Delete trainer_state.json
Browse files- trainer_state.json +0 -695
trainer_state.json
DELETED
|
@@ -1,695 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_metric": null,
|
| 3 |
-
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 5.0,
|
| 5 |
-
"eval_steps": 500,
|
| 6 |
-
"global_step": 9580,
|
| 7 |
-
"is_hyper_param_search": false,
|
| 8 |
-
"is_local_process_zero": true,
|
| 9 |
-
"is_world_process_zero": true,
|
| 10 |
-
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.05,
|
| 13 |
-
"grad_norm": 1.0843530893325806,
|
| 14 |
-
"learning_rate": 0.00019994623498004714,
|
| 15 |
-
"loss": 1.8222,
|
| 16 |
-
"step": 100
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.1,
|
| 20 |
-
"grad_norm": 0.9015785455703735,
|
| 21 |
-
"learning_rate": 0.00019978499773373596,
|
| 22 |
-
"loss": 1.8565,
|
| 23 |
-
"step": 200
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.16,
|
| 27 |
-
"grad_norm": 1.2422339916229248,
|
| 28 |
-
"learning_rate": 0.00019951646163954176,
|
| 29 |
-
"loss": 1.7994,
|
| 30 |
-
"step": 300
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.21,
|
| 34 |
-
"grad_norm": 1.0499353408813477,
|
| 35 |
-
"learning_rate": 0.0001991409154544338,
|
| 36 |
-
"loss": 1.6984,
|
| 37 |
-
"step": 400
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.26,
|
| 41 |
-
"grad_norm": 1.122926950454712,
|
| 42 |
-
"learning_rate": 0.00019865876300337478,
|
| 43 |
-
"loss": 1.6973,
|
| 44 |
-
"step": 500
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.31,
|
| 48 |
-
"grad_norm": 1.3406814336776733,
|
| 49 |
-
"learning_rate": 0.00019807052274508773,
|
| 50 |
-
"loss": 1.7661,
|
| 51 |
-
"step": 600
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.37,
|
| 55 |
-
"grad_norm": 0.7354043126106262,
|
| 56 |
-
"learning_rate": 0.00019737682721455714,
|
| 57 |
-
"loss": 1.6948,
|
| 58 |
-
"step": 700
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.42,
|
| 62 |
-
"grad_norm": 2.118002414703369,
|
| 63 |
-
"learning_rate": 0.0001965784223428638,
|
| 64 |
-
"loss": 1.8664,
|
| 65 |
-
"step": 800
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.47,
|
| 69 |
-
"grad_norm": 1.6859912872314453,
|
| 70 |
-
"learning_rate": 0.00019567616665508485,
|
| 71 |
-
"loss": 1.7087,
|
| 72 |
-
"step": 900
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.52,
|
| 76 |
-
"grad_norm": 1.2144606113433838,
|
| 77 |
-
"learning_rate": 0.0001946710303471214,
|
| 78 |
-
"loss": 1.6826,
|
| 79 |
-
"step": 1000
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.57,
|
| 83 |
-
"grad_norm": 1.680190920829773,
|
| 84 |
-
"learning_rate": 0.00019356409424244655,
|
| 85 |
-
"loss": 1.5399,
|
| 86 |
-
"step": 1100
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.63,
|
| 90 |
-
"grad_norm": 1.834701418876648,
|
| 91 |
-
"learning_rate": 0.00019235654862989537,
|
| 92 |
-
"loss": 1.6394,
|
| 93 |
-
"step": 1200
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.68,
|
| 97 |
-
"grad_norm": 1.9838495254516602,
|
| 98 |
-
"learning_rate": 0.00019104969198374688,
|
| 99 |
-
"loss": 1.645,
|
| 100 |
-
"step": 1300
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.73,
|
| 104 |
-
"grad_norm": 2.141674280166626,
|
| 105 |
-
"learning_rate": 0.00018964492956747425,
|
| 106 |
-
"loss": 1.5853,
|
| 107 |
-
"step": 1400
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.78,
|
| 111 |
-
"grad_norm": 2.6233127117156982,
|
| 112 |
-
"learning_rate": 0.00018814377192266423,
|
| 113 |
-
"loss": 1.5778,
|
| 114 |
-
"step": 1500
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.84,
|
| 118 |
-
"grad_norm": 1.6887329816818237,
|
| 119 |
-
"learning_rate": 0.00018654783324473137,
|
| 120 |
-
"loss": 1.6368,
|
| 121 |
-
"step": 1600
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.89,
|
| 125 |
-
"grad_norm": 3.113426446914673,
|
| 126 |
-
"learning_rate": 0.00018487617447307124,
|
| 127 |
-
"loss": 1.6269,
|
| 128 |
-
"step": 1700
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.94,
|
| 132 |
-
"grad_norm": 2.1365339756011963,
|
| 133 |
-
"learning_rate": 0.00018309682531549338,
|
| 134 |
-
"loss": 1.7073,
|
| 135 |
-
"step": 1800
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.99,
|
| 139 |
-
"grad_norm": 2.3787343502044678,
|
| 140 |
-
"learning_rate": 0.00018122812210849337,
|
| 141 |
-
"loss": 1.5544,
|
| 142 |
-
"step": 1900
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 1.04,
|
| 146 |
-
"grad_norm": 2.1635191440582275,
|
| 147 |
-
"learning_rate": 0.00017927207426937544,
|
| 148 |
-
"loss": 1.3285,
|
| 149 |
-
"step": 2000
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 1.1,
|
| 153 |
-
"grad_norm": 2.331688642501831,
|
| 154 |
-
"learning_rate": 0.00017723078513716157,
|
| 155 |
-
"loss": 1.2529,
|
| 156 |
-
"step": 2100
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 1.15,
|
| 160 |
-
"grad_norm": 5.185802936553955,
|
| 161 |
-
"learning_rate": 0.00017510644971087015,
|
| 162 |
-
"loss": 1.3708,
|
| 163 |
-
"step": 2200
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 1.2,
|
| 167 |
-
"grad_norm": 3.1339027881622314,
|
| 168 |
-
"learning_rate": 0.0001729013522892329,
|
| 169 |
-
"loss": 1.3309,
|
| 170 |
-
"step": 2300
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 1.25,
|
| 174 |
-
"grad_norm": 3.4236795902252197,
|
| 175 |
-
"learning_rate": 0.0001706178640143872,
|
| 176 |
-
"loss": 1.4201,
|
| 177 |
-
"step": 2400
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 1.3,
|
| 181 |
-
"grad_norm": 3.849432945251465,
|
| 182 |
-
"learning_rate": 0.00016825844032218625,
|
| 183 |
-
"loss": 1.2956,
|
| 184 |
-
"step": 2500
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 1.36,
|
| 188 |
-
"grad_norm": 3.8217601776123047,
|
| 189 |
-
"learning_rate": 0.00016582561830186785,
|
| 190 |
-
"loss": 1.3375,
|
| 191 |
-
"step": 2600
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 1.41,
|
| 195 |
-
"grad_norm": 1.5418843030929565,
|
| 196 |
-
"learning_rate": 0.00016332201396792123,
|
| 197 |
-
"loss": 1.2897,
|
| 198 |
-
"step": 2700
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 1.46,
|
| 202 |
-
"grad_norm": 2.2709996700286865,
|
| 203 |
-
"learning_rate": 0.00016075031944708584,
|
| 204 |
-
"loss": 1.3323,
|
| 205 |
-
"step": 2800
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 1.51,
|
| 209 |
-
"grad_norm": 4.869096279144287,
|
| 210 |
-
"learning_rate": 0.0001581133000835061,
|
| 211 |
-
"loss": 1.2841,
|
| 212 |
-
"step": 2900
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 1.57,
|
| 216 |
-
"grad_norm": 3.3247766494750977,
|
| 217 |
-
"learning_rate": 0.00015541379146515603,
|
| 218 |
-
"loss": 1.2666,
|
| 219 |
-
"step": 3000
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 1.62,
|
| 223 |
-
"grad_norm": 4.0471110343933105,
|
| 224 |
-
"learning_rate": 0.0001526546963747302,
|
| 225 |
-
"loss": 1.2336,
|
| 226 |
-
"step": 3100
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 1.67,
|
| 230 |
-
"grad_norm": 2.8835229873657227,
|
| 231 |
-
"learning_rate": 0.00014986740918973633,
|
| 232 |
-
"loss": 1.2627,
|
| 233 |
-
"step": 3200
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 1.72,
|
| 237 |
-
"grad_norm": 4.162999629974365,
|
| 238 |
-
"learning_rate": 0.00014699862334591993,
|
| 239 |
-
"loss": 1.2683,
|
| 240 |
-
"step": 3300
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 1.77,
|
| 244 |
-
"grad_norm": 2.161064624786377,
|
| 245 |
-
"learning_rate": 0.00014407929986366458,
|
| 246 |
-
"loss": 1.3076,
|
| 247 |
-
"step": 3400
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 1.83,
|
| 251 |
-
"grad_norm": 2.2110209465026855,
|
| 252 |
-
"learning_rate": 0.0001411125778926756,
|
| 253 |
-
"loss": 1.3296,
|
| 254 |
-
"step": 3500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 1.88,
|
| 258 |
-
"grad_norm": 3.9512176513671875,
|
| 259 |
-
"learning_rate": 0.0001381016475502724,
|
| 260 |
-
"loss": 1.2321,
|
| 261 |
-
"step": 3600
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 1.93,
|
| 265 |
-
"grad_norm": 1.0244474411010742,
|
| 266 |
-
"learning_rate": 0.00013504974649105364,
|
| 267 |
-
"loss": 1.1754,
|
| 268 |
-
"step": 3700
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 1.98,
|
| 272 |
-
"grad_norm": 4.0455217361450195,
|
| 273 |
-
"learning_rate": 0.0001319601564254462,
|
| 274 |
-
"loss": 1.2665,
|
| 275 |
-
"step": 3800
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 2.04,
|
| 279 |
-
"grad_norm": 4.162383556365967,
|
| 280 |
-
"learning_rate": 0.00012883619959088054,
|
| 281 |
-
"loss": 1.0379,
|
| 282 |
-
"step": 3900
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 2.09,
|
| 286 |
-
"grad_norm": 2.3285574913024902,
|
| 287 |
-
"learning_rate": 0.0001256812351793875,
|
| 288 |
-
"loss": 0.8172,
|
| 289 |
-
"step": 4000
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 2.14,
|
| 293 |
-
"grad_norm": 5.439870834350586,
|
| 294 |
-
"learning_rate": 0.0001224986557254578,
|
| 295 |
-
"loss": 0.8519,
|
| 296 |
-
"step": 4100
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 2.19,
|
| 300 |
-
"grad_norm": 2.932978391647339,
|
| 301 |
-
"learning_rate": 0.00011929188345804825,
|
| 302 |
-
"loss": 0.8636,
|
| 303 |
-
"step": 4200
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 2.24,
|
| 307 |
-
"grad_norm": 5.512314319610596,
|
| 308 |
-
"learning_rate": 0.00011606436662065767,
|
| 309 |
-
"loss": 0.8727,
|
| 310 |
-
"step": 4300
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 2.3,
|
| 314 |
-
"grad_norm": 4.129833698272705,
|
| 315 |
-
"learning_rate": 0.00011281957576342934,
|
| 316 |
-
"loss": 0.8618,
|
| 317 |
-
"step": 4400
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 2.35,
|
| 321 |
-
"grad_norm": 2.5295193195343018,
|
| 322 |
-
"learning_rate": 0.00010956100001126682,
|
| 323 |
-
"loss": 0.7976,
|
| 324 |
-
"step": 4500
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 2.4,
|
| 328 |
-
"grad_norm": 4.148761749267578,
|
| 329 |
-
"learning_rate": 0.00010629214331197683,
|
| 330 |
-
"loss": 0.7871,
|
| 331 |
-
"step": 4600
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 2.45,
|
| 335 |
-
"grad_norm": 5.00084924697876,
|
| 336 |
-
"learning_rate": 0.00010301652066847249,
|
| 337 |
-
"loss": 0.8557,
|
| 338 |
-
"step": 4700
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 2.51,
|
| 342 |
-
"grad_norm": 5.721644878387451,
|
| 343 |
-
"learning_rate": 9.973765435908962e-05,
|
| 344 |
-
"loss": 0.8898,
|
| 345 |
-
"step": 4800
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 2.56,
|
| 349 |
-
"grad_norm": 5.872501850128174,
|
| 350 |
-
"learning_rate": 9.64590701500791e-05,
|
| 351 |
-
"loss": 0.8269,
|
| 352 |
-
"step": 4900
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 2.61,
|
| 356 |
-
"grad_norm": 5.41200065612793,
|
| 357 |
-
"learning_rate": 9.318429350434922e-05,
|
| 358 |
-
"loss": 0.7502,
|
| 359 |
-
"step": 5000
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 2.66,
|
| 363 |
-
"grad_norm": 3.3696563243865967,
|
| 364 |
-
"learning_rate": 8.991684579053403e-05,
|
| 365 |
-
"loss": 0.8201,
|
| 366 |
-
"step": 5100
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 2.71,
|
| 370 |
-
"grad_norm": 3.6817996501922607,
|
| 371 |
-
"learning_rate": 8.666024049646397e-05,
|
| 372 |
-
"loss": 0.8618,
|
| 373 |
-
"step": 5200
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 2.77,
|
| 377 |
-
"grad_norm": 4.630326747894287,
|
| 378 |
-
"learning_rate": 8.341797945111142e-05,
|
| 379 |
-
"loss": 0.8255,
|
| 380 |
-
"step": 5300
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 2.82,
|
| 384 |
-
"grad_norm": 2.672938823699951,
|
| 385 |
-
"learning_rate": 8.019354905907224e-05,
|
| 386 |
-
"loss": 0.759,
|
| 387 |
-
"step": 5400
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 2.87,
|
| 391 |
-
"grad_norm": 3.0967512130737305,
|
| 392 |
-
"learning_rate": 7.699041655163378e-05,
|
| 393 |
-
"loss": 0.8287,
|
| 394 |
-
"step": 5500
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 2.92,
|
| 398 |
-
"grad_norm": 4.052151679992676,
|
| 399 |
-
"learning_rate": 7.381202625845948e-05,
|
| 400 |
-
"loss": 0.8356,
|
| 401 |
-
"step": 5600
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 2.97,
|
| 405 |
-
"grad_norm": 3.5286691188812256,
|
| 406 |
-
"learning_rate": 7.066179590389994e-05,
|
| 407 |
-
"loss": 0.9181,
|
| 408 |
-
"step": 5700
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 3.03,
|
| 412 |
-
"grad_norm": 4.84535026550293,
|
| 413 |
-
"learning_rate": 6.754311293191257e-05,
|
| 414 |
-
"loss": 0.5811,
|
| 415 |
-
"step": 5800
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 3.08,
|
| 419 |
-
"grad_norm": 2.9033737182617188,
|
| 420 |
-
"learning_rate": 6.44593308635417e-05,
|
| 421 |
-
"loss": 0.4604,
|
| 422 |
-
"step": 5900
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 3.13,
|
| 426 |
-
"grad_norm": 1.6759638786315918,
|
| 427 |
-
"learning_rate": 6.1413765690876e-05,
|
| 428 |
-
"loss": 0.412,
|
| 429 |
-
"step": 6000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 3.18,
|
| 433 |
-
"grad_norm": 3.905642032623291,
|
| 434 |
-
"learning_rate": 5.840969231136102e-05,
|
| 435 |
-
"loss": 0.4478,
|
| 436 |
-
"step": 6100
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 3.24,
|
| 440 |
-
"grad_norm": 7.030938148498535,
|
| 441 |
-
"learning_rate": 5.5450341006300535e-05,
|
| 442 |
-
"loss": 0.3905,
|
| 443 |
-
"step": 6200
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 3.29,
|
| 447 |
-
"grad_norm": 5.1767988204956055,
|
| 448 |
-
"learning_rate": 5.2538893967333866e-05,
|
| 449 |
-
"loss": 0.401,
|
| 450 |
-
"step": 6300
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 3.34,
|
| 454 |
-
"grad_norm": 2.0529091358184814,
|
| 455 |
-
"learning_rate": 4.9678481874623836e-05,
|
| 456 |
-
"loss": 0.417,
|
| 457 |
-
"step": 6400
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 3.39,
|
| 461 |
-
"grad_norm": 4.730635643005371,
|
| 462 |
-
"learning_rate": 4.687218053043516e-05,
|
| 463 |
-
"loss": 0.4213,
|
| 464 |
-
"step": 6500
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 3.44,
|
| 468 |
-
"grad_norm": 5.270570278167725,
|
| 469 |
-
"learning_rate": 4.412300755172314e-05,
|
| 470 |
-
"loss": 0.4371,
|
| 471 |
-
"step": 6600
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 3.5,
|
| 475 |
-
"grad_norm": 5.049856185913086,
|
| 476 |
-
"learning_rate": 4.1433919125288914e-05,
|
| 477 |
-
"loss": 0.4686,
|
| 478 |
-
"step": 6700
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 3.55,
|
| 482 |
-
"grad_norm": 5.543980121612549,
|
| 483 |
-
"learning_rate": 3.8807806828990455e-05,
|
| 484 |
-
"loss": 0.4017,
|
| 485 |
-
"step": 6800
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 3.6,
|
| 489 |
-
"grad_norm": 5.048495292663574,
|
| 490 |
-
"learning_rate": 3.624749452242799e-05,
|
| 491 |
-
"loss": 0.4227,
|
| 492 |
-
"step": 6900
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 3.65,
|
| 496 |
-
"grad_norm": 6.684240818023682,
|
| 497 |
-
"learning_rate": 3.375573531044645e-05,
|
| 498 |
-
"loss": 0.4035,
|
| 499 |
-
"step": 7000
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 3.71,
|
| 503 |
-
"grad_norm": 6.066633701324463,
|
| 504 |
-
"learning_rate": 3.1335208582720856e-05,
|
| 505 |
-
"loss": 0.4271,
|
| 506 |
-
"step": 7100
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 3.76,
|
| 510 |
-
"grad_norm": 4.037696361541748,
|
| 511 |
-
"learning_rate": 2.8988517132607428e-05,
|
| 512 |
-
"loss": 0.4546,
|
| 513 |
-
"step": 7200
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 3.81,
|
| 517 |
-
"grad_norm": 4.5075602531433105,
|
| 518 |
-
"learning_rate": 2.6718184358358956e-05,
|
| 519 |
-
"loss": 0.3988,
|
| 520 |
-
"step": 7300
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 3.86,
|
| 524 |
-
"grad_norm": 2.9646005630493164,
|
| 525 |
-
"learning_rate": 2.4526651549713608e-05,
|
| 526 |
-
"loss": 0.3839,
|
| 527 |
-
"step": 7400
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 3.91,
|
| 531 |
-
"grad_norm": 6.707378387451172,
|
| 532 |
-
"learning_rate": 2.2436969776075456e-05,
|
| 533 |
-
"loss": 0.3956,
|
| 534 |
-
"step": 7500
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 3.97,
|
| 538 |
-
"grad_norm": 6.770478248596191,
|
| 539 |
-
"learning_rate": 2.040917410080746e-05,
|
| 540 |
-
"loss": 0.4,
|
| 541 |
-
"step": 7600
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 4.02,
|
| 545 |
-
"grad_norm": 3.7659125328063965,
|
| 546 |
-
"learning_rate": 1.8466962472390136e-05,
|
| 547 |
-
"loss": 0.3287,
|
| 548 |
-
"step": 7700
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 4.07,
|
| 552 |
-
"grad_norm": 2.905404567718506,
|
| 553 |
-
"learning_rate": 1.661242335176261e-05,
|
| 554 |
-
"loss": 0.2069,
|
| 555 |
-
"step": 7800
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 4.12,
|
| 559 |
-
"grad_norm": 5.363249778747559,
|
| 560 |
-
"learning_rate": 1.4847550925581377e-05,
|
| 561 |
-
"loss": 0.1645,
|
| 562 |
-
"step": 7900
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 4.18,
|
| 566 |
-
"grad_norm": 5.363257884979248,
|
| 567 |
-
"learning_rate": 1.3174242961870542e-05,
|
| 568 |
-
"loss": 0.1888,
|
| 569 |
-
"step": 8000
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 4.23,
|
| 573 |
-
"grad_norm": 2.431536912918091,
|
| 574 |
-
"learning_rate": 1.1594298769351286e-05,
|
| 575 |
-
"loss": 0.1644,
|
| 576 |
-
"step": 8100
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 4.28,
|
| 580 |
-
"grad_norm": 2.4400837421417236,
|
| 581 |
-
"learning_rate": 1.0109417262644261e-05,
|
| 582 |
-
"loss": 0.2169,
|
| 583 |
-
"step": 8200
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 4.33,
|
| 587 |
-
"grad_norm": 8.288840293884277,
|
| 588 |
-
"learning_rate": 8.72119513542623e-06,
|
| 589 |
-
"loss": 0.1992,
|
| 590 |
-
"step": 8300
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 4.38,
|
| 594 |
-
"grad_norm": 3.36771297454834,
|
| 595 |
-
"learning_rate": 7.431125143504525e-06,
|
| 596 |
-
"loss": 0.2011,
|
| 597 |
-
"step": 8400
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 4.44,
|
| 601 |
-
"grad_norm": 0.056750617921352386,
|
| 602 |
-
"learning_rate": 6.240594499656316e-06,
|
| 603 |
-
"loss": 0.1783,
|
| 604 |
-
"step": 8500
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 4.49,
|
| 608 |
-
"grad_norm": 4.8992743492126465,
|
| 609 |
-
"learning_rate": 5.150883381957983e-06,
|
| 610 |
-
"loss": 0.1813,
|
| 611 |
-
"step": 8600
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 4.54,
|
| 615 |
-
"grad_norm": 2.5150370597839355,
|
| 616 |
-
"learning_rate": 4.1631635572092706e-06,
|
| 617 |
-
"loss": 0.2027,
|
| 618 |
-
"step": 8700
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 4.59,
|
| 622 |
-
"grad_norm": 4.545584201812744,
|
| 623 |
-
"learning_rate": 3.2784971209318673e-06,
|
| 624 |
-
"loss": 0.216,
|
| 625 |
-
"step": 8800
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 4.65,
|
| 629 |
-
"grad_norm": 6.020129680633545,
|
| 630 |
-
"learning_rate": 2.4978353552977398e-06,
|
| 631 |
-
"loss": 0.202,
|
| 632 |
-
"step": 8900
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 4.7,
|
| 636 |
-
"grad_norm": 5.093991279602051,
|
| 637 |
-
"learning_rate": 1.822017706215029e-06,
|
| 638 |
-
"loss": 0.1762,
|
| 639 |
-
"step": 9000
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 4.75,
|
| 643 |
-
"grad_norm": 2.0878984928131104,
|
| 644 |
-
"learning_rate": 1.2517708806714657e-06,
|
| 645 |
-
"loss": 0.217,
|
| 646 |
-
"step": 9100
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 4.8,
|
| 650 |
-
"grad_norm": 2.145759105682373,
|
| 651 |
-
"learning_rate": 7.877080653061031e-07,
|
| 652 |
-
"loss": 0.2045,
|
| 653 |
-
"step": 9200
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 4.85,
|
| 657 |
-
"grad_norm": 8.151817321777344,
|
| 658 |
-
"learning_rate": 4.303282670495068e-07,
|
| 659 |
-
"loss": 0.2263,
|
| 660 |
-
"step": 9300
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 4.91,
|
| 664 |
-
"grad_norm": 6.520354270935059,
|
| 665 |
-
"learning_rate": 1.800157765413535e-07,
|
| 666 |
-
"loss": 0.192,
|
| 667 |
-
"step": 9400
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 4.96,
|
| 671 |
-
"grad_norm": 7.271907806396484,
|
| 672 |
-
"learning_rate": 3.703975490257916e-08,
|
| 673 |
-
"loss": 0.2436,
|
| 674 |
-
"step": 9500
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 5.0,
|
| 678 |
-
"step": 9580,
|
| 679 |
-
"total_flos": 1.5336527740560998e+18,
|
| 680 |
-
"train_loss": 0.886494568245396,
|
| 681 |
-
"train_runtime": 33369.3902,
|
| 682 |
-
"train_samples_per_second": 1.148,
|
| 683 |
-
"train_steps_per_second": 0.287
|
| 684 |
-
}
|
| 685 |
-
],
|
| 686 |
-
"logging_steps": 100,
|
| 687 |
-
"max_steps": 9580,
|
| 688 |
-
"num_input_tokens_seen": 0,
|
| 689 |
-
"num_train_epochs": 5,
|
| 690 |
-
"save_steps": 1000,
|
| 691 |
-
"total_flos": 1.5336527740560998e+18,
|
| 692 |
-
"train_batch_size": 4,
|
| 693 |
-
"trial_name": null,
|
| 694 |
-
"trial_params": null
|
| 695 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|