_rate": 4.996031968290326e-05,
|
| 470 |
-
"loss": 1.8099,
|
| 471 |
-
"step": 2046
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.06903084286094124,
|
| 475 |
-
"grad_norm": 2.330463409423828,
|
| 476 |
-
"learning_rate": 4.995544899210594e-05,
|
| 477 |
-
"loss": 1.8267,
|
| 478 |
-
"step": 2077
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.07006115394841797,
|
| 482 |
-
"grad_norm": 2.3259341716766357,
|
| 483 |
-
"learning_rate": 4.9950296682458583e-05,
|
| 484 |
-
"loss": 1.7801,
|
| 485 |
-
"step": 2108
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.07109146503589471,
|
| 489 |
-
"grad_norm": 2.1711952686309814,
|
| 490 |
-
"learning_rate": 4.994486281210429e-05,
|
| 491 |
-
"loss": 1.7961,
|
| 492 |
-
"step": 2139
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.07212177612337145,
|
| 496 |
-
"grad_norm": 2.1808884143829346,
|
| 497 |
-
"learning_rate": 4.9939147442363566e-05,
|
| 498 |
-
"loss": 1.8109,
|
| 499 |
-
"step": 2170
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.07315208721084818,
|
| 503 |
-
"grad_norm": 2.089256525039673,
|
| 504 |
-
"learning_rate": 4.9933150637733574e-05,
|
| 505 |
-
"loss": 1.8026,
|
| 506 |
-
"step": 2201
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.07418239829832492,
|
| 510 |
-
"grad_norm": 2.0864951610565186,
|
| 511 |
-
"learning_rate": 4.992687246588743e-05,
|
| 512 |
-
"loss": 1.7753,
|
| 513 |
-
"step": 2232
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.07521270938580164,
|
| 517 |
-
"grad_norm": 2.36157488822937,
|
| 518 |
-
"learning_rate": 4.992031299767347e-05,
|
| 519 |
-
"loss": 1.7746,
|
| 520 |
-
"step": 2263
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.07624302047327838,
|
| 524 |
-
"grad_norm": 2.5334439277648926,
|
| 525 |
-
"learning_rate": 4.9913472307114386e-05,
|
| 526 |
-
"loss": 1.7927,
|
| 527 |
-
"step": 2294
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.07727333156075512,
|
| 531 |
-
"grad_norm": 2.2565715312957764,
|
| 532 |
-
"learning_rate": 4.9906350471406446e-05,
|
| 533 |
-
"loss": 1.7668,
|
| 534 |
-
"step": 2325
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.07830364264823185,
|
| 538 |
-
"grad_norm": 2.1043128967285156,
|
| 539 |
-
"learning_rate": 4.989894757091861e-05,
|
| 540 |
-
"loss": 1.7771,
|
| 541 |
-
"step": 2356
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.07933395373570859,
|
| 545 |
-
"grad_norm": 1.9659819602966309,
|
| 546 |
-
"learning_rate": 4.989126368919158e-05,
|
| 547 |
-
"loss": 1.7666,
|
| 548 |
-
"step": 2387
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.08036426482318532,
|
| 552 |
-
"grad_norm": 2.0778403282165527,
|
| 553 |
-
"learning_rate": 4.988329891293693e-05,
|
| 554 |
-
"loss": 1.7405,
|
| 555 |
-
"step": 2418
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.08139457591066206,
|
| 559 |
-
"grad_norm": 2.1767923831939697,
|
| 560 |
-
"learning_rate": 4.987505333203608e-05,
|
| 561 |
-
"loss": 1.7495,
|
| 562 |
-
"step": 2449
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.0824248869981388,
|
| 566 |
-
"grad_norm": 2.260143280029297,
|
| 567 |
-
"learning_rate": 4.9866527039539276e-05,
|
| 568 |
-
"loss": 1.7504,
|
| 569 |
-
"step": 2480
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.08345519808561552,
|
| 573 |
-
"grad_norm": 2.18271803855896,
|
| 574 |
-
"learning_rate": 4.9857720131664594e-05,
|
| 575 |
-
"loss": 1.7456,
|
| 576 |
-
"step": 2511
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.08448550917309226,
|
| 580 |
-
"grad_norm": 2.209594964981079,
|
| 581 |
-
"learning_rate": 4.9848632707796773e-05,
|
| 582 |
-
"loss": 1.7528,
|
| 583 |
-
"step": 2542
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.085515820260569,
|
| 587 |
-
"grad_norm": 2.0666229724884033,
|
| 588 |
-
"learning_rate": 4.9839264870486155e-05,
|
| 589 |
-
"loss": 1.7517,
|
| 590 |
-
"step": 2573
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.08654613134804573,
|
| 594 |
-
"grad_norm": 2.1070454120635986,
|
| 595 |
-
"learning_rate": 4.9829616725447526e-05,
|
| 596 |
-
"loss": 1.7474,
|
| 597 |
-
"step": 2604
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.08757644243552247,
|
| 601 |
-
"grad_norm": 1.9430303573608398,
|
| 602 |
-
"learning_rate": 4.981968838155888e-05,
|
| 603 |
-
"loss": 1.7348,
|
| 604 |
-
"step": 2635
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.0886067535229992,
|
| 608 |
-
"grad_norm": 1.9638925790786743,
|
| 609 |
-
"learning_rate": 4.980947995086024e-05,
|
| 610 |
-
"loss": 1.7202,
|
| 611 |
-
"step": 2666
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.08963706461047594,
|
| 615 |
-
"grad_norm": 1.8845652341842651,
|
| 616 |
-
"learning_rate": 4.979899154855234e-05,
|
| 617 |
-
"loss": 1.7375,
|
| 618 |
-
"step": 2697
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.09066737569795268,
|
| 622 |
-
"grad_norm": 5.712058067321777,
|
| 623 |
-
"learning_rate": 4.9788223292995386e-05,
|
| 624 |
-
"loss": 1.7379,
|
| 625 |
-
"step": 2728
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.0916976867854294,
|
| 629 |
-
"grad_norm": 1.9520670175552368,
|
| 630 |
-
"learning_rate": 4.977717530570768e-05,
|
| 631 |
-
"loss": 1.7302,
|
| 632 |
-
"step": 2759
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.09272799787290614,
|
| 636 |
-
"grad_norm": 1.8802224397659302,
|
| 637 |
-
"learning_rate": 4.976584771136425e-05,
|
| 638 |
-
"loss": 1.74,
|
| 639 |
-
"step": 2790
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.09375830896038288,
|
| 643 |
-
"grad_norm": 2.1098153591156006,
|
| 644 |
-
"learning_rate": 4.975424063779547e-05,
|
| 645 |
-
"loss": 1.7024,
|
| 646 |
-
"step": 2821
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 0.09478862004785961,
|
| 650 |
-
"grad_norm": 2.1568291187286377,
|
| 651 |
-
"learning_rate": 4.974235421598557e-05,
|
| 652 |
-
"loss": 1.7131,
|
| 653 |
-
"step": 2852
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.09581893113533635,
|
| 657 |
-
"grad_norm": 1.8769980669021606,
|
| 658 |
-
"learning_rate": 4.973018858007122e-05,
|
| 659 |
-
"loss": 1.7008,
|
| 660 |
-
"step": 2883
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 0.09684924222281308,
|
| 664 |
-
"grad_norm": 1.8325533866882324,
|
| 665 |
-
"learning_rate": 4.9717743867339963e-05,
|
| 666 |
-
"loss": 1.7058,
|
| 667 |
-
"step": 2914
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.09787955331028982,
|
| 671 |
-
"grad_norm": 2.086416721343994,
|
| 672 |
-
"learning_rate": 4.9705020218228695e-05,
|
| 673 |
-
"loss": 1.711,
|
| 674 |
-
"step": 2945
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 0.09890986439776656,
|
| 678 |
-
"grad_norm": 1.8294793367385864,
|
| 679 |
-
"learning_rate": 4.969201777632205e-05,
|
| 680 |
-
"loss": 1.6998,
|
| 681 |
-
"step": 2976
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.09994017548524328,
|
| 685 |
-
"grad_norm": 2.0608153343200684,
|
| 686 |
-
"learning_rate": 4.9678736688350846e-05,
|
| 687 |
-
"loss": 1.6948,
|
| 688 |
-
"step": 3007
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 0.10097048657272002,
|
| 692 |
-
"grad_norm": 3.2166008949279785,
|
| 693 |
-
"learning_rate": 4.966517710419033e-05,
|
| 694 |
-
"loss": 1.6788,
|
| 695 |
-
"step": 3038
|
| 696 |
-
}
|
| 697 |
-
],
|
| 698 |
-
"logging_steps": 31,
|
| 699 |
-
"max_steps": 30517,
|
| 700 |
-
"num_input_tokens_seen": 0,
|
| 701 |
-
"num_train_epochs": 2,
|
| 702 |
-
"save_steps": 3052,
|
| 703 |
-
"stateful_callbacks": {
|
| 704 |
-
"TrainerControl": {
|
| 705 |
-
"args": {
|
| 706 |
-
"should_epoch_stop": false,
|
| 707 |
-
"should_evaluate": false,
|
| 708 |
-
"should_log": false,
|
| 709 |
-
"should_save": true,
|
| 710 |
-
"should_training_stop": false
|
| 711 |
-
},
|
| 712 |
-
"attributes": {}
|
| 713 |
-
}
|
| 714 |
-
},
|
| 715 |
-
"total_flos": 2.263945054022271e+18,
|
| 716 |
-
"train_batch_size": 8,
|
| 717 |
-
"trial_name": null,
|
| 718 |
-
"trial_params": null
|
| 719 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|