Training in progress, step 500, checkpoint
Browse files
last-checkpoint/optimizer.pt
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 sha256:98fd124bf42aea777816d09325e7ffdff5b493ffa0dd09d7c3a34559fbbc57a1
|
| 3 |
+
size 386379
|
last-checkpoint/pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1540661735
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8db1924741cee0f05554aeca4688871e1e114cabeb2ea4420e919cde46a884c9
|
| 3 |
size 1540661735
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14455
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fb8116194227284009af0f32eb6c9bf39b120912678b54e07fcb6539f29b5b2
|
| 3 |
size 14455
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b3161ebc22e167d90d7ee5cc1bfde367c6af5f0b04db44c912617eb26a593223
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,3528 +2,99 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
-
"eval_steps":
|
| 7 |
"global_step": 500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch": 0.
|
| 14 |
-
"grad_norm":
|
| 15 |
-
"learning_rate": 0.
|
| 16 |
-
"loss":
|
| 17 |
-
"step":
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.004,
|
| 21 |
-
"grad_norm": 1.6919429302215576,
|
| 22 |
-
"learning_rate": 1.6000000000000003e-05,
|
| 23 |
-
"loss": 10.938454627990723,
|
| 24 |
-
"step": 2
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.006,
|
| 28 |
-
"grad_norm": 1.5634195804595947,
|
| 29 |
-
"learning_rate": 3.2000000000000005e-05,
|
| 30 |
-
"loss": 10.937177658081055,
|
| 31 |
-
"step": 3
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008,
|
| 35 |
-
"grad_norm": 1.7329031229019165,
|
| 36 |
-
"learning_rate": 4.8e-05,
|
| 37 |
-
"loss": 10.934802055358887,
|
| 38 |
-
"step": 4
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.01,
|
| 42 |
-
"grad_norm": 1.4461477994918823,
|
| 43 |
-
"learning_rate": 6.400000000000001e-05,
|
| 44 |
-
"loss": 10.912322998046875,
|
| 45 |
-
"step": 5
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.012,
|
| 49 |
-
"grad_norm": 1.4622466564178467,
|
| 50 |
-
"learning_rate": 8e-05,
|
| 51 |
-
"loss": 10.895112037658691,
|
| 52 |
-
"step": 6
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.014,
|
| 56 |
-
"grad_norm": 1.5276347398757935,
|
| 57 |
-
"learning_rate": 9.6e-05,
|
| 58 |
-
"loss": 10.920489311218262,
|
| 59 |
-
"step": 7
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.016,
|
| 63 |
-
"grad_norm": 1.4473192691802979,
|
| 64 |
-
"learning_rate": 0.00011200000000000001,
|
| 65 |
-
"loss": 10.93116283416748,
|
| 66 |
-
"step": 8
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.018,
|
| 70 |
-
"grad_norm": 1.54355788230896,
|
| 71 |
-
"learning_rate": 0.00012800000000000002,
|
| 72 |
-
"loss": 10.912033081054688,
|
| 73 |
-
"step": 9
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.02,
|
| 77 |
-
"grad_norm": 1.6697673797607422,
|
| 78 |
-
"learning_rate": 0.000144,
|
| 79 |
-
"loss": 10.880549430847168,
|
| 80 |
-
"step": 10
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.022,
|
| 84 |
-
"grad_norm": 2.032745122909546,
|
| 85 |
-
"learning_rate": 0.00016,
|
| 86 |
-
"loss": 10.903897285461426,
|
| 87 |
-
"step": 11
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.024,
|
| 91 |
-
"grad_norm": 1.6386597156524658,
|
| 92 |
-
"learning_rate": 0.00017600000000000002,
|
| 93 |
-
"loss": 10.850340843200684,
|
| 94 |
-
"step": 12
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.026,
|
| 98 |
-
"grad_norm": 1.6451835632324219,
|
| 99 |
-
"learning_rate": 0.000192,
|
| 100 |
-
"loss": 10.818197250366211,
|
| 101 |
-
"step": 13
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.028,
|
| 105 |
-
"grad_norm": 1.8559101819992065,
|
| 106 |
-
"learning_rate": 0.00020800000000000001,
|
| 107 |
-
"loss": 10.789449691772461,
|
| 108 |
-
"step": 14
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.03,
|
| 112 |
-
"grad_norm": 1.6030349731445312,
|
| 113 |
-
"learning_rate": 0.00022400000000000002,
|
| 114 |
-
"loss": 10.799813270568848,
|
| 115 |
-
"step": 15
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.032,
|
| 119 |
-
"grad_norm": 1.724048376083374,
|
| 120 |
-
"learning_rate": 0.00024,
|
| 121 |
-
"loss": 10.5698823928833,
|
| 122 |
-
"step": 16
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.034,
|
| 126 |
-
"grad_norm": 1.7329089641571045,
|
| 127 |
-
"learning_rate": 0.00025600000000000004,
|
| 128 |
-
"loss": 10.664511680603027,
|
| 129 |
-
"step": 17
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.036,
|
| 133 |
-
"grad_norm": 1.6604479551315308,
|
| 134 |
-
"learning_rate": 0.00027200000000000005,
|
| 135 |
-
"loss": 10.609597206115723,
|
| 136 |
-
"step": 18
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.038,
|
| 140 |
-
"grad_norm": 1.889926552772522,
|
| 141 |
-
"learning_rate": 0.000288,
|
| 142 |
-
"loss": 10.495404243469238,
|
| 143 |
-
"step": 19
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.04,
|
| 147 |
-
"grad_norm": 1.8469852209091187,
|
| 148 |
-
"learning_rate": 0.000304,
|
| 149 |
-
"loss": 10.616316795349121,
|
| 150 |
-
"step": 20
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.042,
|
| 154 |
-
"grad_norm": 1.7306843996047974,
|
| 155 |
-
"learning_rate": 0.00032,
|
| 156 |
-
"loss": 10.589288711547852,
|
| 157 |
-
"step": 21
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.044,
|
| 161 |
-
"grad_norm": 1.800308346748352,
|
| 162 |
-
"learning_rate": 0.000336,
|
| 163 |
-
"loss": 10.518178939819336,
|
| 164 |
-
"step": 22
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.046,
|
| 168 |
-
"grad_norm": 1.9531307220458984,
|
| 169 |
-
"learning_rate": 0.00035200000000000005,
|
| 170 |
-
"loss": 10.387950897216797,
|
| 171 |
-
"step": 23
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.048,
|
| 175 |
-
"grad_norm": 1.828633189201355,
|
| 176 |
-
"learning_rate": 0.00036800000000000005,
|
| 177 |
-
"loss": 10.544182777404785,
|
| 178 |
-
"step": 24
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.05,
|
| 182 |
-
"grad_norm": 2.3012185096740723,
|
| 183 |
-
"learning_rate": 0.000384,
|
| 184 |
-
"loss": 10.537701606750488,
|
| 185 |
-
"step": 25
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.052,
|
| 189 |
-
"grad_norm": 1.8945454359054565,
|
| 190 |
-
"learning_rate": 0.0004,
|
| 191 |
-
"loss": 10.43015193939209,
|
| 192 |
-
"step": 26
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.054,
|
| 196 |
-
"grad_norm": 1.6789051294326782,
|
| 197 |
-
"learning_rate": 0.000399995625676045,
|
| 198 |
-
"loss": 10.502100944519043,
|
| 199 |
-
"step": 27
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.056,
|
| 203 |
-
"grad_norm": 1.7954329252243042,
|
| 204 |
-
"learning_rate": 0.0003999825028955268,
|
| 205 |
-
"loss": 10.42210865020752,
|
| 206 |
-
"step": 28
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.058,
|
| 210 |
-
"grad_norm": 1.7050436735153198,
|
| 211 |
-
"learning_rate": 0.0003999606322324786,
|
| 212 |
-
"loss": 10.32676887512207,
|
| 213 |
-
"step": 29
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.06,
|
| 217 |
-
"grad_norm": 1.847233533859253,
|
| 218 |
-
"learning_rate": 0.0003999300146435939,
|
| 219 |
-
"loss": 10.175254821777344,
|
| 220 |
-
"step": 30
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.062,
|
| 224 |
-
"grad_norm": 1.7930752038955688,
|
| 225 |
-
"learning_rate": 0.00039989065146818525,
|
| 226 |
-
"loss": 10.2400541305542,
|
| 227 |
-
"step": 31
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.064,
|
| 231 |
-
"grad_norm": 1.6788480281829834,
|
| 232 |
-
"learning_rate": 0.0003998425444281255,
|
| 233 |
-
"loss": 10.1865234375,
|
| 234 |
-
"step": 32
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.066,
|
| 238 |
-
"grad_norm": 2.166168689727783,
|
| 239 |
-
"learning_rate": 0.00039978569562777234,
|
| 240 |
-
"loss": 10.016301155090332,
|
| 241 |
-
"step": 33
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.068,
|
| 245 |
-
"grad_norm": 1.9033219814300537,
|
| 246 |
-
"learning_rate": 0.0003997201075538765,
|
| 247 |
-
"loss": 10.151909828186035,
|
| 248 |
-
"step": 34
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.07,
|
| 252 |
-
"grad_norm": 1.8884942531585693,
|
| 253 |
-
"learning_rate": 0.0003996457830754729,
|
| 254 |
-
"loss": 10.137393951416016,
|
| 255 |
-
"step": 35
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.072,
|
| 259 |
-
"grad_norm": 2.0840940475463867,
|
| 260 |
-
"learning_rate": 0.00039956272544375493,
|
| 261 |
-
"loss": 10.099255561828613,
|
| 262 |
-
"step": 36
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.074,
|
| 266 |
-
"grad_norm": 2.003985643386841,
|
| 267 |
-
"learning_rate": 0.00039947093829193245,
|
| 268 |
-
"loss": 9.89750862121582,
|
| 269 |
-
"step": 37
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.076,
|
| 273 |
-
"grad_norm": 2.0030078887939453,
|
| 274 |
-
"learning_rate": 0.00039937042563507283,
|
| 275 |
-
"loss": 9.982844352722168,
|
| 276 |
-
"step": 38
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.078,
|
| 280 |
-
"grad_norm": 1.91128671169281,
|
| 281 |
-
"learning_rate": 0.00039926119186992537,
|
| 282 |
-
"loss": 9.850743293762207,
|
| 283 |
-
"step": 39
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.08,
|
| 287 |
-
"grad_norm": 2.099236488342285,
|
| 288 |
-
"learning_rate": 0.0003991432417747288,
|
| 289 |
-
"loss": 9.631217002868652,
|
| 290 |
-
"step": 40
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.082,
|
| 294 |
-
"grad_norm": 2.3838231563568115,
|
| 295 |
-
"learning_rate": 0.0003990165805090023,
|
| 296 |
-
"loss": 9.456477165222168,
|
| 297 |
-
"step": 41
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.084,
|
| 301 |
-
"grad_norm": 2.2251834869384766,
|
| 302 |
-
"learning_rate": 0.00039888121361332003,
|
| 303 |
-
"loss": 9.866691589355469,
|
| 304 |
-
"step": 42
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.086,
|
| 308 |
-
"grad_norm": 2.0647263526916504,
|
| 309 |
-
"learning_rate": 0.0003987371470090686,
|
| 310 |
-
"loss": 9.930168151855469,
|
| 311 |
-
"step": 43
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.088,
|
| 315 |
-
"grad_norm": 3.4656786918640137,
|
| 316 |
-
"learning_rate": 0.00039858438699818784,
|
| 317 |
-
"loss": 10.008134841918945,
|
| 318 |
-
"step": 44
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.09,
|
| 322 |
-
"grad_norm": 5.32875394821167,
|
| 323 |
-
"learning_rate": 0.0003984229402628956,
|
| 324 |
-
"loss": 10.024931907653809,
|
| 325 |
-
"step": 45
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.092,
|
| 329 |
-
"grad_norm": 4.914247989654541,
|
| 330 |
-
"learning_rate": 0.00039825281386539503,
|
| 331 |
-
"loss": 9.620004653930664,
|
| 332 |
-
"step": 46
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.094,
|
| 336 |
-
"grad_norm": 4.497557163238525,
|
| 337 |
-
"learning_rate": 0.000398074015247566,
|
| 338 |
-
"loss": 9.497236251831055,
|
| 339 |
-
"step": 47
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.096,
|
| 343 |
-
"grad_norm": 3.8763749599456787,
|
| 344 |
-
"learning_rate": 0.0003978865522306392,
|
| 345 |
-
"loss": 9.334505081176758,
|
| 346 |
-
"step": 48
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.098,
|
| 350 |
-
"grad_norm": 2.562915802001953,
|
| 351 |
-
"learning_rate": 0.0003976904330148543,
|
| 352 |
-
"loss": 9.646927833557129,
|
| 353 |
-
"step": 49
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.1,
|
| 357 |
-
"grad_norm": 2.843885660171509,
|
| 358 |
-
"learning_rate": 0.00039748566617910113,
|
| 359 |
-
"loss": 9.876285552978516,
|
| 360 |
-
"step": 50
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.102,
|
| 364 |
-
"grad_norm": 2.1500260829925537,
|
| 365 |
-
"learning_rate": 0.0003972722606805445,
|
| 366 |
-
"loss": 9.864785194396973,
|
| 367 |
-
"step": 51
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.104,
|
| 371 |
-
"grad_norm": 2.2095565795898438,
|
| 372 |
-
"learning_rate": 0.00039705022585423216,
|
| 373 |
-
"loss": 9.655417442321777,
|
| 374 |
-
"step": 52
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.106,
|
| 378 |
-
"grad_norm": 2.728620767593384,
|
| 379 |
-
"learning_rate": 0.0003968195714126868,
|
| 380 |
-
"loss": 9.74680233001709,
|
| 381 |
-
"step": 53
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.108,
|
| 385 |
-
"grad_norm": 2.514153003692627,
|
| 386 |
-
"learning_rate": 0.00039658030744548075,
|
| 387 |
-
"loss": 9.557250022888184,
|
| 388 |
-
"step": 54
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.11,
|
| 392 |
-
"grad_norm": 2.290548324584961,
|
| 393 |
-
"learning_rate": 0.0003963324444187952,
|
| 394 |
-
"loss": 9.096907615661621,
|
| 395 |
-
"step": 55
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.112,
|
| 399 |
-
"grad_norm": 2.582338809967041,
|
| 400 |
-
"learning_rate": 0.0003960759931749619,
|
| 401 |
-
"loss": 9.14653491973877,
|
| 402 |
-
"step": 56
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.114,
|
| 406 |
-
"grad_norm": 2.1877808570861816,
|
| 407 |
-
"learning_rate": 0.00039581096493198893,
|
| 408 |
-
"loss": 9.360384941101074,
|
| 409 |
-
"step": 57
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.116,
|
| 413 |
-
"grad_norm": 2.369631290435791,
|
| 414 |
-
"learning_rate": 0.0003955373712830703,
|
| 415 |
-
"loss": 9.222617149353027,
|
| 416 |
-
"step": 58
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.118,
|
| 420 |
-
"grad_norm": 2.3022987842559814,
|
| 421 |
-
"learning_rate": 0.00039525522419607854,
|
| 422 |
-
"loss": 9.208744049072266,
|
| 423 |
-
"step": 59
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.12,
|
| 427 |
-
"grad_norm": 2.5634045600891113,
|
| 428 |
-
"learning_rate": 0.0003949645360130412,
|
| 429 |
-
"loss": 8.993451118469238,
|
| 430 |
-
"step": 60
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.122,
|
| 434 |
-
"grad_norm": 2.15937876701355,
|
| 435 |
-
"learning_rate": 0.0003946653194496012,
|
| 436 |
-
"loss": 9.19296646118164,
|
| 437 |
-
"step": 61
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.124,
|
| 441 |
-
"grad_norm": 2.7941417694091797,
|
| 442 |
-
"learning_rate": 0.00039435758759446025,
|
| 443 |
-
"loss": 9.373156547546387,
|
| 444 |
-
"step": 62
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.126,
|
| 448 |
-
"grad_norm": 2.6045541763305664,
|
| 449 |
-
"learning_rate": 0.00039404135390880664,
|
| 450 |
-
"loss": 9.560800552368164,
|
| 451 |
-
"step": 63
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.128,
|
| 455 |
-
"grad_norm": 2.3164615631103516,
|
| 456 |
-
"learning_rate": 0.0003937166322257262,
|
| 457 |
-
"loss": 9.892736434936523,
|
| 458 |
-
"step": 64
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.13,
|
| 462 |
-
"grad_norm": 2.236987590789795,
|
| 463 |
-
"learning_rate": 0.00039338343674959745,
|
| 464 |
-
"loss": 9.644782066345215,
|
| 465 |
-
"step": 65
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.132,
|
| 469 |
-
"grad_norm": 2.5886218547821045,
|
| 470 |
-
"learning_rate": 0.00039304178205546976,
|
| 471 |
-
"loss": 9.261398315429688,
|
| 472 |
-
"step": 66
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.134,
|
| 476 |
-
"grad_norm": 2.18086838722229,
|
| 477 |
-
"learning_rate": 0.00039269168308842634,
|
| 478 |
-
"loss": 9.878418922424316,
|
| 479 |
-
"step": 67
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.136,
|
| 483 |
-
"grad_norm": 2.366741180419922,
|
| 484 |
-
"learning_rate": 0.00039233315516293006,
|
| 485 |
-
"loss": 9.529150009155273,
|
| 486 |
-
"step": 68
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.138,
|
| 490 |
-
"grad_norm": 3.0622975826263428,
|
| 491 |
-
"learning_rate": 0.00039196621396215403,
|
| 492 |
-
"loss": 8.992331504821777,
|
| 493 |
-
"step": 69
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.14,
|
| 497 |
-
"grad_norm": 2.5137059688568115,
|
| 498 |
-
"learning_rate": 0.000391590875537295,
|
| 499 |
-
"loss": 9.463154792785645,
|
| 500 |
-
"step": 70
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.142,
|
| 504 |
-
"grad_norm": 2.401258707046509,
|
| 505 |
-
"learning_rate": 0.00039120715630687155,
|
| 506 |
-
"loss": 9.673027038574219,
|
| 507 |
-
"step": 71
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.144,
|
| 511 |
-
"grad_norm": 2.664299249649048,
|
| 512 |
-
"learning_rate": 0.000390815073056006,
|
| 513 |
-
"loss": 8.944525718688965,
|
| 514 |
-
"step": 72
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.146,
|
| 518 |
-
"grad_norm": 2.288097620010376,
|
| 519 |
-
"learning_rate": 0.00039041464293568983,
|
| 520 |
-
"loss": 8.93644905090332,
|
| 521 |
-
"step": 73
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.148,
|
| 525 |
-
"grad_norm": 2.3277499675750732,
|
| 526 |
-
"learning_rate": 0.00039000588346203374,
|
| 527 |
-
"loss": 9.651083946228027,
|
| 528 |
-
"step": 74
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.15,
|
| 532 |
-
"grad_norm": 2.259471893310547,
|
| 533 |
-
"learning_rate": 0.0003895888125155014,
|
| 534 |
-
"loss": 9.38897705078125,
|
| 535 |
-
"step": 75
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.152,
|
| 539 |
-
"grad_norm": 2.279477834701538,
|
| 540 |
-
"learning_rate": 0.00038916344834012695,
|
| 541 |
-
"loss": 9.122760772705078,
|
| 542 |
-
"step": 76
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.154,
|
| 546 |
-
"grad_norm": 2.9229204654693604,
|
| 547 |
-
"learning_rate": 0.00038872980954271757,
|
| 548 |
-
"loss": 9.61400032043457,
|
| 549 |
-
"step": 77
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.156,
|
| 553 |
-
"grad_norm": 2.203185558319092,
|
| 554 |
-
"learning_rate": 0.00038828791509203895,
|
| 555 |
-
"loss": 9.049810409545898,
|
| 556 |
-
"step": 78
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.158,
|
| 560 |
-
"grad_norm": 2.204108476638794,
|
| 561 |
-
"learning_rate": 0.00038783778431798597,
|
| 562 |
-
"loss": 8.968871116638184,
|
| 563 |
-
"step": 79
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.16,
|
| 567 |
-
"grad_norm": 2.291256904602051,
|
| 568 |
-
"learning_rate": 0.0003873794369107369,
|
| 569 |
-
"loss": 9.418950080871582,
|
| 570 |
-
"step": 80
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.162,
|
| 574 |
-
"grad_norm": 2.3404266834259033,
|
| 575 |
-
"learning_rate": 0.0003869128929198922,
|
| 576 |
-
"loss": 9.489933967590332,
|
| 577 |
-
"step": 81
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.164,
|
| 581 |
-
"grad_norm": 2.5765013694763184,
|
| 582 |
-
"learning_rate": 0.0003864381727535973,
|
| 583 |
-
"loss": 9.591105461120605,
|
| 584 |
-
"step": 82
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.166,
|
| 588 |
-
"grad_norm": 2.9069576263427734,
|
| 589 |
-
"learning_rate": 0.00038595529717765027,
|
| 590 |
-
"loss": 10.357954978942871,
|
| 591 |
-
"step": 83
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.168,
|
| 595 |
-
"grad_norm": 2.1355342864990234,
|
| 596 |
-
"learning_rate": 0.0003854642873145931,
|
| 597 |
-
"loss": 9.309526443481445,
|
| 598 |
-
"step": 84
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.17,
|
| 602 |
-
"grad_norm": 2.4784109592437744,
|
| 603 |
-
"learning_rate": 0.00038496516464278776,
|
| 604 |
-
"loss": 9.193565368652344,
|
| 605 |
-
"step": 85
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.172,
|
| 609 |
-
"grad_norm": 2.4962222576141357,
|
| 610 |
-
"learning_rate": 0.00038445795099547697,
|
| 611 |
-
"loss": 9.196398735046387,
|
| 612 |
-
"step": 86
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.174,
|
| 616 |
-
"grad_norm": 2.187791585922241,
|
| 617 |
-
"learning_rate": 0.0003839426685598287,
|
| 618 |
-
"loss": 9.315688133239746,
|
| 619 |
-
"step": 87
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.176,
|
| 623 |
-
"grad_norm": 2.1691067218780518,
|
| 624 |
-
"learning_rate": 0.000383419339875966,
|
| 625 |
-
"loss": 9.588054656982422,
|
| 626 |
-
"step": 88
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.178,
|
| 630 |
-
"grad_norm": 2.293802261352539,
|
| 631 |
-
"learning_rate": 0.00038288798783598087,
|
| 632 |
-
"loss": 9.176922798156738,
|
| 633 |
-
"step": 89
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.18,
|
| 637 |
-
"grad_norm": 2.548900842666626,
|
| 638 |
-
"learning_rate": 0.0003823486356829329,
|
| 639 |
-
"loss": 8.861899375915527,
|
| 640 |
-
"step": 90
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.182,
|
| 644 |
-
"grad_norm": 2.685309648513794,
|
| 645 |
-
"learning_rate": 0.0003818013070098325,
|
| 646 |
-
"loss": 9.092292785644531,
|
| 647 |
-
"step": 91
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.184,
|
| 651 |
-
"grad_norm": 2.1843101978302,
|
| 652 |
-
"learning_rate": 0.0003812460257586089,
|
| 653 |
-
"loss": 9.847696304321289,
|
| 654 |
-
"step": 92
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.186,
|
| 658 |
-
"grad_norm": 2.262296199798584,
|
| 659 |
-
"learning_rate": 0.000380682816219063,
|
| 660 |
-
"loss": 9.236616134643555,
|
| 661 |
-
"step": 93
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.188,
|
| 665 |
-
"grad_norm": 2.5018444061279297,
|
| 666 |
-
"learning_rate": 0.00038011170302780446,
|
| 667 |
-
"loss": 9.138856887817383,
|
| 668 |
-
"step": 94
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.19,
|
| 672 |
-
"grad_norm": 2.5166268348693848,
|
| 673 |
-
"learning_rate": 0.00037953271116717444,
|
| 674 |
-
"loss": 9.022369384765625,
|
| 675 |
-
"step": 95
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.192,
|
| 679 |
-
"grad_norm": 2.3215460777282715,
|
| 680 |
-
"learning_rate": 0.0003789458659641527,
|
| 681 |
-
"loss": 8.769148826599121,
|
| 682 |
-
"step": 96
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 0.194,
|
| 686 |
-
"grad_norm": 2.047666072845459,
|
| 687 |
-
"learning_rate": 0.0003783511930892495,
|
| 688 |
-
"loss": 9.138738632202148,
|
| 689 |
-
"step": 97
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 0.196,
|
| 693 |
-
"grad_norm": 2.1690986156463623,
|
| 694 |
-
"learning_rate": 0.00037774871855538275,
|
| 695 |
-
"loss": 9.50068187713623,
|
| 696 |
-
"step": 98
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 0.198,
|
| 700 |
-
"grad_norm": 2.152500867843628,
|
| 701 |
-
"learning_rate": 0.00037713846871674045,
|
| 702 |
-
"loss": 9.252716064453125,
|
| 703 |
-
"step": 99
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 0.2,
|
| 707 |
-
"grad_norm": 2.470381498336792,
|
| 708 |
-
"learning_rate": 0.0003765204702676274,
|
| 709 |
-
"loss": 9.554462432861328,
|
| 710 |
-
"step": 100
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.202,
|
| 714 |
-
"grad_norm": 2.3412070274353027,
|
| 715 |
-
"learning_rate": 0.0003758947502412978,
|
| 716 |
-
"loss": 9.199671745300293,
|
| 717 |
-
"step": 101
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.204,
|
| 721 |
-
"grad_norm": 2.3880722522735596,
|
| 722 |
-
"learning_rate": 0.0003752613360087727,
|
| 723 |
-
"loss": 9.079456329345703,
|
| 724 |
-
"step": 102
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 0.206,
|
| 728 |
-
"grad_norm": 2.360715389251709,
|
| 729 |
-
"learning_rate": 0.00037462025527764265,
|
| 730 |
-
"loss": 8.943620681762695,
|
| 731 |
-
"step": 103
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 0.208,
|
| 735 |
-
"grad_norm": 2.2601282596588135,
|
| 736 |
-
"learning_rate": 0.00037397153609085553,
|
| 737 |
-
"loss": 8.811064720153809,
|
| 738 |
-
"step": 104
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 0.21,
|
| 742 |
-
"grad_norm": 2.937379837036133,
|
| 743 |
-
"learning_rate": 0.0003733152068254901,
|
| 744 |
-
"loss": 8.938572883605957,
|
| 745 |
-
"step": 105
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.212,
|
| 749 |
-
"grad_norm": 2.2584245204925537,
|
| 750 |
-
"learning_rate": 0.00037265129619151483,
|
| 751 |
-
"loss": 8.90784740447998,
|
| 752 |
-
"step": 106
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 0.214,
|
| 756 |
-
"grad_norm": 2.4078965187072754,
|
| 757 |
-
"learning_rate": 0.00037197983323053143,
|
| 758 |
-
"loss": 8.889952659606934,
|
| 759 |
-
"step": 107
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.216,
|
| 763 |
-
"grad_norm": 2.513396978378296,
|
| 764 |
-
"learning_rate": 0.00037130084731450515,
|
| 765 |
-
"loss": 9.039155006408691,
|
| 766 |
-
"step": 108
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.218,
|
| 770 |
-
"grad_norm": 2.323331356048584,
|
| 771 |
-
"learning_rate": 0.0003706143681444795,
|
| 772 |
-
"loss": 8.843852043151855,
|
| 773 |
-
"step": 109
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.22,
|
| 777 |
-
"grad_norm": 2.4571354389190674,
|
| 778 |
-
"learning_rate": 0.0003699204257492774,
|
| 779 |
-
"loss": 8.732458114624023,
|
| 780 |
-
"step": 110
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 0.222,
|
| 784 |
-
"grad_norm": 2.1594672203063965,
|
| 785 |
-
"learning_rate": 0.0003692190504841871,
|
| 786 |
-
"loss": 8.942313194274902,
|
| 787 |
-
"step": 111
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.224,
|
| 791 |
-
"grad_norm": 2.1859757900238037,
|
| 792 |
-
"learning_rate": 0.00036851027302963493,
|
| 793 |
-
"loss": 8.837929725646973,
|
| 794 |
-
"step": 112
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 0.226,
|
| 798 |
-
"grad_norm": 2.2536919116973877,
|
| 799 |
-
"learning_rate": 0.00036779412438984294,
|
| 800 |
-
"loss": 8.81443977355957,
|
| 801 |
-
"step": 113
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 0.228,
|
| 805 |
-
"grad_norm": 2.30411958694458,
|
| 806 |
-
"learning_rate": 0.0003670706358914725,
|
| 807 |
-
"loss": 8.665769577026367,
|
| 808 |
-
"step": 114
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"epoch": 0.23,
|
| 812 |
-
"grad_norm": 2.203463315963745,
|
| 813 |
-
"learning_rate": 0.0003663398391822543,
|
| 814 |
-
"loss": 8.481475830078125,
|
| 815 |
-
"step": 115
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 0.232,
|
| 819 |
-
"grad_norm": 2.2214648723602295,
|
| 820 |
-
"learning_rate": 0.00036560176622960403,
|
| 821 |
-
"loss": 8.336684226989746,
|
| 822 |
-
"step": 116
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.234,
|
| 826 |
-
"grad_norm": 2.953378677368164,
|
| 827 |
-
"learning_rate": 0.00036485644931922353,
|
| 828 |
-
"loss": 8.88039493560791,
|
| 829 |
-
"step": 117
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 0.236,
|
| 833 |
-
"grad_norm": 2.736071825027466,
|
| 834 |
-
"learning_rate": 0.0003641039210536889,
|
| 835 |
-
"loss": 9.217086791992188,
|
| 836 |
-
"step": 118
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 0.238,
|
| 840 |
-
"grad_norm": 2.491441249847412,
|
| 841 |
-
"learning_rate": 0.0003633442143510245,
|
| 842 |
-
"loss": 9.156209945678711,
|
| 843 |
-
"step": 119
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 0.24,
|
| 847 |
-
"grad_norm": 2.2380411624908447,
|
| 848 |
-
"learning_rate": 0.00036257736244326246,
|
| 849 |
-
"loss": 8.874693870544434,
|
| 850 |
-
"step": 120
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 0.242,
|
| 854 |
-
"grad_norm": 2.2316958904266357,
|
| 855 |
-
"learning_rate": 0.0003618033988749895,
|
| 856 |
-
"loss": 9.17328929901123,
|
| 857 |
-
"step": 121
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 0.244,
|
| 861 |
-
"grad_norm": 2.2098398208618164,
|
| 862 |
-
"learning_rate": 0.0003610223575018795,
|
| 863 |
-
"loss": 8.89964771270752,
|
| 864 |
-
"step": 122
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 0.246,
|
| 868 |
-
"grad_norm": 2.1878061294555664,
|
| 869 |
-
"learning_rate": 0.00036023427248921215,
|
| 870 |
-
"loss": 9.016963005065918,
|
| 871 |
-
"step": 123
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.248,
|
| 875 |
-
"grad_norm": 2.713280200958252,
|
| 876 |
-
"learning_rate": 0.0003594391783103792,
|
| 877 |
-
"loss": 9.224175453186035,
|
| 878 |
-
"step": 124
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 0.25,
|
| 882 |
-
"grad_norm": 2.5514464378356934,
|
| 883 |
-
"learning_rate": 0.00035863710974537563,
|
| 884 |
-
"loss": 8.910709381103516,
|
| 885 |
-
"step": 125
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 0.252,
|
| 889 |
-
"grad_norm": 2.7256574630737305,
|
| 890 |
-
"learning_rate": 0.00035782810187927875,
|
| 891 |
-
"loss": 8.939787864685059,
|
| 892 |
-
"step": 126
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 0.254,
|
| 896 |
-
"grad_norm": 2.1270618438720703,
|
| 897 |
-
"learning_rate": 0.0003570121901007136,
|
| 898 |
-
"loss": 9.144500732421875,
|
| 899 |
-
"step": 127
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 0.256,
|
| 903 |
-
"grad_norm": 2.2637150287628174,
|
| 904 |
-
"learning_rate": 0.0003561894101003044,
|
| 905 |
-
"loss": 8.90536880493164,
|
| 906 |
-
"step": 128
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 0.258,
|
| 910 |
-
"grad_norm": 2.6754376888275146,
|
| 911 |
-
"learning_rate": 0.00035535979786911396,
|
| 912 |
-
"loss": 8.82153606414795,
|
| 913 |
-
"step": 129
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 0.26,
|
| 917 |
-
"grad_norm": 2.232227087020874,
|
| 918 |
-
"learning_rate": 0.00035452338969706876,
|
| 919 |
-
"loss": 9.127659797668457,
|
| 920 |
-
"step": 130
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.262,
|
| 924 |
-
"grad_norm": 2.1317057609558105,
|
| 925 |
-
"learning_rate": 0.00035368022217137184,
|
| 926 |
-
"loss": 8.978923797607422,
|
| 927 |
-
"step": 131
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.264,
|
| 931 |
-
"grad_norm": 2.4739177227020264,
|
| 932 |
-
"learning_rate": 0.00035283033217490227,
|
| 933 |
-
"loss": 8.894933700561523,
|
| 934 |
-
"step": 132
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 0.266,
|
| 938 |
-
"grad_norm": 2.2630388736724854,
|
| 939 |
-
"learning_rate": 0.00035197375688460176,
|
| 940 |
-
"loss": 8.912936210632324,
|
| 941 |
-
"step": 133
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 0.268,
|
| 945 |
-
"grad_norm": 2.420799732208252,
|
| 946 |
-
"learning_rate": 0.0003511105337698484,
|
| 947 |
-
"loss": 8.535635948181152,
|
| 948 |
-
"step": 134
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 0.27,
|
| 952 |
-
"grad_norm": 2.297384262084961,
|
| 953 |
-
"learning_rate": 0.0003502407005908177,
|
| 954 |
-
"loss": 8.684527397155762,
|
| 955 |
-
"step": 135
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 0.272,
|
| 959 |
-
"grad_norm": 2.3078620433807373,
|
| 960 |
-
"learning_rate": 0.0003493642953968308,
|
| 961 |
-
"loss": 8.654274940490723,
|
| 962 |
-
"step": 136
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 0.274,
|
| 966 |
-
"grad_norm": 2.411306858062744,
|
| 967 |
-
"learning_rate": 0.00034848135652469,
|
| 968 |
-
"loss": 8.775450706481934,
|
| 969 |
-
"step": 137
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 0.276,
|
| 973 |
-
"grad_norm": 2.4742720127105713,
|
| 974 |
-
"learning_rate": 0.00034759192259700196,
|
| 975 |
-
"loss": 8.752528190612793,
|
| 976 |
-
"step": 138
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 0.278,
|
| 980 |
-
"grad_norm": 2.3106977939605713,
|
| 981 |
-
"learning_rate": 0.000346696032520488,
|
| 982 |
-
"loss": 9.044780731201172,
|
| 983 |
-
"step": 139
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.28,
|
| 987 |
-
"grad_norm": 2.25669264793396,
|
| 988 |
-
"learning_rate": 0.00034579372548428235,
|
| 989 |
-
"loss": 8.638384819030762,
|
| 990 |
-
"step": 140
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 0.282,
|
| 994 |
-
"grad_norm": 2.569659471511841,
|
| 995 |
-
"learning_rate": 0.00034488504095821784,
|
| 996 |
-
"loss": 8.72479248046875,
|
| 997 |
-
"step": 141
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 0.284,
|
| 1001 |
-
"grad_norm": 2.371952772140503,
|
| 1002 |
-
"learning_rate": 0.0003439700186910993,
|
| 1003 |
-
"loss": 8.62860107421875,
|
| 1004 |
-
"step": 142
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 0.286,
|
| 1008 |
-
"grad_norm": 2.845393657684326,
|
| 1009 |
-
"learning_rate": 0.00034304869870896513,
|
| 1010 |
-
"loss": 8.54886531829834,
|
| 1011 |
-
"step": 143
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 0.288,
|
| 1015 |
-
"grad_norm": 2.3225581645965576,
|
| 1016 |
-
"learning_rate": 0.00034212112131333587,
|
| 1017 |
-
"loss": 8.782685279846191,
|
| 1018 |
-
"step": 144
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 0.29,
|
| 1022 |
-
"grad_norm": 3.1657943725585938,
|
| 1023 |
-
"learning_rate": 0.0003411873270794518,
|
| 1024 |
-
"loss": 8.93557357788086,
|
| 1025 |
-
"step": 145
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 0.292,
|
| 1029 |
-
"grad_norm": 2.854094982147217,
|
| 1030 |
-
"learning_rate": 0.00034024735685449773,
|
| 1031 |
-
"loss": 9.170632362365723,
|
| 1032 |
-
"step": 146
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.294,
|
| 1036 |
-
"grad_norm": 4.226919651031494,
|
| 1037 |
-
"learning_rate": 0.00033930125175581647,
|
| 1038 |
-
"loss": 8.83270263671875,
|
| 1039 |
-
"step": 147
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 0.296,
|
| 1043 |
-
"grad_norm": 2.5496034622192383,
|
| 1044 |
-
"learning_rate": 0.0003383490531691099,
|
| 1045 |
-
"loss": 9.363445281982422,
|
| 1046 |
-
"step": 148
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 0.298,
|
| 1050 |
-
"grad_norm": 2.2917191982269287,
|
| 1051 |
-
"learning_rate": 0.0003373908027466289,
|
| 1052 |
-
"loss": 10.128934860229492,
|
| 1053 |
-
"step": 149
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 0.3,
|
| 1057 |
-
"grad_norm": 1.8361351490020752,
|
| 1058 |
-
"learning_rate": 0.00033642654240535134,
|
| 1059 |
-
"loss": 10.009841918945312,
|
| 1060 |
-
"step": 150
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 0.302,
|
| 1064 |
-
"grad_norm": 2.388427734375,
|
| 1065 |
-
"learning_rate": 0.00033545631432514825,
|
| 1066 |
-
"loss": 9.989909172058105,
|
| 1067 |
-
"step": 151
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 0.304,
|
| 1071 |
-
"grad_norm": 2.2267374992370605,
|
| 1072 |
-
"learning_rate": 0.00033448016094693895,
|
| 1073 |
-
"loss": 8.923678398132324,
|
| 1074 |
-
"step": 152
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.306,
|
| 1078 |
-
"grad_norm": 1.80934739112854,
|
| 1079 |
-
"learning_rate": 0.0003334981249708345,
|
| 1080 |
-
"loss": 8.479927062988281,
|
| 1081 |
-
"step": 153
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 0.308,
|
| 1085 |
-
"grad_norm": 2.195918560028076,
|
| 1086 |
-
"learning_rate": 0.00033251024935427,
|
| 1087 |
-
"loss": 8.912820816040039,
|
| 1088 |
-
"step": 154
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.31,
|
| 1092 |
-
"grad_norm": 1.9725319147109985,
|
| 1093 |
-
"learning_rate": 0.0003315165773101249,
|
| 1094 |
-
"loss": 8.949131965637207,
|
| 1095 |
-
"step": 155
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 0.312,
|
| 1099 |
-
"grad_norm": 2.4333691596984863,
|
| 1100 |
-
"learning_rate": 0.00033051715230483374,
|
| 1101 |
-
"loss": 9.121983528137207,
|
| 1102 |
-
"step": 156
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 0.314,
|
| 1106 |
-
"grad_norm": 2.921964406967163,
|
| 1107 |
-
"learning_rate": 0.0003295120180564838,
|
| 1108 |
-
"loss": 9.040019989013672,
|
| 1109 |
-
"step": 157
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 0.316,
|
| 1113 |
-
"grad_norm": 2.979886770248413,
|
| 1114 |
-
"learning_rate": 0.00032850121853290334,
|
| 1115 |
-
"loss": 8.351471900939941,
|
| 1116 |
-
"step": 158
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 0.318,
|
| 1120 |
-
"grad_norm": 2.5775766372680664,
|
| 1121 |
-
"learning_rate": 0.000327484797949738,
|
| 1122 |
-
"loss": 8.370685577392578,
|
| 1123 |
-
"step": 159
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 0.32,
|
| 1127 |
-
"grad_norm": 2.3292393684387207,
|
| 1128 |
-
"learning_rate": 0.00032646280076851684,
|
| 1129 |
-
"loss": 9.214851379394531,
|
| 1130 |
-
"step": 160
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.322,
|
| 1134 |
-
"grad_norm": 2.1481590270996094,
|
| 1135 |
-
"learning_rate": 0.0003254352716947074,
|
| 1136 |
-
"loss": 9.306310653686523,
|
| 1137 |
-
"step": 161
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.324,
|
| 1141 |
-
"grad_norm": 2.6631641387939453,
|
| 1142 |
-
"learning_rate": 0.0003244022556757602,
|
| 1143 |
-
"loss": 8.955159187316895,
|
| 1144 |
-
"step": 162
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.326,
|
| 1148 |
-
"grad_norm": 2.16222882270813,
|
| 1149 |
-
"learning_rate": 0.0003233637978991422,
|
| 1150 |
-
"loss": 9.106553077697754,
|
| 1151 |
-
"step": 163
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.328,
|
| 1155 |
-
"grad_norm": 2.115537166595459,
|
| 1156 |
-
"learning_rate": 0.00032231994379036086,
|
| 1157 |
-
"loss": 9.03366470336914,
|
| 1158 |
-
"step": 164
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.33,
|
| 1162 |
-
"grad_norm": 2.3213510513305664,
|
| 1163 |
-
"learning_rate": 0.0003212707390109765,
|
| 1164 |
-
"loss": 8.890856742858887,
|
| 1165 |
-
"step": 165
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.332,
|
| 1169 |
-
"grad_norm": 2.095850944519043,
|
| 1170 |
-
"learning_rate": 0.00032021622945660504,
|
| 1171 |
-
"loss": 9.074614524841309,
|
| 1172 |
-
"step": 166
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.334,
|
| 1176 |
-
"grad_norm": 2.2268483638763428,
|
| 1177 |
-
"learning_rate": 0.0003191564612549106,
|
| 1178 |
-
"loss": 9.12753963470459,
|
| 1179 |
-
"step": 167
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.336,
|
| 1183 |
-
"grad_norm": 2.3109869956970215,
|
| 1184 |
-
"learning_rate": 0.0003180914807635874,
|
| 1185 |
-
"loss": 8.67584228515625,
|
| 1186 |
-
"step": 168
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.338,
|
| 1190 |
-
"grad_norm": 2.303342580795288,
|
| 1191 |
-
"learning_rate": 0.00031702133456833236,
|
| 1192 |
-
"loss": 8.74753189086914,
|
| 1193 |
-
"step": 169
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.34,
|
| 1197 |
-
"grad_norm": 2.3582310676574707,
|
| 1198 |
-
"learning_rate": 0.00031594606948080663,
|
| 1199 |
-
"loss": 8.63679313659668,
|
| 1200 |
-
"step": 170
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 0.342,
|
| 1204 |
-
"grad_norm": 2.2999212741851807,
|
| 1205 |
-
"learning_rate": 0.00031486573253658874,
|
| 1206 |
-
"loss": 8.993260383605957,
|
| 1207 |
-
"step": 171
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 0.344,
|
| 1211 |
-
"grad_norm": 2.059751033782959,
|
| 1212 |
-
"learning_rate": 0.00031378037099311627,
|
| 1213 |
-
"loss": 8.988717079162598,
|
| 1214 |
-
"step": 172
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.346,
|
| 1218 |
-
"grad_norm": 2.098874568939209,
|
| 1219 |
-
"learning_rate": 0.00031269003232761933,
|
| 1220 |
-
"loss": 9.038498878479004,
|
| 1221 |
-
"step": 173
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 0.348,
|
| 1225 |
-
"grad_norm": 2.051058292388916,
|
| 1226 |
-
"learning_rate": 0.0003115947642350433,
|
| 1227 |
-
"loss": 8.815178871154785,
|
| 1228 |
-
"step": 174
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 0.35,
|
| 1232 |
-
"grad_norm": 1.9609602689743042,
|
| 1233 |
-
"learning_rate": 0.00031049461462596267,
|
| 1234 |
-
"loss": 9.118207931518555,
|
| 1235 |
-
"step": 175
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 0.352,
|
| 1239 |
-
"grad_norm": 2.215872287750244,
|
| 1240 |
-
"learning_rate": 0.00030938963162448544,
|
| 1241 |
-
"loss": 9.338445663452148,
|
| 1242 |
-
"step": 176
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 0.354,
|
| 1246 |
-
"grad_norm": 2.4297356605529785,
|
| 1247 |
-
"learning_rate": 0.0003082798635661476,
|
| 1248 |
-
"loss": 10.221427917480469,
|
| 1249 |
-
"step": 177
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 0.356,
|
| 1253 |
-
"grad_norm": 1.8971245288848877,
|
| 1254 |
-
"learning_rate": 0.0003071653589957993,
|
| 1255 |
-
"loss": 9.243640899658203,
|
| 1256 |
-
"step": 178
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 0.358,
|
| 1260 |
-
"grad_norm": 2.039586067199707,
|
| 1261 |
-
"learning_rate": 0.000306046166665481,
|
| 1262 |
-
"loss": 9.404755592346191,
|
| 1263 |
-
"step": 179
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 0.36,
|
| 1267 |
-
"grad_norm": 2.120884418487549,
|
| 1268 |
-
"learning_rate": 0.00030492233553229076,
|
| 1269 |
-
"loss": 9.160200119018555,
|
| 1270 |
-
"step": 180
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 0.362,
|
| 1274 |
-
"grad_norm": 2.1105194091796875,
|
| 1275 |
-
"learning_rate": 0.00030379391475624304,
|
| 1276 |
-
"loss": 8.598636627197266,
|
| 1277 |
-
"step": 181
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 0.364,
|
| 1281 |
-
"grad_norm": 2.1490981578826904,
|
| 1282 |
-
"learning_rate": 0.0003026609536981183,
|
| 1283 |
-
"loss": 8.572158813476562,
|
| 1284 |
-
"step": 182
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 0.366,
|
| 1288 |
-
"grad_norm": 2.309242010116577,
|
| 1289 |
-
"learning_rate": 0.0003015235019173034,
|
| 1290 |
-
"loss": 8.578638076782227,
|
| 1291 |
-
"step": 183
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 0.368,
|
| 1295 |
-
"grad_norm": 2.3543789386749268,
|
| 1296 |
-
"learning_rate": 0.00030038160916962404,
|
| 1297 |
-
"loss": 8.930187225341797,
|
| 1298 |
-
"step": 184
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 0.37,
|
| 1302 |
-
"grad_norm": 2.687485933303833,
|
| 1303 |
-
"learning_rate": 0.00029923532540516843,
|
| 1304 |
-
"loss": 8.654874801635742,
|
| 1305 |
-
"step": 185
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 0.372,
|
| 1309 |
-
"grad_norm": 2.1869587898254395,
|
| 1310 |
-
"learning_rate": 0.00029808470076610167,
|
| 1311 |
-
"loss": 8.435524940490723,
|
| 1312 |
-
"step": 186
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 0.374,
|
| 1316 |
-
"grad_norm": 2.1635005474090576,
|
| 1317 |
-
"learning_rate": 0.00029692978558447305,
|
| 1318 |
-
"loss": 9.194635391235352,
|
| 1319 |
-
"step": 187
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 0.376,
|
| 1323 |
-
"grad_norm": 1.98250150680542,
|
| 1324 |
-
"learning_rate": 0.0002957706303800139,
|
| 1325 |
-
"loss": 9.433363914489746,
|
| 1326 |
-
"step": 188
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 0.378,
|
| 1330 |
-
"grad_norm": 1.8844033479690552,
|
| 1331 |
-
"learning_rate": 0.0002946072858579282,
|
| 1332 |
-
"loss": 8.961788177490234,
|
| 1333 |
-
"step": 189
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 0.38,
|
| 1337 |
-
"grad_norm": 2.6384198665618896,
|
| 1338 |
-
"learning_rate": 0.0002934398029066739,
|
| 1339 |
-
"loss": 8.384697914123535,
|
| 1340 |
-
"step": 190
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 0.382,
|
| 1344 |
-
"grad_norm": 2.100079298019409,
|
| 1345 |
-
"learning_rate": 0.0002922682325957376,
|
| 1346 |
-
"loss": 8.51193618774414,
|
| 1347 |
-
"step": 191
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 0.384,
|
| 1351 |
-
"grad_norm": 1.9416857957839966,
|
| 1352 |
-
"learning_rate": 0.00029109262617339987,
|
| 1353 |
-
"loss": 9.050522804260254,
|
| 1354 |
-
"step": 192
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 0.386,
|
| 1358 |
-
"grad_norm": 1.92392897605896,
|
| 1359 |
-
"learning_rate": 0.0002899130350644941,
|
| 1360 |
-
"loss": 9.127717971801758,
|
| 1361 |
-
"step": 193
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 0.388,
|
| 1365 |
-
"grad_norm": 1.9244396686553955,
|
| 1366 |
-
"learning_rate": 0.00028872951086815685,
|
| 1367 |
-
"loss": 8.932775497436523,
|
| 1368 |
-
"step": 194
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 0.39,
|
| 1372 |
-
"grad_norm": 2.156482219696045,
|
| 1373 |
-
"learning_rate": 0.00028754210535557036,
|
| 1374 |
-
"loss": 8.989794731140137,
|
| 1375 |
-
"step": 195
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 0.392,
|
| 1379 |
-
"grad_norm": 2.0522656440734863,
|
| 1380 |
-
"learning_rate": 0.00028635087046769857,
|
| 1381 |
-
"loss": 9.00429630279541,
|
| 1382 |
-
"step": 196
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 0.394,
|
| 1386 |
-
"grad_norm": 1.985840916633606,
|
| 1387 |
-
"learning_rate": 0.00028515585831301456,
|
| 1388 |
-
"loss": 8.986638069152832,
|
| 1389 |
-
"step": 197
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 0.396,
|
| 1393 |
-
"grad_norm": 2.1469404697418213,
|
| 1394 |
-
"learning_rate": 0.0002839571211652212,
|
| 1395 |
-
"loss": 8.83475112915039,
|
| 1396 |
-
"step": 198
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 0.398,
|
| 1400 |
-
"grad_norm": 2.973970890045166,
|
| 1401 |
-
"learning_rate": 0.00028275471146096466,
|
| 1402 |
-
"loss": 8.771541595458984,
|
| 1403 |
-
"step": 199
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 0.4,
|
| 1407 |
-
"grad_norm": 2.049677848815918,
|
| 1408 |
-
"learning_rate": 0.00028154868179754074,
|
| 1409 |
-
"loss": 9.055174827575684,
|
| 1410 |
-
"step": 200
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 0.402,
|
| 1414 |
-
"grad_norm": 2.141853094100952,
|
| 1415 |
-
"learning_rate": 0.0002803390849305939,
|
| 1416 |
-
"loss": 9.262734413146973,
|
| 1417 |
-
"step": 201
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 0.404,
|
| 1421 |
-
"grad_norm": 2.5649197101593018,
|
| 1422 |
-
"learning_rate": 0.0002791259737718097,
|
| 1423 |
-
"loss": 8.846776962280273,
|
| 1424 |
-
"step": 202
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 0.406,
|
| 1428 |
-
"grad_norm": 2.330873727798462,
|
| 1429 |
-
"learning_rate": 0.0002779094013866001,
|
| 1430 |
-
"loss": 8.51368522644043,
|
| 1431 |
-
"step": 203
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 0.408,
|
| 1435 |
-
"grad_norm": 2.319931983947754,
|
| 1436 |
-
"learning_rate": 0.00027668942099178234,
|
| 1437 |
-
"loss": 8.636953353881836,
|
| 1438 |
-
"step": 204
|
| 1439 |
-
},
|
| 1440 |
-
{
|
| 1441 |
-
"epoch": 0.41,
|
| 1442 |
-
"grad_norm": 2.190286159515381,
|
| 1443 |
-
"learning_rate": 0.00027546608595325117,
|
| 1444 |
-
"loss": 8.872692108154297,
|
| 1445 |
-
"step": 205
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 0.412,
|
| 1449 |
-
"grad_norm": 2.1084277629852295,
|
| 1450 |
-
"learning_rate": 0.00027423944978364416,
|
| 1451 |
-
"loss": 8.60301399230957,
|
| 1452 |
-
"step": 206
|
| 1453 |
-
},
|
| 1454 |
-
{
|
| 1455 |
-
"epoch": 0.414,
|
| 1456 |
-
"grad_norm": 2.131725788116455,
|
| 1457 |
-
"learning_rate": 0.00027300956614000115,
|
| 1458 |
-
"loss": 8.761425018310547,
|
| 1459 |
-
"step": 207
|
| 1460 |
-
},
|
| 1461 |
-
{
|
| 1462 |
-
"epoch": 0.416,
|
| 1463 |
-
"grad_norm": 2.091268301010132,
|
| 1464 |
-
"learning_rate": 0.00027177648882141704,
|
| 1465 |
-
"loss": 8.531584739685059,
|
| 1466 |
-
"step": 208
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 0.418,
|
| 1470 |
-
"grad_norm": 3.2136242389678955,
|
| 1471 |
-
"learning_rate": 0.0002705402717666883,
|
| 1472 |
-
"loss": 9.191469192504883,
|
| 1473 |
-
"step": 209
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 0.42,
|
| 1477 |
-
"grad_norm": 2.223203420639038,
|
| 1478 |
-
"learning_rate": 0.00026930096905195363,
|
| 1479 |
-
"loss": 8.968942642211914,
|
| 1480 |
-
"step": 210
|
| 1481 |
-
},
|
| 1482 |
-
{
|
| 1483 |
-
"epoch": 0.422,
|
| 1484 |
-
"grad_norm": 2.1849513053894043,
|
| 1485 |
-
"learning_rate": 0.00026805863488832865,
|
| 1486 |
-
"loss": 8.568798065185547,
|
| 1487 |
-
"step": 211
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 0.424,
|
| 1491 |
-
"grad_norm": 2.341390609741211,
|
| 1492 |
-
"learning_rate": 0.00026681332361953424,
|
| 1493 |
-
"loss": 8.58562183380127,
|
| 1494 |
-
"step": 212
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 0.426,
|
| 1498 |
-
"grad_norm": 2.2670180797576904,
|
| 1499 |
-
"learning_rate": 0.0002655650897195195,
|
| 1500 |
-
"loss": 8.619218826293945,
|
| 1501 |
-
"step": 213
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 0.428,
|
| 1505 |
-
"grad_norm": 2.0588316917419434,
|
| 1506 |
-
"learning_rate": 0.0002643139877900791,
|
| 1507 |
-
"loss": 8.445850372314453,
|
| 1508 |
-
"step": 214
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 0.43,
|
| 1512 |
-
"grad_norm": 2.206193208694458,
|
| 1513 |
-
"learning_rate": 0.00026306007255846436,
|
| 1514 |
-
"loss": 8.351473808288574,
|
| 1515 |
-
"step": 215
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 0.432,
|
| 1519 |
-
"grad_norm": 2.011192798614502,
|
| 1520 |
-
"learning_rate": 0.00026180339887498953,
|
| 1521 |
-
"loss": 8.77007007598877,
|
| 1522 |
-
"step": 216
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 0.434,
|
| 1526 |
-
"grad_norm": 2.5190346240997314,
|
| 1527 |
-
"learning_rate": 0.00026054402171063267,
|
| 1528 |
-
"loss": 8.483141899108887,
|
| 1529 |
-
"step": 217
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 0.436,
|
| 1533 |
-
"grad_norm": 1.955390214920044,
|
| 1534 |
-
"learning_rate": 0.0002592819961546308,
|
| 1535 |
-
"loss": 8.786367416381836,
|
| 1536 |
-
"step": 218
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 0.438,
|
| 1540 |
-
"grad_norm": 2.0526981353759766,
|
| 1541 |
-
"learning_rate": 0.00025801737741207005,
|
| 1542 |
-
"loss": 8.833505630493164,
|
| 1543 |
-
"step": 219
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 0.44,
|
| 1547 |
-
"grad_norm": 2.5714111328125,
|
| 1548 |
-
"learning_rate": 0.000256750220801471,
|
| 1549 |
-
"loss": 8.71342945098877,
|
| 1550 |
-
"step": 220
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 0.442,
|
| 1554 |
-
"grad_norm": 2.1315982341766357,
|
| 1555 |
-
"learning_rate": 0.0002554805817523689,
|
| 1556 |
-
"loss": 9.144265174865723,
|
| 1557 |
-
"step": 221
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 0.444,
|
| 1561 |
-
"grad_norm": 2.163686990737915,
|
| 1562 |
-
"learning_rate": 0.0002542085158028889,
|
| 1563 |
-
"loss": 8.4553804397583,
|
| 1564 |
-
"step": 222
|
| 1565 |
-
},
|
| 1566 |
-
{
|
| 1567 |
-
"epoch": 0.446,
|
| 1568 |
-
"grad_norm": 2.1909923553466797,
|
| 1569 |
-
"learning_rate": 0.00025293407859731633,
|
| 1570 |
-
"loss": 8.719898223876953,
|
| 1571 |
-
"step": 223
|
| 1572 |
-
},
|
| 1573 |
-
{
|
| 1574 |
-
"epoch": 0.448,
|
| 1575 |
-
"grad_norm": 2.1163313388824463,
|
| 1576 |
-
"learning_rate": 0.00025165732588366334,
|
| 1577 |
-
"loss": 8.743754386901855,
|
| 1578 |
-
"step": 224
|
| 1579 |
-
},
|
| 1580 |
-
{
|
| 1581 |
-
"epoch": 0.45,
|
| 1582 |
-
"grad_norm": 2.163228750228882,
|
| 1583 |
-
"learning_rate": 0.00025037831351122967,
|
| 1584 |
-
"loss": 8.655962944030762,
|
| 1585 |
-
"step": 225
|
| 1586 |
-
},
|
| 1587 |
-
{
|
| 1588 |
-
"epoch": 0.452,
|
| 1589 |
-
"grad_norm": 2.136554718017578,
|
| 1590 |
-
"learning_rate": 0.0002490970974281599,
|
| 1591 |
-
"loss": 8.541149139404297,
|
| 1592 |
-
"step": 226
|
| 1593 |
-
},
|
| 1594 |
-
{
|
| 1595 |
-
"epoch": 0.454,
|
| 1596 |
-
"grad_norm": 2.1031856536865234,
|
| 1597 |
-
"learning_rate": 0.00024781373367899597,
|
| 1598 |
-
"loss": 8.4634428024292,
|
| 1599 |
-
"step": 227
|
| 1600 |
-
},
|
| 1601 |
-
{
|
| 1602 |
-
"epoch": 0.456,
|
| 1603 |
-
"grad_norm": 2.2057831287384033,
|
| 1604 |
-
"learning_rate": 0.00024652827840222606,
|
| 1605 |
-
"loss": 8.664609909057617,
|
| 1606 |
-
"step": 228
|
| 1607 |
-
},
|
| 1608 |
-
{
|
| 1609 |
-
"epoch": 0.458,
|
| 1610 |
-
"grad_norm": 2.016057014465332,
|
| 1611 |
-
"learning_rate": 0.00024524078782782807,
|
| 1612 |
-
"loss": 8.324894905090332,
|
| 1613 |
-
"step": 229
|
| 1614 |
-
},
|
| 1615 |
-
{
|
| 1616 |
-
"epoch": 0.46,
|
| 1617 |
-
"grad_norm": 2.0009148120880127,
|
| 1618 |
-
"learning_rate": 0.00024395131827481062,
|
| 1619 |
-
"loss": 9.179727554321289,
|
| 1620 |
-
"step": 230
|
| 1621 |
-
},
|
| 1622 |
-
{
|
| 1623 |
-
"epoch": 0.462,
|
| 1624 |
-
"grad_norm": 1.954477310180664,
|
| 1625 |
-
"learning_rate": 0.0002426599261487494,
|
| 1626 |
-
"loss": 8.644003868103027,
|
| 1627 |
-
"step": 231
|
| 1628 |
-
},
|
| 1629 |
-
{
|
| 1630 |
-
"epoch": 0.464,
|
| 1631 |
-
"grad_norm": 2.198920249938965,
|
| 1632 |
-
"learning_rate": 0.00024136666793931935,
|
| 1633 |
-
"loss": 8.694100379943848,
|
| 1634 |
-
"step": 232
|
| 1635 |
-
},
|
| 1636 |
-
{
|
| 1637 |
-
"epoch": 0.466,
|
| 1638 |
-
"grad_norm": 2.0328779220581055,
|
| 1639 |
-
"learning_rate": 0.00024007160021782427,
|
| 1640 |
-
"loss": 8.79814624786377,
|
| 1641 |
-
"step": 233
|
| 1642 |
-
},
|
| 1643 |
-
{
|
| 1644 |
-
"epoch": 0.468,
|
| 1645 |
-
"grad_norm": 2.0648465156555176,
|
| 1646 |
-
"learning_rate": 0.0002387747796347217,
|
| 1647 |
-
"loss": 8.87083911895752,
|
| 1648 |
-
"step": 234
|
| 1649 |
-
},
|
| 1650 |
-
{
|
| 1651 |
-
"epoch": 0.47,
|
| 1652 |
-
"grad_norm": 2.0518767833709717,
|
| 1653 |
-
"learning_rate": 0.00023747626291714498,
|
| 1654 |
-
"loss": 9.036115646362305,
|
| 1655 |
-
"step": 235
|
| 1656 |
-
},
|
| 1657 |
-
{
|
| 1658 |
-
"epoch": 0.472,
|
| 1659 |
-
"grad_norm": 2.0349044799804688,
|
| 1660 |
-
"learning_rate": 0.000236176106866422,
|
| 1661 |
-
"loss": 8.789681434631348,
|
| 1662 |
-
"step": 236
|
| 1663 |
-
},
|
| 1664 |
-
{
|
| 1665 |
-
"epoch": 0.474,
|
| 1666 |
-
"grad_norm": 2.528453826904297,
|
| 1667 |
-
"learning_rate": 0.00023487436835559035,
|
| 1668 |
-
"loss": 8.664607048034668,
|
| 1669 |
-
"step": 237
|
| 1670 |
-
},
|
| 1671 |
-
{
|
| 1672 |
-
"epoch": 0.476,
|
| 1673 |
-
"grad_norm": 1.9791979789733887,
|
| 1674 |
-
"learning_rate": 0.00023357110432690954,
|
| 1675 |
-
"loss": 8.811439514160156,
|
| 1676 |
-
"step": 238
|
| 1677 |
-
},
|
| 1678 |
-
{
|
| 1679 |
-
"epoch": 0.478,
|
| 1680 |
-
"grad_norm": 1.9924676418304443,
|
| 1681 |
-
"learning_rate": 0.00023226637178937022,
|
| 1682 |
-
"loss": 8.76052188873291,
|
| 1683 |
-
"step": 239
|
| 1684 |
-
},
|
| 1685 |
-
{
|
| 1686 |
-
"epoch": 0.48,
|
| 1687 |
-
"grad_norm": 2.254997968673706,
|
| 1688 |
-
"learning_rate": 0.00023096022781620034,
|
| 1689 |
-
"loss": 9.341845512390137,
|
| 1690 |
-
"step": 240
|
| 1691 |
-
},
|
| 1692 |
-
{
|
| 1693 |
-
"epoch": 0.482,
|
| 1694 |
-
"grad_norm": 2.0505363941192627,
|
| 1695 |
-
"learning_rate": 0.0002296527295423684,
|
| 1696 |
-
"loss": 8.936700820922852,
|
| 1697 |
-
"step": 241
|
| 1698 |
-
},
|
| 1699 |
-
{
|
| 1700 |
-
"epoch": 0.484,
|
| 1701 |
-
"grad_norm": 2.202488422393799,
|
| 1702 |
-
"learning_rate": 0.00022834393416208486,
|
| 1703 |
-
"loss": 8.593229293823242,
|
| 1704 |
-
"step": 242
|
| 1705 |
-
},
|
| 1706 |
-
{
|
| 1707 |
-
"epoch": 0.486,
|
| 1708 |
-
"grad_norm": 2.2660021781921387,
|
| 1709 |
-
"learning_rate": 0.0002270338989262994,
|
| 1710 |
-
"loss": 8.762115478515625,
|
| 1711 |
-
"step": 243
|
| 1712 |
-
},
|
| 1713 |
-
{
|
| 1714 |
-
"epoch": 0.488,
|
| 1715 |
-
"grad_norm": 2.243227481842041,
|
| 1716 |
-
"learning_rate": 0.00022572268114019726,
|
| 1717 |
-
"loss": 8.501476287841797,
|
| 1718 |
-
"step": 244
|
| 1719 |
-
},
|
| 1720 |
-
{
|
| 1721 |
-
"epoch": 0.49,
|
| 1722 |
-
"grad_norm": 2.0970733165740967,
|
| 1723 |
-
"learning_rate": 0.00022441033816069202,
|
| 1724 |
-
"loss": 8.231463432312012,
|
| 1725 |
-
"step": 245
|
| 1726 |
-
},
|
| 1727 |
-
{
|
| 1728 |
-
"epoch": 0.492,
|
| 1729 |
-
"grad_norm": 2.0461721420288086,
|
| 1730 |
-
"learning_rate": 0.00022309692739391727,
|
| 1731 |
-
"loss": 8.305780410766602,
|
| 1732 |
-
"step": 246
|
| 1733 |
-
},
|
| 1734 |
-
{
|
| 1735 |
-
"epoch": 0.494,
|
| 1736 |
-
"grad_norm": 2.3393101692199707,
|
| 1737 |
-
"learning_rate": 0.00022178250629271452,
|
| 1738 |
-
"loss": 8.002923011779785,
|
| 1739 |
-
"step": 247
|
| 1740 |
-
},
|
| 1741 |
-
{
|
| 1742 |
-
"epoch": 0.496,
|
| 1743 |
-
"grad_norm": 1.9498422145843506,
|
| 1744 |
-
"learning_rate": 0.00022046713235412103,
|
| 1745 |
-
"loss": 8.305809020996094,
|
| 1746 |
-
"step": 248
|
| 1747 |
-
},
|
| 1748 |
-
{
|
| 1749 |
-
"epoch": 0.498,
|
| 1750 |
-
"grad_norm": 1.9709067344665527,
|
| 1751 |
-
"learning_rate": 0.00021915086311685404,
|
| 1752 |
-
"loss": 8.464375495910645,
|
| 1753 |
-
"step": 249
|
| 1754 |
-
},
|
| 1755 |
-
{
|
| 1756 |
-
"epoch": 0.5,
|
| 1757 |
-
"grad_norm": 2.030928373336792,
|
| 1758 |
-
"learning_rate": 0.00021783375615879415,
|
| 1759 |
-
"loss": 8.108865737915039,
|
| 1760 |
-
"step": 250
|
| 1761 |
-
},
|
| 1762 |
-
{
|
| 1763 |
-
"epoch": 0.502,
|
| 1764 |
-
"grad_norm": 1.876850962638855,
|
| 1765 |
-
"learning_rate": 0.0002165158690944665,
|
| 1766 |
-
"loss": 8.170710563659668,
|
| 1767 |
-
"step": 251
|
| 1768 |
-
},
|
| 1769 |
-
{
|
| 1770 |
-
"epoch": 0.504,
|
| 1771 |
-
"grad_norm": 1.9732722043991089,
|
| 1772 |
-
"learning_rate": 0.00021519725957252063,
|
| 1773 |
-
"loss": 8.505722045898438,
|
| 1774 |
-
"step": 252
|
| 1775 |
-
},
|
| 1776 |
-
{
|
| 1777 |
-
"epoch": 0.506,
|
| 1778 |
-
"grad_norm": 2.0012569427490234,
|
| 1779 |
-
"learning_rate": 0.00021387798527320882,
|
| 1780 |
-
"loss": 8.21572494506836,
|
| 1781 |
-
"step": 253
|
| 1782 |
-
},
|
| 1783 |
-
{
|
| 1784 |
-
"epoch": 0.508,
|
| 1785 |
-
"grad_norm": 1.8451972007751465,
|
| 1786 |
-
"learning_rate": 0.0002125581039058627,
|
| 1787 |
-
"loss": 8.303668022155762,
|
| 1788 |
-
"step": 254
|
| 1789 |
-
},
|
| 1790 |
-
{
|
| 1791 |
-
"epoch": 0.51,
|
| 1792 |
-
"grad_norm": 1.9150594472885132,
|
| 1793 |
-
"learning_rate": 0.0002112376732063691,
|
| 1794 |
-
"loss": 8.306048393249512,
|
| 1795 |
-
"step": 255
|
| 1796 |
-
},
|
| 1797 |
-
{
|
| 1798 |
-
"epoch": 0.512,
|
| 1799 |
-
"grad_norm": 2.209169864654541,
|
| 1800 |
-
"learning_rate": 0.00020991675093464448,
|
| 1801 |
-
"loss": 7.939265727996826,
|
| 1802 |
-
"step": 256
|
| 1803 |
-
},
|
| 1804 |
-
{
|
| 1805 |
-
"epoch": 0.514,
|
| 1806 |
-
"grad_norm": 2.7872672080993652,
|
| 1807 |
-
"learning_rate": 0.00020859539487210813,
|
| 1808 |
-
"loss": 8.67353343963623,
|
| 1809 |
-
"step": 257
|
| 1810 |
-
},
|
| 1811 |
-
{
|
| 1812 |
-
"epoch": 0.516,
|
| 1813 |
-
"grad_norm": 2.120419979095459,
|
| 1814 |
-
"learning_rate": 0.0002072736628191549,
|
| 1815 |
-
"loss": 8.481735229492188,
|
| 1816 |
-
"step": 258
|
| 1817 |
-
},
|
| 1818 |
-
{
|
| 1819 |
-
"epoch": 0.518,
|
| 1820 |
-
"grad_norm": 2.115609884262085,
|
| 1821 |
-
"learning_rate": 0.0002059516125926265,
|
| 1822 |
-
"loss": 8.521812438964844,
|
| 1823 |
-
"step": 259
|
| 1824 |
-
},
|
| 1825 |
-
{
|
| 1826 |
-
"epoch": 0.52,
|
| 1827 |
-
"grad_norm": 2.0753695964813232,
|
| 1828 |
-
"learning_rate": 0.00020462930202328278,
|
| 1829 |
-
"loss": 8.759242057800293,
|
| 1830 |
-
"step": 260
|
| 1831 |
-
},
|
| 1832 |
-
{
|
| 1833 |
-
"epoch": 0.522,
|
| 1834 |
-
"grad_norm": 2.149508237838745,
|
| 1835 |
-
"learning_rate": 0.00020330678895327174,
|
| 1836 |
-
"loss": 8.718214988708496,
|
| 1837 |
-
"step": 261
|
| 1838 |
-
},
|
| 1839 |
-
{
|
| 1840 |
-
"epoch": 0.524,
|
| 1841 |
-
"grad_norm": 2.004070997238159,
|
| 1842 |
-
"learning_rate": 0.00020198413123359926,
|
| 1843 |
-
"loss": 8.574090003967285,
|
| 1844 |
-
"step": 262
|
| 1845 |
-
},
|
| 1846 |
-
{
|
| 1847 |
-
"epoch": 0.526,
|
| 1848 |
-
"grad_norm": 2.2298269271850586,
|
| 1849 |
-
"learning_rate": 0.00020066138672159903,
|
| 1850 |
-
"loss": 8.622062683105469,
|
| 1851 |
-
"step": 263
|
| 1852 |
-
},
|
| 1853 |
-
{
|
| 1854 |
-
"epoch": 0.528,
|
| 1855 |
-
"grad_norm": 2.1374661922454834,
|
| 1856 |
-
"learning_rate": 0.00019933861327840098,
|
| 1857 |
-
"loss": 8.323304176330566,
|
| 1858 |
-
"step": 264
|
| 1859 |
-
},
|
| 1860 |
-
{
|
| 1861 |
-
"epoch": 0.53,
|
| 1862 |
-
"grad_norm": 1.9791207313537598,
|
| 1863 |
-
"learning_rate": 0.00019801586876640073,
|
| 1864 |
-
"loss": 8.170138359069824,
|
| 1865 |
-
"step": 265
|
| 1866 |
-
},
|
| 1867 |
-
{
|
| 1868 |
-
"epoch": 0.532,
|
| 1869 |
-
"grad_norm": 2.0265350341796875,
|
| 1870 |
-
"learning_rate": 0.0001966932110467283,
|
| 1871 |
-
"loss": 8.584282875061035,
|
| 1872 |
-
"step": 266
|
| 1873 |
-
},
|
| 1874 |
-
{
|
| 1875 |
-
"epoch": 0.534,
|
| 1876 |
-
"grad_norm": 2.100043535232544,
|
| 1877 |
-
"learning_rate": 0.00019537069797671724,
|
| 1878 |
-
"loss": 8.808910369873047,
|
| 1879 |
-
"step": 267
|
| 1880 |
-
},
|
| 1881 |
-
{
|
| 1882 |
-
"epoch": 0.536,
|
| 1883 |
-
"grad_norm": 2.1547322273254395,
|
| 1884 |
-
"learning_rate": 0.0001940483874073735,
|
| 1885 |
-
"loss": 8.47503662109375,
|
| 1886 |
-
"step": 268
|
| 1887 |
-
},
|
| 1888 |
-
{
|
| 1889 |
-
"epoch": 0.538,
|
| 1890 |
-
"grad_norm": 2.0436670780181885,
|
| 1891 |
-
"learning_rate": 0.00019272633718084517,
|
| 1892 |
-
"loss": 8.857638359069824,
|
| 1893 |
-
"step": 269
|
| 1894 |
-
},
|
| 1895 |
-
{
|
| 1896 |
-
"epoch": 0.54,
|
| 1897 |
-
"grad_norm": 2.5172436237335205,
|
| 1898 |
-
"learning_rate": 0.0001914046051278919,
|
| 1899 |
-
"loss": 8.935922622680664,
|
| 1900 |
-
"step": 270
|
| 1901 |
-
},
|
| 1902 |
-
{
|
| 1903 |
-
"epoch": 0.542,
|
| 1904 |
-
"grad_norm": 2.363839864730835,
|
| 1905 |
-
"learning_rate": 0.00019008324906535554,
|
| 1906 |
-
"loss": 8.704509735107422,
|
| 1907 |
-
"step": 271
|
| 1908 |
-
},
|
| 1909 |
-
{
|
| 1910 |
-
"epoch": 0.544,
|
| 1911 |
-
"grad_norm": 2.258408546447754,
|
| 1912 |
-
"learning_rate": 0.0001887623267936309,
|
| 1913 |
-
"loss": 8.630513191223145,
|
| 1914 |
-
"step": 272
|
| 1915 |
-
},
|
| 1916 |
-
{
|
| 1917 |
-
"epoch": 0.546,
|
| 1918 |
-
"grad_norm": 2.2246406078338623,
|
| 1919 |
-
"learning_rate": 0.00018744189609413734,
|
| 1920 |
-
"loss": 8.515497207641602,
|
| 1921 |
-
"step": 273
|
| 1922 |
-
},
|
| 1923 |
-
{
|
| 1924 |
-
"epoch": 0.548,
|
| 1925 |
-
"grad_norm": 2.1859161853790283,
|
| 1926 |
-
"learning_rate": 0.0001861220147267912,
|
| 1927 |
-
"loss": 8.569636344909668,
|
| 1928 |
-
"step": 274
|
| 1929 |
-
},
|
| 1930 |
-
{
|
| 1931 |
-
"epoch": 0.55,
|
| 1932 |
-
"grad_norm": 2.306729555130005,
|
| 1933 |
-
"learning_rate": 0.0001848027404274794,
|
| 1934 |
-
"loss": 8.568589210510254,
|
| 1935 |
-
"step": 275
|
| 1936 |
-
},
|
| 1937 |
-
{
|
| 1938 |
-
"epoch": 0.552,
|
| 1939 |
-
"grad_norm": 3.5251359939575195,
|
| 1940 |
-
"learning_rate": 0.00018348413090553354,
|
| 1941 |
-
"loss": 8.712056159973145,
|
| 1942 |
-
"step": 276
|
| 1943 |
-
},
|
| 1944 |
-
{
|
| 1945 |
-
"epoch": 0.554,
|
| 1946 |
-
"grad_norm": 2.266549587249756,
|
| 1947 |
-
"learning_rate": 0.00018216624384120595,
|
| 1948 |
-
"loss": 8.456377029418945,
|
| 1949 |
-
"step": 277
|
| 1950 |
-
},
|
| 1951 |
-
{
|
| 1952 |
-
"epoch": 0.556,
|
| 1953 |
-
"grad_norm": 2.2839274406433105,
|
| 1954 |
-
"learning_rate": 0.00018084913688314597,
|
| 1955 |
-
"loss": 8.64017105102539,
|
| 1956 |
-
"step": 278
|
| 1957 |
-
},
|
| 1958 |
-
{
|
| 1959 |
-
"epoch": 0.558,
|
| 1960 |
-
"grad_norm": 2.145022392272949,
|
| 1961 |
-
"learning_rate": 0.000179532867645879,
|
| 1962 |
-
"loss": 8.920992851257324,
|
| 1963 |
-
"step": 279
|
| 1964 |
-
},
|
| 1965 |
-
{
|
| 1966 |
-
"epoch": 0.56,
|
| 1967 |
-
"grad_norm": 2.3669328689575195,
|
| 1968 |
-
"learning_rate": 0.0001782174937072855,
|
| 1969 |
-
"loss": 8.96733570098877,
|
| 1970 |
-
"step": 280
|
| 1971 |
-
},
|
| 1972 |
-
{
|
| 1973 |
-
"epoch": 0.562,
|
| 1974 |
-
"grad_norm": 2.46120548248291,
|
| 1975 |
-
"learning_rate": 0.00017690307260608278,
|
| 1976 |
-
"loss": 8.669891357421875,
|
| 1977 |
-
"step": 281
|
| 1978 |
-
},
|
| 1979 |
-
{
|
| 1980 |
-
"epoch": 0.564,
|
| 1981 |
-
"grad_norm": 1.99955415725708,
|
| 1982 |
-
"learning_rate": 0.000175589661839308,
|
| 1983 |
-
"loss": 8.476048469543457,
|
| 1984 |
-
"step": 282
|
| 1985 |
-
},
|
| 1986 |
-
{
|
| 1987 |
-
"epoch": 0.566,
|
| 1988 |
-
"grad_norm": 2.3627235889434814,
|
| 1989 |
-
"learning_rate": 0.00017427731885980282,
|
| 1990 |
-
"loss": 8.750870704650879,
|
| 1991 |
-
"step": 283
|
| 1992 |
-
},
|
| 1993 |
-
{
|
| 1994 |
-
"epoch": 0.568,
|
| 1995 |
-
"grad_norm": 2.3135578632354736,
|
| 1996 |
-
"learning_rate": 0.0001729661010737007,
|
| 1997 |
-
"loss": 8.49306583404541,
|
| 1998 |
-
"step": 284
|
| 1999 |
-
},
|
| 2000 |
-
{
|
| 2001 |
-
"epoch": 0.57,
|
| 2002 |
-
"grad_norm": 2.4493324756622314,
|
| 2003 |
-
"learning_rate": 0.00017165606583791515,
|
| 2004 |
-
"loss": 9.049016952514648,
|
| 2005 |
-
"step": 285
|
| 2006 |
-
},
|
| 2007 |
-
{
|
| 2008 |
-
"epoch": 0.572,
|
| 2009 |
-
"grad_norm": 2.399137258529663,
|
| 2010 |
-
"learning_rate": 0.00017034727045763158,
|
| 2011 |
-
"loss": 9.218151092529297,
|
| 2012 |
-
"step": 286
|
| 2013 |
-
},
|
| 2014 |
-
{
|
| 2015 |
-
"epoch": 0.574,
|
| 2016 |
-
"grad_norm": 3.1089603900909424,
|
| 2017 |
-
"learning_rate": 0.00016903977218379974,
|
| 2018 |
-
"loss": 8.707843780517578,
|
| 2019 |
-
"step": 287
|
| 2020 |
-
},
|
| 2021 |
-
{
|
| 2022 |
-
"epoch": 0.576,
|
| 2023 |
-
"grad_norm": 2.3197054862976074,
|
| 2024 |
-
"learning_rate": 0.00016773362821062983,
|
| 2025 |
-
"loss": 8.859163284301758,
|
| 2026 |
-
"step": 288
|
| 2027 |
-
},
|
| 2028 |
-
{
|
| 2029 |
-
"epoch": 0.578,
|
| 2030 |
-
"grad_norm": 2.16019344329834,
|
| 2031 |
-
"learning_rate": 0.00016642889567309048,
|
| 2032 |
-
"loss": 8.65565013885498,
|
| 2033 |
-
"step": 289
|
| 2034 |
-
},
|
| 2035 |
-
{
|
| 2036 |
-
"epoch": 0.58,
|
| 2037 |
-
"grad_norm": 2.4339723587036133,
|
| 2038 |
-
"learning_rate": 0.0001651256316444097,
|
| 2039 |
-
"loss": 8.664863586425781,
|
| 2040 |
-
"step": 290
|
| 2041 |
-
},
|
| 2042 |
-
{
|
| 2043 |
-
"epoch": 0.582,
|
| 2044 |
-
"grad_norm": 2.289949417114258,
|
| 2045 |
-
"learning_rate": 0.0001638238931335781,
|
| 2046 |
-
"loss": 8.4475736618042,
|
| 2047 |
-
"step": 291
|
| 2048 |
-
},
|
| 2049 |
-
{
|
| 2050 |
-
"epoch": 0.584,
|
| 2051 |
-
"grad_norm": 2.2416627407073975,
|
| 2052 |
-
"learning_rate": 0.00016252373708285504,
|
| 2053 |
-
"loss": 8.673498153686523,
|
| 2054 |
-
"step": 292
|
| 2055 |
-
},
|
| 2056 |
-
{
|
| 2057 |
-
"epoch": 0.586,
|
| 2058 |
-
"grad_norm": 2.304426908493042,
|
| 2059 |
-
"learning_rate": 0.00016122522036527838,
|
| 2060 |
-
"loss": 8.329156875610352,
|
| 2061 |
-
"step": 293
|
| 2062 |
-
},
|
| 2063 |
-
{
|
| 2064 |
-
"epoch": 0.588,
|
| 2065 |
-
"grad_norm": 2.325382947921753,
|
| 2066 |
-
"learning_rate": 0.00015992839978217578,
|
| 2067 |
-
"loss": 8.132695198059082,
|
| 2068 |
-
"step": 294
|
| 2069 |
-
},
|
| 2070 |
-
{
|
| 2071 |
-
"epoch": 0.59,
|
| 2072 |
-
"grad_norm": 2.322619676589966,
|
| 2073 |
-
"learning_rate": 0.00015863333206068067,
|
| 2074 |
-
"loss": 8.164306640625,
|
| 2075 |
-
"step": 295
|
| 2076 |
-
},
|
| 2077 |
-
{
|
| 2078 |
-
"epoch": 0.592,
|
| 2079 |
-
"grad_norm": 5.08868408203125,
|
| 2080 |
-
"learning_rate": 0.00015734007385125067,
|
| 2081 |
-
"loss": 7.9473876953125,
|
| 2082 |
-
"step": 296
|
| 2083 |
-
},
|
| 2084 |
-
{
|
| 2085 |
-
"epoch": 0.594,
|
| 2086 |
-
"grad_norm": 2.907891035079956,
|
| 2087 |
-
"learning_rate": 0.0001560486817251894,
|
| 2088 |
-
"loss": 8.277129173278809,
|
| 2089 |
-
"step": 297
|
| 2090 |
-
},
|
| 2091 |
-
{
|
| 2092 |
-
"epoch": 0.596,
|
| 2093 |
-
"grad_norm": 2.378244638442993,
|
| 2094 |
-
"learning_rate": 0.000154759212172172,
|
| 2095 |
-
"loss": 9.207996368408203,
|
| 2096 |
-
"step": 298
|
| 2097 |
-
},
|
| 2098 |
-
{
|
| 2099 |
-
"epoch": 0.598,
|
| 2100 |
-
"grad_norm": 2.2020370960235596,
|
| 2101 |
-
"learning_rate": 0.00015347172159777396,
|
| 2102 |
-
"loss": 8.603676795959473,
|
| 2103 |
-
"step": 299
|
| 2104 |
-
},
|
| 2105 |
-
{
|
| 2106 |
-
"epoch": 0.6,
|
| 2107 |
-
"grad_norm": 2.5474672317504883,
|
| 2108 |
-
"learning_rate": 0.000152186266321004,
|
| 2109 |
-
"loss": 8.99532413482666,
|
| 2110 |
-
"step": 300
|
| 2111 |
-
},
|
| 2112 |
-
{
|
| 2113 |
-
"epoch": 0.602,
|
| 2114 |
-
"grad_norm": 2.336972713470459,
|
| 2115 |
-
"learning_rate": 0.0001509029025718402,
|
| 2116 |
-
"loss": 8.95500659942627,
|
| 2117 |
-
"step": 301
|
| 2118 |
-
},
|
| 2119 |
-
{
|
| 2120 |
-
"epoch": 0.604,
|
| 2121 |
-
"grad_norm": 3.2240536212921143,
|
| 2122 |
-
"learning_rate": 0.0001496216864887704,
|
| 2123 |
-
"loss": 8.39035415649414,
|
| 2124 |
-
"step": 302
|
| 2125 |
-
},
|
| 2126 |
-
{
|
| 2127 |
-
"epoch": 0.606,
|
| 2128 |
-
"grad_norm": 2.964778423309326,
|
| 2129 |
-
"learning_rate": 0.00014834267411633674,
|
| 2130 |
-
"loss": 8.482704162597656,
|
| 2131 |
-
"step": 303
|
| 2132 |
-
},
|
| 2133 |
-
{
|
| 2134 |
-
"epoch": 0.608,
|
| 2135 |
-
"grad_norm": 2.2110812664031982,
|
| 2136 |
-
"learning_rate": 0.0001470659214026837,
|
| 2137 |
-
"loss": 8.503761291503906,
|
| 2138 |
-
"step": 304
|
| 2139 |
-
},
|
| 2140 |
-
{
|
| 2141 |
-
"epoch": 0.61,
|
| 2142 |
-
"grad_norm": 2.7591614723205566,
|
| 2143 |
-
"learning_rate": 0.00014579148419711119,
|
| 2144 |
-
"loss": 8.46738052368164,
|
| 2145 |
-
"step": 305
|
| 2146 |
-
},
|
| 2147 |
-
{
|
| 2148 |
-
"epoch": 0.612,
|
| 2149 |
-
"grad_norm": 2.773846387863159,
|
| 2150 |
-
"learning_rate": 0.00014451941824763113,
|
| 2151 |
-
"loss": 8.264008522033691,
|
| 2152 |
-
"step": 306
|
| 2153 |
-
},
|
| 2154 |
-
{
|
| 2155 |
-
"epoch": 0.614,
|
| 2156 |
-
"grad_norm": 2.6419637203216553,
|
| 2157 |
-
"learning_rate": 0.000143249779198529,
|
| 2158 |
-
"loss": 7.882493495941162,
|
| 2159 |
-
"step": 307
|
| 2160 |
-
},
|
| 2161 |
-
{
|
| 2162 |
-
"epoch": 0.616,
|
| 2163 |
-
"grad_norm": 2.802663564682007,
|
| 2164 |
-
"learning_rate": 0.00014198262258793002,
|
| 2165 |
-
"loss": 7.871265888214111,
|
| 2166 |
-
"step": 308
|
| 2167 |
-
},
|
| 2168 |
-
{
|
| 2169 |
-
"epoch": 0.618,
|
| 2170 |
-
"grad_norm": 2.509293794631958,
|
| 2171 |
-
"learning_rate": 0.00014071800384536927,
|
| 2172 |
-
"loss": 8.024494171142578,
|
| 2173 |
-
"step": 309
|
| 2174 |
-
},
|
| 2175 |
-
{
|
| 2176 |
-
"epoch": 0.62,
|
| 2177 |
-
"grad_norm": 2.420099973678589,
|
| 2178 |
-
"learning_rate": 0.00013945597828936737,
|
| 2179 |
-
"loss": 7.889064311981201,
|
| 2180 |
-
"step": 310
|
| 2181 |
-
},
|
| 2182 |
-
{
|
| 2183 |
-
"epoch": 0.622,
|
| 2184 |
-
"grad_norm": 2.6111817359924316,
|
| 2185 |
-
"learning_rate": 0.00013819660112501054,
|
| 2186 |
-
"loss": 8.478769302368164,
|
| 2187 |
-
"step": 311
|
| 2188 |
-
},
|
| 2189 |
-
{
|
| 2190 |
-
"epoch": 0.624,
|
| 2191 |
-
"grad_norm": 2.3872926235198975,
|
| 2192 |
-
"learning_rate": 0.00013693992744153572,
|
| 2193 |
-
"loss": 8.076985359191895,
|
| 2194 |
-
"step": 312
|
| 2195 |
-
},
|
| 2196 |
-
{
|
| 2197 |
-
"epoch": 0.626,
|
| 2198 |
-
"grad_norm": 2.2811145782470703,
|
| 2199 |
-
"learning_rate": 0.00013568601220992097,
|
| 2200 |
-
"loss": 8.777323722839355,
|
| 2201 |
-
"step": 313
|
| 2202 |
-
},
|
| 2203 |
-
{
|
| 2204 |
-
"epoch": 0.628,
|
| 2205 |
-
"grad_norm": 2.441464424133301,
|
| 2206 |
-
"learning_rate": 0.00013443491028048045,
|
| 2207 |
-
"loss": 8.753079414367676,
|
| 2208 |
-
"step": 314
|
| 2209 |
-
},
|
| 2210 |
-
{
|
| 2211 |
-
"epoch": 0.63,
|
| 2212 |
-
"grad_norm": 2.3295352458953857,
|
| 2213 |
-
"learning_rate": 0.0001331866763804658,
|
| 2214 |
-
"loss": 8.511877059936523,
|
| 2215 |
-
"step": 315
|
| 2216 |
-
},
|
| 2217 |
-
{
|
| 2218 |
-
"epoch": 0.632,
|
| 2219 |
-
"grad_norm": 2.2267494201660156,
|
| 2220 |
-
"learning_rate": 0.0001319413651116714,
|
| 2221 |
-
"loss": 8.434760093688965,
|
| 2222 |
-
"step": 316
|
| 2223 |
-
},
|
| 2224 |
-
{
|
| 2225 |
-
"epoch": 0.634,
|
| 2226 |
-
"grad_norm": 2.2881898880004883,
|
| 2227 |
-
"learning_rate": 0.00013069903094804644,
|
| 2228 |
-
"loss": 8.588081359863281,
|
| 2229 |
-
"step": 317
|
| 2230 |
-
},
|
| 2231 |
-
{
|
| 2232 |
-
"epoch": 0.636,
|
| 2233 |
-
"grad_norm": 2.3703527450561523,
|
| 2234 |
-
"learning_rate": 0.0001294597282333118,
|
| 2235 |
-
"loss": 8.745709419250488,
|
| 2236 |
-
"step": 318
|
| 2237 |
-
},
|
| 2238 |
-
{
|
| 2239 |
-
"epoch": 0.638,
|
| 2240 |
-
"grad_norm": 2.7186853885650635,
|
| 2241 |
-
"learning_rate": 0.00012822351117858303,
|
| 2242 |
-
"loss": 9.173514366149902,
|
| 2243 |
-
"step": 319
|
| 2244 |
-
},
|
| 2245 |
-
{
|
| 2246 |
-
"epoch": 0.64,
|
| 2247 |
-
"grad_norm": 2.4356775283813477,
|
| 2248 |
-
"learning_rate": 0.0001269904338599989,
|
| 2249 |
-
"loss": 8.819295883178711,
|
| 2250 |
-
"step": 320
|
| 2251 |
-
},
|
| 2252 |
-
{
|
| 2253 |
-
"epoch": 0.642,
|
| 2254 |
-
"grad_norm": 2.705989360809326,
|
| 2255 |
-
"learning_rate": 0.0001257605502163558,
|
| 2256 |
-
"loss": 8.731491088867188,
|
| 2257 |
-
"step": 321
|
| 2258 |
-
},
|
| 2259 |
-
{
|
| 2260 |
-
"epoch": 0.644,
|
| 2261 |
-
"grad_norm": 2.4525508880615234,
|
| 2262 |
-
"learning_rate": 0.00012453391404674885,
|
| 2263 |
-
"loss": 8.663737297058105,
|
| 2264 |
-
"step": 322
|
| 2265 |
-
},
|
| 2266 |
-
{
|
| 2267 |
-
"epoch": 0.646,
|
| 2268 |
-
"grad_norm": 2.3270366191864014,
|
| 2269 |
-
"learning_rate": 0.00012331057900821768,
|
| 2270 |
-
"loss": 8.994048118591309,
|
| 2271 |
-
"step": 323
|
| 2272 |
-
},
|
| 2273 |
-
{
|
| 2274 |
-
"epoch": 0.648,
|
| 2275 |
-
"grad_norm": 2.415705680847168,
|
| 2276 |
-
"learning_rate": 0.0001220905986134,
|
| 2277 |
-
"loss": 8.778101921081543,
|
| 2278 |
-
"step": 324
|
| 2279 |
-
},
|
| 2280 |
-
{
|
| 2281 |
-
"epoch": 0.65,
|
| 2282 |
-
"grad_norm": 2.230379581451416,
|
| 2283 |
-
"learning_rate": 0.00012087402622819039,
|
| 2284 |
-
"loss": 9.104836463928223,
|
| 2285 |
-
"step": 325
|
| 2286 |
-
},
|
| 2287 |
-
{
|
| 2288 |
-
"epoch": 0.652,
|
| 2289 |
-
"grad_norm": 2.1804556846618652,
|
| 2290 |
-
"learning_rate": 0.00011966091506940616,
|
| 2291 |
-
"loss": 8.784355163574219,
|
| 2292 |
-
"step": 326
|
| 2293 |
-
},
|
| 2294 |
-
{
|
| 2295 |
-
"epoch": 0.654,
|
| 2296 |
-
"grad_norm": 2.2515757083892822,
|
| 2297 |
-
"learning_rate": 0.00011845131820245934,
|
| 2298 |
-
"loss": 8.509770393371582,
|
| 2299 |
-
"step": 327
|
| 2300 |
-
},
|
| 2301 |
-
{
|
| 2302 |
-
"epoch": 0.656,
|
| 2303 |
-
"grad_norm": 2.2556521892547607,
|
| 2304 |
-
"learning_rate": 0.00011724528853903536,
|
| 2305 |
-
"loss": 8.566987037658691,
|
| 2306 |
-
"step": 328
|
| 2307 |
-
},
|
| 2308 |
-
{
|
| 2309 |
-
"epoch": 0.658,
|
| 2310 |
-
"grad_norm": 2.3027467727661133,
|
| 2311 |
-
"learning_rate": 0.00011604287883477889,
|
| 2312 |
-
"loss": 8.478706359863281,
|
| 2313 |
-
"step": 329
|
| 2314 |
-
},
|
| 2315 |
-
{
|
| 2316 |
-
"epoch": 0.66,
|
| 2317 |
-
"grad_norm": 2.430137872695923,
|
| 2318 |
-
"learning_rate": 0.00011484414168698547,
|
| 2319 |
-
"loss": 8.29089069366455,
|
| 2320 |
-
"step": 330
|
| 2321 |
-
},
|
| 2322 |
-
{
|
| 2323 |
-
"epoch": 0.662,
|
| 2324 |
-
"grad_norm": 2.3346168994903564,
|
| 2325 |
-
"learning_rate": 0.00011364912953230145,
|
| 2326 |
-
"loss": 8.463544845581055,
|
| 2327 |
-
"step": 331
|
| 2328 |
-
},
|
| 2329 |
-
{
|
| 2330 |
-
"epoch": 0.664,
|
| 2331 |
-
"grad_norm": 2.498368501663208,
|
| 2332 |
-
"learning_rate": 0.00011245789464442964,
|
| 2333 |
-
"loss": 8.549954414367676,
|
| 2334 |
-
"step": 332
|
| 2335 |
-
},
|
| 2336 |
-
{
|
| 2337 |
-
"epoch": 0.666,
|
| 2338 |
-
"grad_norm": 2.607229709625244,
|
| 2339 |
-
"learning_rate": 0.00011127048913184326,
|
| 2340 |
-
"loss": 8.48581600189209,
|
| 2341 |
-
"step": 333
|
| 2342 |
-
},
|
| 2343 |
-
{
|
| 2344 |
-
"epoch": 0.668,
|
| 2345 |
-
"grad_norm": 2.5177390575408936,
|
| 2346 |
-
"learning_rate": 0.00011008696493550599,
|
| 2347 |
-
"loss": 8.414435386657715,
|
| 2348 |
-
"step": 334
|
| 2349 |
-
},
|
| 2350 |
-
{
|
| 2351 |
-
"epoch": 0.67,
|
| 2352 |
-
"grad_norm": 2.489049196243286,
|
| 2353 |
-
"learning_rate": 0.00010890737382660015,
|
| 2354 |
-
"loss": 8.452604293823242,
|
| 2355 |
-
"step": 335
|
| 2356 |
-
},
|
| 2357 |
-
{
|
| 2358 |
-
"epoch": 0.672,
|
| 2359 |
-
"grad_norm": 2.3974552154541016,
|
| 2360 |
-
"learning_rate": 0.00010773176740426248,
|
| 2361 |
-
"loss": 8.629395484924316,
|
| 2362 |
-
"step": 336
|
| 2363 |
-
},
|
| 2364 |
-
{
|
| 2365 |
-
"epoch": 0.674,
|
| 2366 |
-
"grad_norm": 2.2448816299438477,
|
| 2367 |
-
"learning_rate": 0.00010656019709332606,
|
| 2368 |
-
"loss": 8.524725914001465,
|
| 2369 |
-
"step": 337
|
| 2370 |
-
},
|
| 2371 |
-
{
|
| 2372 |
-
"epoch": 0.676,
|
| 2373 |
-
"grad_norm": 2.31128191947937,
|
| 2374 |
-
"learning_rate": 0.00010539271414207186,
|
| 2375 |
-
"loss": 8.495088577270508,
|
| 2376 |
-
"step": 338
|
| 2377 |
-
},
|
| 2378 |
-
{
|
| 2379 |
-
"epoch": 0.678,
|
| 2380 |
-
"grad_norm": 2.312697649002075,
|
| 2381 |
-
"learning_rate": 0.00010422936961998609,
|
| 2382 |
-
"loss": 8.031539916992188,
|
| 2383 |
-
"step": 339
|
| 2384 |
-
},
|
| 2385 |
-
{
|
| 2386 |
-
"epoch": 0.68,
|
| 2387 |
-
"grad_norm": 2.313143491744995,
|
| 2388 |
-
"learning_rate": 0.00010307021441552707,
|
| 2389 |
-
"loss": 8.151517868041992,
|
| 2390 |
-
"step": 340
|
| 2391 |
-
},
|
| 2392 |
-
{
|
| 2393 |
-
"epoch": 0.682,
|
| 2394 |
-
"grad_norm": 2.7047691345214844,
|
| 2395 |
-
"learning_rate": 0.00010191529923389845,
|
| 2396 |
-
"loss": 8.297797203063965,
|
| 2397 |
-
"step": 341
|
| 2398 |
-
},
|
| 2399 |
-
{
|
| 2400 |
-
"epoch": 0.684,
|
| 2401 |
-
"grad_norm": 2.522658586502075,
|
| 2402 |
-
"learning_rate": 0.00010076467459483155,
|
| 2403 |
-
"loss": 8.585305213928223,
|
| 2404 |
-
"step": 342
|
| 2405 |
-
},
|
| 2406 |
-
{
|
| 2407 |
-
"epoch": 0.686,
|
| 2408 |
-
"grad_norm": 2.7083234786987305,
|
| 2409 |
-
"learning_rate": 9.961839083037592e-05,
|
| 2410 |
-
"loss": 8.55811595916748,
|
| 2411 |
-
"step": 343
|
| 2412 |
-
},
|
| 2413 |
-
{
|
| 2414 |
-
"epoch": 0.688,
|
| 2415 |
-
"grad_norm": 2.555201768875122,
|
| 2416 |
-
"learning_rate": 9.847649808269658e-05,
|
| 2417 |
-
"loss": 8.705409049987793,
|
| 2418 |
-
"step": 344
|
| 2419 |
-
},
|
| 2420 |
-
{
|
| 2421 |
-
"epoch": 0.69,
|
| 2422 |
-
"grad_norm": 2.6906630992889404,
|
| 2423 |
-
"learning_rate": 9.733904630188176e-05,
|
| 2424 |
-
"loss": 8.549100875854492,
|
| 2425 |
-
"step": 345
|
| 2426 |
-
},
|
| 2427 |
-
{
|
| 2428 |
-
"epoch": 0.692,
|
| 2429 |
-
"grad_norm": 2.321979522705078,
|
| 2430 |
-
"learning_rate": 9.620608524375703e-05,
|
| 2431 |
-
"loss": 8.56401252746582,
|
| 2432 |
-
"step": 346
|
| 2433 |
-
},
|
| 2434 |
-
{
|
| 2435 |
-
"epoch": 0.694,
|
| 2436 |
-
"grad_norm": 3.2417521476745605,
|
| 2437 |
-
"learning_rate": 9.507766446770934e-05,
|
| 2438 |
-
"loss": 8.489413261413574,
|
| 2439 |
-
"step": 347
|
| 2440 |
-
},
|
| 2441 |
-
{
|
| 2442 |
-
"epoch": 0.696,
|
| 2443 |
-
"grad_norm": 2.9581375122070312,
|
| 2444 |
-
"learning_rate": 9.39538333345191e-05,
|
| 2445 |
-
"loss": 8.679829597473145,
|
| 2446 |
-
"step": 348
|
| 2447 |
-
},
|
| 2448 |
-
{
|
| 2449 |
-
"epoch": 0.698,
|
| 2450 |
-
"grad_norm": 2.6846065521240234,
|
| 2451 |
-
"learning_rate": 9.283464100420063e-05,
|
| 2452 |
-
"loss": 8.613353729248047,
|
| 2453 |
-
"step": 349
|
| 2454 |
-
},
|
| 2455 |
-
{
|
| 2456 |
-
"epoch": 0.7,
|
| 2457 |
-
"grad_norm": 2.293097734451294,
|
| 2458 |
-
"learning_rate": 9.17201364338524e-05,
|
| 2459 |
-
"loss": 8.863085746765137,
|
| 2460 |
-
"step": 350
|
| 2461 |
-
},
|
| 2462 |
-
{
|
| 2463 |
-
"epoch": 0.702,
|
| 2464 |
-
"grad_norm": 2.689634084701538,
|
| 2465 |
-
"learning_rate": 9.061036837551466e-05,
|
| 2466 |
-
"loss": 8.898355484008789,
|
| 2467 |
-
"step": 351
|
| 2468 |
-
},
|
| 2469 |
-
{
|
| 2470 |
-
"epoch": 0.704,
|
| 2471 |
-
"grad_norm": 2.5077967643737793,
|
| 2472 |
-
"learning_rate": 8.950538537403736e-05,
|
| 2473 |
-
"loss": 8.917553901672363,
|
| 2474 |
-
"step": 352
|
| 2475 |
-
},
|
| 2476 |
-
{
|
| 2477 |
-
"epoch": 0.706,
|
| 2478 |
-
"grad_norm": 2.525681734085083,
|
| 2479 |
-
"learning_rate": 8.840523576495681e-05,
|
| 2480 |
-
"loss": 8.700565338134766,
|
| 2481 |
-
"step": 353
|
| 2482 |
-
},
|
| 2483 |
-
{
|
| 2484 |
-
"epoch": 0.708,
|
| 2485 |
-
"grad_norm": 2.3807644844055176,
|
| 2486 |
-
"learning_rate": 8.730996767238072e-05,
|
| 2487 |
-
"loss": 8.507399559020996,
|
| 2488 |
-
"step": 354
|
| 2489 |
-
},
|
| 2490 |
-
{
|
| 2491 |
-
"epoch": 0.71,
|
| 2492 |
-
"grad_norm": 2.367744207382202,
|
| 2493 |
-
"learning_rate": 8.621962900688378e-05,
|
| 2494 |
-
"loss": 8.729925155639648,
|
| 2495 |
-
"step": 355
|
| 2496 |
-
},
|
| 2497 |
-
{
|
| 2498 |
-
"epoch": 0.712,
|
| 2499 |
-
"grad_norm": 2.3890011310577393,
|
| 2500 |
-
"learning_rate": 8.513426746341128e-05,
|
| 2501 |
-
"loss": 8.4448823928833,
|
| 2502 |
-
"step": 356
|
| 2503 |
-
},
|
| 2504 |
-
{
|
| 2505 |
-
"epoch": 0.714,
|
| 2506 |
-
"grad_norm": 2.440182685852051,
|
| 2507 |
-
"learning_rate": 8.405393051919333e-05,
|
| 2508 |
-
"loss": 8.661076545715332,
|
| 2509 |
-
"step": 357
|
| 2510 |
-
},
|
| 2511 |
-
{
|
| 2512 |
-
"epoch": 0.716,
|
| 2513 |
-
"grad_norm": 2.2747421264648438,
|
| 2514 |
-
"learning_rate": 8.29786654316677e-05,
|
| 2515 |
-
"loss": 8.94270133972168,
|
| 2516 |
-
"step": 358
|
| 2517 |
-
},
|
| 2518 |
-
{
|
| 2519 |
-
"epoch": 0.718,
|
| 2520 |
-
"grad_norm": 2.3551080226898193,
|
| 2521 |
-
"learning_rate": 8.190851923641259e-05,
|
| 2522 |
-
"loss": 8.860125541687012,
|
| 2523 |
-
"step": 359
|
| 2524 |
-
},
|
| 2525 |
-
{
|
| 2526 |
-
"epoch": 0.72,
|
| 2527 |
-
"grad_norm": 2.6194686889648438,
|
| 2528 |
-
"learning_rate": 8.084353874508947e-05,
|
| 2529 |
-
"loss": 8.445380210876465,
|
| 2530 |
-
"step": 360
|
| 2531 |
-
},
|
| 2532 |
-
{
|
| 2533 |
-
"epoch": 0.722,
|
| 2534 |
-
"grad_norm": 2.5639469623565674,
|
| 2535 |
-
"learning_rate": 7.978377054339499e-05,
|
| 2536 |
-
"loss": 8.40030574798584,
|
| 2537 |
-
"step": 361
|
| 2538 |
-
},
|
| 2539 |
-
{
|
| 2540 |
-
"epoch": 0.724,
|
| 2541 |
-
"grad_norm": 2.6488072872161865,
|
| 2542 |
-
"learning_rate": 7.872926098902358e-05,
|
| 2543 |
-
"loss": 8.693204879760742,
|
| 2544 |
-
"step": 362
|
| 2545 |
-
},
|
| 2546 |
-
{
|
| 2547 |
-
"epoch": 0.726,
|
| 2548 |
-
"grad_norm": 2.3694777488708496,
|
| 2549 |
-
"learning_rate": 7.768005620963916e-05,
|
| 2550 |
-
"loss": 8.157784461975098,
|
| 2551 |
-
"step": 363
|
| 2552 |
-
},
|
| 2553 |
-
{
|
| 2554 |
-
"epoch": 0.728,
|
| 2555 |
-
"grad_norm": 2.6463561058044434,
|
| 2556 |
-
"learning_rate": 7.663620210085781e-05,
|
| 2557 |
-
"loss": 8.414434432983398,
|
| 2558 |
-
"step": 364
|
| 2559 |
-
},
|
| 2560 |
-
{
|
| 2561 |
-
"epoch": 0.73,
|
| 2562 |
-
"grad_norm": 2.853205680847168,
|
| 2563 |
-
"learning_rate": 7.55977443242399e-05,
|
| 2564 |
-
"loss": 7.885345458984375,
|
| 2565 |
-
"step": 365
|
| 2566 |
-
},
|
| 2567 |
-
{
|
| 2568 |
-
"epoch": 0.732,
|
| 2569 |
-
"grad_norm": 2.3680975437164307,
|
| 2570 |
-
"learning_rate": 7.456472830529259e-05,
|
| 2571 |
-
"loss": 8.413387298583984,
|
| 2572 |
-
"step": 366
|
| 2573 |
-
},
|
| 2574 |
-
{
|
| 2575 |
-
"epoch": 0.734,
|
| 2576 |
-
"grad_norm": 3.5096073150634766,
|
| 2577 |
-
"learning_rate": 7.353719923148324e-05,
|
| 2578 |
-
"loss": 8.39874267578125,
|
| 2579 |
-
"step": 367
|
| 2580 |
-
},
|
| 2581 |
-
{
|
| 2582 |
-
"epoch": 0.736,
|
| 2583 |
-
"grad_norm": 2.7081639766693115,
|
| 2584 |
-
"learning_rate": 7.251520205026205e-05,
|
| 2585 |
-
"loss": 8.545775413513184,
|
| 2586 |
-
"step": 368
|
| 2587 |
-
},
|
| 2588 |
-
{
|
| 2589 |
-
"epoch": 0.738,
|
| 2590 |
-
"grad_norm": 2.63431453704834,
|
| 2591 |
-
"learning_rate": 7.149878146709676e-05,
|
| 2592 |
-
"loss": 8.475957870483398,
|
| 2593 |
-
"step": 369
|
| 2594 |
-
},
|
| 2595 |
-
{
|
| 2596 |
-
"epoch": 0.74,
|
| 2597 |
-
"grad_norm": 2.6949734687805176,
|
| 2598 |
-
"learning_rate": 7.048798194351625e-05,
|
| 2599 |
-
"loss": 9.073458671569824,
|
| 2600 |
-
"step": 370
|
| 2601 |
-
},
|
| 2602 |
-
{
|
| 2603 |
-
"epoch": 0.742,
|
| 2604 |
-
"grad_norm": 2.6518638134002686,
|
| 2605 |
-
"learning_rate": 6.948284769516627e-05,
|
| 2606 |
-
"loss": 9.169623374938965,
|
| 2607 |
-
"step": 371
|
| 2608 |
-
},
|
| 2609 |
-
{
|
| 2610 |
-
"epoch": 0.744,
|
| 2611 |
-
"grad_norm": 2.63506817817688,
|
| 2612 |
-
"learning_rate": 6.848342268987511e-05,
|
| 2613 |
-
"loss": 9.147165298461914,
|
| 2614 |
-
"step": 372
|
| 2615 |
-
},
|
| 2616 |
-
{
|
| 2617 |
-
"epoch": 0.746,
|
| 2618 |
-
"grad_norm": 2.960604190826416,
|
| 2619 |
-
"learning_rate": 6.748975064573007e-05,
|
| 2620 |
-
"loss": 8.527523040771484,
|
| 2621 |
-
"step": 373
|
| 2622 |
-
},
|
| 2623 |
-
{
|
| 2624 |
-
"epoch": 0.748,
|
| 2625 |
-
"grad_norm": 2.6785247325897217,
|
| 2626 |
-
"learning_rate": 6.650187502916552e-05,
|
| 2627 |
-
"loss": 8.792768478393555,
|
| 2628 |
-
"step": 374
|
| 2629 |
-
},
|
| 2630 |
-
{
|
| 2631 |
-
"epoch": 0.75,
|
| 2632 |
-
"grad_norm": 2.8223087787628174,
|
| 2633 |
-
"learning_rate": 6.551983905306107e-05,
|
| 2634 |
-
"loss": 8.6453857421875,
|
| 2635 |
-
"step": 375
|
| 2636 |
-
},
|
| 2637 |
-
{
|
| 2638 |
-
"epoch": 0.752,
|
| 2639 |
-
"grad_norm": 2.6297085285186768,
|
| 2640 |
-
"learning_rate": 6.454368567485183e-05,
|
| 2641 |
-
"loss": 8.664865493774414,
|
| 2642 |
-
"step": 376
|
| 2643 |
-
},
|
| 2644 |
-
{
|
| 2645 |
-
"epoch": 0.754,
|
| 2646 |
-
"grad_norm": 2.673945903778076,
|
| 2647 |
-
"learning_rate": 6.35734575946487e-05,
|
| 2648 |
-
"loss": 8.782684326171875,
|
| 2649 |
-
"step": 377
|
| 2650 |
-
},
|
| 2651 |
-
{
|
| 2652 |
-
"epoch": 0.756,
|
| 2653 |
-
"grad_norm": 2.5158681869506836,
|
| 2654 |
-
"learning_rate": 6.260919725337109e-05,
|
| 2655 |
-
"loss": 8.47056770324707,
|
| 2656 |
-
"step": 378
|
| 2657 |
-
},
|
| 2658 |
-
{
|
| 2659 |
-
"epoch": 0.758,
|
| 2660 |
-
"grad_norm": 2.592374801635742,
|
| 2661 |
-
"learning_rate": 6.165094683089015e-05,
|
| 2662 |
-
"loss": 8.751676559448242,
|
| 2663 |
-
"step": 379
|
| 2664 |
-
},
|
| 2665 |
-
{
|
| 2666 |
-
"epoch": 0.76,
|
| 2667 |
-
"grad_norm": 2.9102237224578857,
|
| 2668 |
-
"learning_rate": 6.069874824418356e-05,
|
| 2669 |
-
"loss": 8.525209426879883,
|
| 2670 |
-
"step": 380
|
| 2671 |
-
},
|
| 2672 |
-
{
|
| 2673 |
-
"epoch": 0.762,
|
| 2674 |
-
"grad_norm": 2.4679441452026367,
|
| 2675 |
-
"learning_rate": 5.975264314550229e-05,
|
| 2676 |
-
"loss": 8.670839309692383,
|
| 2677 |
-
"step": 381
|
| 2678 |
-
},
|
| 2679 |
-
{
|
| 2680 |
-
"epoch": 0.764,
|
| 2681 |
-
"grad_norm": 2.483761787414551,
|
| 2682 |
-
"learning_rate": 5.881267292054828e-05,
|
| 2683 |
-
"loss": 8.52482795715332,
|
| 2684 |
-
"step": 382
|
| 2685 |
-
},
|
| 2686 |
-
{
|
| 2687 |
-
"epoch": 0.766,
|
| 2688 |
-
"grad_norm": 2.954261541366577,
|
| 2689 |
-
"learning_rate": 5.787887868666417e-05,
|
| 2690 |
-
"loss": 8.72692584991455,
|
| 2691 |
-
"step": 383
|
| 2692 |
-
},
|
| 2693 |
-
{
|
| 2694 |
-
"epoch": 0.768,
|
| 2695 |
-
"grad_norm": 3.019418239593506,
|
| 2696 |
-
"learning_rate": 5.6951301291034945e-05,
|
| 2697 |
-
"loss": 8.547119140625,
|
| 2698 |
-
"step": 384
|
| 2699 |
-
},
|
| 2700 |
-
{
|
| 2701 |
-
"epoch": 0.77,
|
| 2702 |
-
"grad_norm": 2.727445363998413,
|
| 2703 |
-
"learning_rate": 5.602998130890065e-05,
|
| 2704 |
-
"loss": 8.689985275268555,
|
| 2705 |
-
"step": 385
|
| 2706 |
-
},
|
| 2707 |
-
{
|
| 2708 |
-
"epoch": 0.772,
|
| 2709 |
-
"grad_norm": 2.708590507507324,
|
| 2710 |
-
"learning_rate": 5.511495904178221e-05,
|
| 2711 |
-
"loss": 8.477546691894531,
|
| 2712 |
-
"step": 386
|
| 2713 |
-
},
|
| 2714 |
-
{
|
| 2715 |
-
"epoch": 0.774,
|
| 2716 |
-
"grad_norm": 2.6245064735412598,
|
| 2717 |
-
"learning_rate": 5.4206274515717736e-05,
|
| 2718 |
-
"loss": 8.42736530303955,
|
| 2719 |
-
"step": 387
|
| 2720 |
-
},
|
| 2721 |
-
{
|
| 2722 |
-
"epoch": 0.776,
|
| 2723 |
-
"grad_norm": 2.5949771404266357,
|
| 2724 |
-
"learning_rate": 5.330396747951205e-05,
|
| 2725 |
-
"loss": 8.770480155944824,
|
| 2726 |
-
"step": 388
|
| 2727 |
-
},
|
| 2728 |
-
{
|
| 2729 |
-
"epoch": 0.778,
|
| 2730 |
-
"grad_norm": 2.57098126411438,
|
| 2731 |
-
"learning_rate": 5.240807740299811e-05,
|
| 2732 |
-
"loss": 8.526888847351074,
|
| 2733 |
-
"step": 389
|
| 2734 |
-
},
|
| 2735 |
-
{
|
| 2736 |
-
"epoch": 0.78,
|
| 2737 |
-
"grad_norm": 3.136831045150757,
|
| 2738 |
-
"learning_rate": 5.1518643475310034e-05,
|
| 2739 |
-
"loss": 8.408879280090332,
|
| 2740 |
-
"step": 390
|
| 2741 |
-
},
|
| 2742 |
-
{
|
| 2743 |
-
"epoch": 0.782,
|
| 2744 |
-
"grad_norm": 3.5655157566070557,
|
| 2745 |
-
"learning_rate": 5.0635704603169287e-05,
|
| 2746 |
-
"loss": 8.627933502197266,
|
| 2747 |
-
"step": 391
|
| 2748 |
-
},
|
| 2749 |
-
{
|
| 2750 |
-
"epoch": 0.784,
|
| 2751 |
-
"grad_norm": 2.917283296585083,
|
| 2752 |
-
"learning_rate": 4.975929940918236e-05,
|
| 2753 |
-
"loss": 8.751184463500977,
|
| 2754 |
-
"step": 392
|
| 2755 |
-
},
|
| 2756 |
-
{
|
| 2757 |
-
"epoch": 0.786,
|
| 2758 |
-
"grad_norm": 2.7184383869171143,
|
| 2759 |
-
"learning_rate": 4.8889466230151646e-05,
|
| 2760 |
-
"loss": 8.992475509643555,
|
| 2761 |
-
"step": 393
|
| 2762 |
-
},
|
| 2763 |
-
{
|
| 2764 |
-
"epoch": 0.788,
|
| 2765 |
-
"grad_norm": 2.911776304244995,
|
| 2766 |
-
"learning_rate": 4.8026243115398314e-05,
|
| 2767 |
-
"loss": 8.160642623901367,
|
| 2768 |
-
"step": 394
|
| 2769 |
},
|
| 2770 |
{
|
| 2771 |
-
"epoch": 0.
|
| 2772 |
-
"
|
| 2773 |
-
"
|
| 2774 |
-
"
|
| 2775 |
-
"
|
|
|
|
|
|
|
| 2776 |
},
|
| 2777 |
{
|
| 2778 |
-
"epoch": 0.
|
| 2779 |
-
"grad_norm": 2.
|
| 2780 |
-
"learning_rate":
|
| 2781 |
-
"loss": 8.
|
| 2782 |
-
"step":
|
| 2783 |
},
|
| 2784 |
{
|
| 2785 |
-
"epoch": 0.
|
| 2786 |
-
"
|
| 2787 |
-
"
|
| 2788 |
-
"
|
| 2789 |
-
"
|
|
|
|
|
|
|
| 2790 |
},
|
| 2791 |
{
|
| 2792 |
-
"epoch": 0.
|
| 2793 |
-
"grad_norm":
|
| 2794 |
-
"learning_rate":
|
| 2795 |
-
"loss": 8.
|
| 2796 |
-
"step":
|
| 2797 |
},
|
| 2798 |
{
|
| 2799 |
-
"epoch": 0.
|
| 2800 |
-
"
|
| 2801 |
-
"
|
| 2802 |
-
"
|
| 2803 |
-
"
|
|
|
|
|
|
|
| 2804 |
},
|
| 2805 |
{
|
| 2806 |
-
"epoch": 0.
|
| 2807 |
-
"grad_norm":
|
| 2808 |
-
"learning_rate":
|
| 2809 |
-
"loss": 8.
|
| 2810 |
"step": 400
|
| 2811 |
},
|
| 2812 |
{
|
| 2813 |
-
"epoch": 0.
|
| 2814 |
-
"
|
| 2815 |
-
"
|
| 2816 |
-
"
|
| 2817 |
-
"
|
| 2818 |
-
|
| 2819 |
-
|
| 2820 |
-
"epoch": 0.804,
|
| 2821 |
-
"grad_norm": 2.688173770904541,
|
| 2822 |
-
"learning_rate": 4.136289025462443e-05,
|
| 2823 |
-
"loss": 8.424184799194336,
|
| 2824 |
-
"step": 402
|
| 2825 |
-
},
|
| 2826 |
-
{
|
| 2827 |
-
"epoch": 0.806,
|
| 2828 |
-
"grad_norm": 2.8252346515655518,
|
| 2829 |
-
"learning_rate": 4.0560821689620856e-05,
|
| 2830 |
-
"loss": 8.815681457519531,
|
| 2831 |
-
"step": 403
|
| 2832 |
-
},
|
| 2833 |
-
{
|
| 2834 |
-
"epoch": 0.808,
|
| 2835 |
-
"grad_norm": 3.3969688415527344,
|
| 2836 |
-
"learning_rate": 3.976572751078782e-05,
|
| 2837 |
-
"loss": 8.551783561706543,
|
| 2838 |
-
"step": 404
|
| 2839 |
-
},
|
| 2840 |
-
{
|
| 2841 |
-
"epoch": 0.81,
|
| 2842 |
-
"grad_norm": 3.14737606048584,
|
| 2843 |
-
"learning_rate": 3.8977642498120594e-05,
|
| 2844 |
-
"loss": 8.264228820800781,
|
| 2845 |
-
"step": 405
|
| 2846 |
-
},
|
| 2847 |
-
{
|
| 2848 |
-
"epoch": 0.812,
|
| 2849 |
-
"grad_norm": 3.2091007232666016,
|
| 2850 |
-
"learning_rate": 3.819660112501053e-05,
|
| 2851 |
-
"loss": 8.55145263671875,
|
| 2852 |
-
"step": 406
|
| 2853 |
-
},
|
| 2854 |
-
{
|
| 2855 |
-
"epoch": 0.814,
|
| 2856 |
-
"grad_norm": 3.197000503540039,
|
| 2857 |
-
"learning_rate": 3.742263755673758e-05,
|
| 2858 |
-
"loss": 8.59677791595459,
|
| 2859 |
-
"step": 407
|
| 2860 |
-
},
|
| 2861 |
-
{
|
| 2862 |
-
"epoch": 0.816,
|
| 2863 |
-
"grad_norm": 3.0365517139434814,
|
| 2864 |
-
"learning_rate": 3.6655785648975585e-05,
|
| 2865 |
-
"loss": 8.459378242492676,
|
| 2866 |
-
"step": 408
|
| 2867 |
-
},
|
| 2868 |
-
{
|
| 2869 |
-
"epoch": 0.818,
|
| 2870 |
-
"grad_norm": 3.6148812770843506,
|
| 2871 |
-
"learning_rate": 3.589607894631111e-05,
|
| 2872 |
-
"loss": 9.298233032226562,
|
| 2873 |
-
"step": 409
|
| 2874 |
-
},
|
| 2875 |
-
{
|
| 2876 |
-
"epoch": 0.82,
|
| 2877 |
-
"grad_norm": 3.607222080230713,
|
| 2878 |
-
"learning_rate": 3.514355068077655e-05,
|
| 2879 |
-
"loss": 8.133456230163574,
|
| 2880 |
-
"step": 410
|
| 2881 |
-
},
|
| 2882 |
-
{
|
| 2883 |
-
"epoch": 0.822,
|
| 2884 |
-
"grad_norm": 3.6667959690093994,
|
| 2885 |
-
"learning_rate": 3.439823377039599e-05,
|
| 2886 |
-
"loss": 8.69001579284668,
|
| 2887 |
-
"step": 411
|
| 2888 |
-
},
|
| 2889 |
-
{
|
| 2890 |
-
"epoch": 0.824,
|
| 2891 |
-
"grad_norm": 3.8627288341522217,
|
| 2892 |
-
"learning_rate": 3.36601608177457e-05,
|
| 2893 |
-
"loss": 9.25894546508789,
|
| 2894 |
-
"step": 412
|
| 2895 |
-
},
|
| 2896 |
-
{
|
| 2897 |
-
"epoch": 0.826,
|
| 2898 |
-
"grad_norm": 3.4121053218841553,
|
| 2899 |
-
"learning_rate": 3.292936410852754e-05,
|
| 2900 |
-
"loss": 8.787903785705566,
|
| 2901 |
-
"step": 413
|
| 2902 |
-
},
|
| 2903 |
-
{
|
| 2904 |
-
"epoch": 0.828,
|
| 2905 |
-
"grad_norm": 4.225803852081299,
|
| 2906 |
-
"learning_rate": 3.220587561015709e-05,
|
| 2907 |
-
"loss": 8.680237770080566,
|
| 2908 |
-
"step": 414
|
| 2909 |
-
},
|
| 2910 |
-
{
|
| 2911 |
-
"epoch": 0.83,
|
| 2912 |
-
"grad_norm": 4.281456470489502,
|
| 2913 |
-
"learning_rate": 3.148972697036507e-05,
|
| 2914 |
-
"loss": 8.895009994506836,
|
| 2915 |
-
"step": 415
|
| 2916 |
-
},
|
| 2917 |
-
{
|
| 2918 |
-
"epoch": 0.832,
|
| 2919 |
-
"grad_norm": 3.286329746246338,
|
| 2920 |
-
"learning_rate": 3.078094951581289e-05,
|
| 2921 |
-
"loss": 8.653852462768555,
|
| 2922 |
-
"step": 416
|
| 2923 |
-
},
|
| 2924 |
-
{
|
| 2925 |
-
"epoch": 0.834,
|
| 2926 |
-
"grad_norm": 3.1779611110687256,
|
| 2927 |
-
"learning_rate": 3.007957425072265e-05,
|
| 2928 |
-
"loss": 8.923322677612305,
|
| 2929 |
-
"step": 417
|
| 2930 |
-
},
|
| 2931 |
-
{
|
| 2932 |
-
"epoch": 0.836,
|
| 2933 |
-
"grad_norm": 3.193608522415161,
|
| 2934 |
-
"learning_rate": 2.9385631855520546e-05,
|
| 2935 |
-
"loss": 8.368019104003906,
|
| 2936 |
-
"step": 418
|
| 2937 |
-
},
|
| 2938 |
-
{
|
| 2939 |
-
"epoch": 0.838,
|
| 2940 |
-
"grad_norm": 3.545006036758423,
|
| 2941 |
-
"learning_rate": 2.8699152685494925e-05,
|
| 2942 |
-
"loss": 8.54011058807373,
|
| 2943 |
-
"step": 419
|
| 2944 |
-
},
|
| 2945 |
-
{
|
| 2946 |
-
"epoch": 0.84,
|
| 2947 |
-
"grad_norm": 5.63062858581543,
|
| 2948 |
-
"learning_rate": 2.8020166769468616e-05,
|
| 2949 |
-
"loss": 8.574893951416016,
|
| 2950 |
-
"step": 420
|
| 2951 |
-
},
|
| 2952 |
-
{
|
| 2953 |
-
"epoch": 0.842,
|
| 2954 |
-
"grad_norm": 5.535733222961426,
|
| 2955 |
-
"learning_rate": 2.7348703808485223e-05,
|
| 2956 |
-
"loss": 8.756441116333008,
|
| 2957 |
-
"step": 421
|
| 2958 |
-
},
|
| 2959 |
-
{
|
| 2960 |
-
"epoch": 0.844,
|
| 2961 |
-
"grad_norm": 3.555478572845459,
|
| 2962 |
-
"learning_rate": 2.6684793174509915e-05,
|
| 2963 |
-
"loss": 8.810291290283203,
|
| 2964 |
-
"step": 422
|
| 2965 |
-
},
|
| 2966 |
-
{
|
| 2967 |
-
"epoch": 0.846,
|
| 2968 |
-
"grad_norm": 4.984836101531982,
|
| 2969 |
-
"learning_rate": 2.6028463909144574e-05,
|
| 2970 |
-
"loss": 8.217482566833496,
|
| 2971 |
-
"step": 423
|
| 2972 |
-
},
|
| 2973 |
-
{
|
| 2974 |
-
"epoch": 0.848,
|
| 2975 |
-
"grad_norm": 5.291404724121094,
|
| 2976 |
-
"learning_rate": 2.5379744722357403e-05,
|
| 2977 |
-
"loss": 9.00054931640625,
|
| 2978 |
-
"step": 424
|
| 2979 |
-
},
|
| 2980 |
-
{
|
| 2981 |
-
"epoch": 0.85,
|
| 2982 |
-
"grad_norm": 4.323802471160889,
|
| 2983 |
-
"learning_rate": 2.473866399122733e-05,
|
| 2984 |
-
"loss": 8.62179183959961,
|
| 2985 |
-
"step": 425
|
| 2986 |
-
},
|
| 2987 |
-
{
|
| 2988 |
-
"epoch": 0.852,
|
| 2989 |
-
"grad_norm": 3.6583471298217773,
|
| 2990 |
-
"learning_rate": 2.410524975870221e-05,
|
| 2991 |
-
"loss": 8.83240795135498,
|
| 2992 |
-
"step": 426
|
| 2993 |
-
},
|
| 2994 |
-
{
|
| 2995 |
-
"epoch": 0.854,
|
| 2996 |
-
"grad_norm": 4.0204854011535645,
|
| 2997 |
-
"learning_rate": 2.347952973237262e-05,
|
| 2998 |
-
"loss": 9.110936164855957,
|
| 2999 |
-
"step": 427
|
| 3000 |
-
},
|
| 3001 |
-
{
|
| 3002 |
-
"epoch": 0.856,
|
| 3003 |
-
"grad_norm": 3.66446852684021,
|
| 3004 |
-
"learning_rate": 2.286153128325954e-05,
|
| 3005 |
-
"loss": 9.406879425048828,
|
| 3006 |
-
"step": 428
|
| 3007 |
-
},
|
| 3008 |
-
{
|
| 3009 |
-
"epoch": 0.858,
|
| 3010 |
-
"grad_norm": 4.050625801086426,
|
| 3011 |
-
"learning_rate": 2.2251281444617257e-05,
|
| 3012 |
-
"loss": 8.864500045776367,
|
| 3013 |
-
"step": 429
|
| 3014 |
-
},
|
| 3015 |
-
{
|
| 3016 |
-
"epoch": 0.86,
|
| 3017 |
-
"grad_norm": 3.9661545753479004,
|
| 3018 |
-
"learning_rate": 2.1648806910750575e-05,
|
| 3019 |
-
"loss": 8.571767807006836,
|
| 3020 |
-
"step": 430
|
| 3021 |
-
},
|
| 3022 |
-
{
|
| 3023 |
-
"epoch": 0.862,
|
| 3024 |
-
"grad_norm": 3.8447558879852295,
|
| 3025 |
-
"learning_rate": 2.1054134035847307e-05,
|
| 3026 |
-
"loss": 8.509228706359863,
|
| 3027 |
-
"step": 431
|
| 3028 |
-
},
|
| 3029 |
-
{
|
| 3030 |
-
"epoch": 0.864,
|
| 3031 |
-
"grad_norm": 4.050955295562744,
|
| 3032 |
-
"learning_rate": 2.0467288832825583e-05,
|
| 3033 |
-
"loss": 8.64655876159668,
|
| 3034 |
-
"step": 432
|
| 3035 |
-
},
|
| 3036 |
-
{
|
| 3037 |
-
"epoch": 0.866,
|
| 3038 |
-
"grad_norm": 3.8370473384857178,
|
| 3039 |
-
"learning_rate": 1.9888296972195587e-05,
|
| 3040 |
-
"loss": 8.47394847869873,
|
| 3041 |
-
"step": 433
|
| 3042 |
-
},
|
| 3043 |
-
{
|
| 3044 |
-
"epoch": 0.868,
|
| 3045 |
-
"grad_norm": 6.3575334548950195,
|
| 3046 |
-
"learning_rate": 1.931718378093703e-05,
|
| 3047 |
-
"loss": 9.061661720275879,
|
| 3048 |
-
"step": 434
|
| 3049 |
-
},
|
| 3050 |
-
{
|
| 3051 |
-
"epoch": 0.87,
|
| 3052 |
-
"grad_norm": 5.828470706939697,
|
| 3053 |
-
"learning_rate": 1.875397424139109e-05,
|
| 3054 |
-
"loss": 8.819586753845215,
|
| 3055 |
-
"step": 435
|
| 3056 |
-
},
|
| 3057 |
-
{
|
| 3058 |
-
"epoch": 0.872,
|
| 3059 |
-
"grad_norm": 6.855267524719238,
|
| 3060 |
-
"learning_rate": 1.81986929901675e-05,
|
| 3061 |
-
"loss": 9.065621376037598,
|
| 3062 |
-
"step": 436
|
| 3063 |
-
},
|
| 3064 |
-
{
|
| 3065 |
-
"epoch": 0.874,
|
| 3066 |
-
"grad_norm": 7.711219310760498,
|
| 3067 |
-
"learning_rate": 1.765136431706711e-05,
|
| 3068 |
-
"loss": 8.79701042175293,
|
| 3069 |
-
"step": 437
|
| 3070 |
-
},
|
| 3071 |
-
{
|
| 3072 |
-
"epoch": 0.876,
|
| 3073 |
-
"grad_norm": 6.396193981170654,
|
| 3074 |
-
"learning_rate": 1.711201216401912e-05,
|
| 3075 |
-
"loss": 8.910534858703613,
|
| 3076 |
-
"step": 438
|
| 3077 |
-
},
|
| 3078 |
-
{
|
| 3079 |
-
"epoch": 0.878,
|
| 3080 |
-
"grad_norm": 4.767712116241455,
|
| 3081 |
-
"learning_rate": 1.6580660124034032e-05,
|
| 3082 |
-
"loss": 8.69563102722168,
|
| 3083 |
-
"step": 439
|
| 3084 |
-
},
|
| 3085 |
-
{
|
| 3086 |
-
"epoch": 0.88,
|
| 3087 |
-
"grad_norm": 6.486217498779297,
|
| 3088 |
-
"learning_rate": 1.605733144017132e-05,
|
| 3089 |
-
"loss": 8.826662063598633,
|
| 3090 |
-
"step": 440
|
| 3091 |
-
},
|
| 3092 |
-
{
|
| 3093 |
-
"epoch": 0.882,
|
| 3094 |
-
"grad_norm": 6.275097370147705,
|
| 3095 |
-
"learning_rate": 1.5542049004523053e-05,
|
| 3096 |
-
"loss": 8.957749366760254,
|
| 3097 |
-
"step": 441
|
| 3098 |
-
},
|
| 3099 |
-
{
|
| 3100 |
-
"epoch": 0.884,
|
| 3101 |
-
"grad_norm": 7.697714328765869,
|
| 3102 |
-
"learning_rate": 1.503483535721224e-05,
|
| 3103 |
-
"loss": 8.713743209838867,
|
| 3104 |
-
"step": 442
|
| 3105 |
-
},
|
| 3106 |
-
{
|
| 3107 |
-
"epoch": 0.886,
|
| 3108 |
-
"grad_norm": 5.000650882720947,
|
| 3109 |
-
"learning_rate": 1.4535712685406921e-05,
|
| 3110 |
-
"loss": 8.948147773742676,
|
| 3111 |
-
"step": 443
|
| 3112 |
-
},
|
| 3113 |
-
{
|
| 3114 |
-
"epoch": 0.888,
|
| 3115 |
-
"grad_norm": 6.006772994995117,
|
| 3116 |
-
"learning_rate": 1.4044702822349731e-05,
|
| 3117 |
-
"loss": 9.16224479675293,
|
| 3118 |
-
"step": 444
|
| 3119 |
-
},
|
| 3120 |
-
{
|
| 3121 |
-
"epoch": 0.89,
|
| 3122 |
-
"grad_norm": 5.949647426605225,
|
| 3123 |
-
"learning_rate": 1.3561827246402692e-05,
|
| 3124 |
-
"loss": 8.794259071350098,
|
| 3125 |
-
"step": 445
|
| 3126 |
-
},
|
| 3127 |
-
{
|
| 3128 |
-
"epoch": 0.892,
|
| 3129 |
-
"grad_norm": 8.286492347717285,
|
| 3130 |
-
"learning_rate": 1.3087107080107853e-05,
|
| 3131 |
-
"loss": 8.689149856567383,
|
| 3132 |
-
"step": 446
|
| 3133 |
-
},
|
| 3134 |
-
{
|
| 3135 |
-
"epoch": 0.894,
|
| 3136 |
-
"grad_norm": 7.273082256317139,
|
| 3137 |
-
"learning_rate": 1.2620563089263093e-05,
|
| 3138 |
-
"loss": 9.098934173583984,
|
| 3139 |
-
"step": 447
|
| 3140 |
-
},
|
| 3141 |
-
{
|
| 3142 |
-
"epoch": 0.896,
|
| 3143 |
-
"grad_norm": 6.687265396118164,
|
| 3144 |
-
"learning_rate": 1.2162215682014012e-05,
|
| 3145 |
-
"loss": 9.410205841064453,
|
| 3146 |
-
"step": 448
|
| 3147 |
-
},
|
| 3148 |
-
{
|
| 3149 |
-
"epoch": 0.898,
|
| 3150 |
-
"grad_norm": 8.580451011657715,
|
| 3151 |
-
"learning_rate": 1.1712084907961053e-05,
|
| 3152 |
-
"loss": 9.154172897338867,
|
| 3153 |
-
"step": 449
|
| 3154 |
-
},
|
| 3155 |
-
{
|
| 3156 |
-
"epoch": 0.9,
|
| 3157 |
-
"grad_norm": 6.76349401473999,
|
| 3158 |
-
"learning_rate": 1.127019045728246e-05,
|
| 3159 |
-
"loss": 9.504095077514648,
|
| 3160 |
-
"step": 450
|
| 3161 |
-
},
|
| 3162 |
-
{
|
| 3163 |
-
"epoch": 0.902,
|
| 3164 |
-
"grad_norm": 5.852334499359131,
|
| 3165 |
-
"learning_rate": 1.0836551659873074e-05,
|
| 3166 |
-
"loss": 9.026161193847656,
|
| 3167 |
-
"step": 451
|
| 3168 |
-
},
|
| 3169 |
-
{
|
| 3170 |
-
"epoch": 0.904,
|
| 3171 |
-
"grad_norm": 5.423657417297363,
|
| 3172 |
-
"learning_rate": 1.0411187484498652e-05,
|
| 3173 |
-
"loss": 9.154343605041504,
|
| 3174 |
-
"step": 452
|
| 3175 |
-
},
|
| 3176 |
-
{
|
| 3177 |
-
"epoch": 0.906,
|
| 3178 |
-
"grad_norm": 5.974980354309082,
|
| 3179 |
-
"learning_rate": 9.99411653796627e-06,
|
| 3180 |
-
"loss": 8.913445472717285,
|
| 3181 |
-
"step": 453
|
| 3182 |
-
},
|
| 3183 |
-
{
|
| 3184 |
-
"epoch": 0.908,
|
| 3185 |
-
"grad_norm": 8.780969619750977,
|
| 3186 |
-
"learning_rate": 9.58535706431023e-06,
|
| 3187 |
-
"loss": 9.358975410461426,
|
| 3188 |
-
"step": 454
|
| 3189 |
-
},
|
| 3190 |
-
{
|
| 3191 |
-
"epoch": 0.91,
|
| 3192 |
-
"grad_norm": 9.879934310913086,
|
| 3193 |
-
"learning_rate": 9.184926943994044e-06,
|
| 3194 |
-
"loss": 9.263197898864746,
|
| 3195 |
-
"step": 455
|
| 3196 |
-
},
|
| 3197 |
-
{
|
| 3198 |
-
"epoch": 0.912,
|
| 3199 |
-
"grad_norm": 11.006145477294922,
|
| 3200 |
-
"learning_rate": 8.792843693128471e-06,
|
| 3201 |
-
"loss": 9.612791061401367,
|
| 3202 |
-
"step": 456
|
| 3203 |
-
},
|
| 3204 |
-
{
|
| 3205 |
-
"epoch": 0.914,
|
| 3206 |
-
"grad_norm": 7.865578651428223,
|
| 3207 |
-
"learning_rate": 8.409124462705032e-06,
|
| 3208 |
-
"loss": 9.142531394958496,
|
| 3209 |
-
"step": 457
|
| 3210 |
-
},
|
| 3211 |
-
{
|
| 3212 |
-
"epoch": 0.916,
|
| 3213 |
-
"grad_norm": 8.890719413757324,
|
| 3214 |
-
"learning_rate": 8.033786037845992e-06,
|
| 3215 |
-
"loss": 9.212858200073242,
|
| 3216 |
-
"step": 458
|
| 3217 |
-
},
|
| 3218 |
-
{
|
| 3219 |
-
"epoch": 0.918,
|
| 3220 |
-
"grad_norm": 7.437742710113525,
|
| 3221 |
-
"learning_rate": 7.66684483706992e-06,
|
| 3222 |
-
"loss": 9.169017791748047,
|
| 3223 |
-
"step": 459
|
| 3224 |
-
},
|
| 3225 |
-
{
|
| 3226 |
-
"epoch": 0.92,
|
| 3227 |
-
"grad_norm": 11.727737426757812,
|
| 3228 |
-
"learning_rate": 7.308316911573721e-06,
|
| 3229 |
-
"loss": 8.899211883544922,
|
| 3230 |
-
"step": 460
|
| 3231 |
-
},
|
| 3232 |
-
{
|
| 3233 |
-
"epoch": 0.922,
|
| 3234 |
-
"grad_norm": 8.83476734161377,
|
| 3235 |
-
"learning_rate": 6.958217944530287e-06,
|
| 3236 |
-
"loss": 9.01573657989502,
|
| 3237 |
-
"step": 461
|
| 3238 |
-
},
|
| 3239 |
-
{
|
| 3240 |
-
"epoch": 0.924,
|
| 3241 |
-
"grad_norm": 7.458086967468262,
|
| 3242 |
-
"learning_rate": 6.616563250402585e-06,
|
| 3243 |
-
"loss": 9.113874435424805,
|
| 3244 |
-
"step": 462
|
| 3245 |
-
},
|
| 3246 |
-
{
|
| 3247 |
-
"epoch": 0.926,
|
| 3248 |
-
"grad_norm": 10.21794605255127,
|
| 3249 |
-
"learning_rate": 6.283367774273785e-06,
|
| 3250 |
-
"loss": 9.42385196685791,
|
| 3251 |
-
"step": 463
|
| 3252 |
-
},
|
| 3253 |
-
{
|
| 3254 |
-
"epoch": 0.928,
|
| 3255 |
-
"grad_norm": 8.564632415771484,
|
| 3256 |
-
"learning_rate": 5.958646091193387e-06,
|
| 3257 |
-
"loss": 9.44645881652832,
|
| 3258 |
-
"step": 464
|
| 3259 |
-
},
|
| 3260 |
-
{
|
| 3261 |
-
"epoch": 0.93,
|
| 3262 |
-
"grad_norm": 12.505248069763184,
|
| 3263 |
-
"learning_rate": 5.642412405539798e-06,
|
| 3264 |
-
"loss": 9.598602294921875,
|
| 3265 |
-
"step": 465
|
| 3266 |
-
},
|
| 3267 |
-
{
|
| 3268 |
-
"epoch": 0.932,
|
| 3269 |
-
"grad_norm": 12.143242835998535,
|
| 3270 |
-
"learning_rate": 5.334680550398852e-06,
|
| 3271 |
-
"loss": 10.075709342956543,
|
| 3272 |
-
"step": 466
|
| 3273 |
-
},
|
| 3274 |
-
{
|
| 3275 |
-
"epoch": 0.934,
|
| 3276 |
-
"grad_norm": 7.087433815002441,
|
| 3277 |
-
"learning_rate": 5.0354639869588e-06,
|
| 3278 |
-
"loss": 8.941017150878906,
|
| 3279 |
-
"step": 467
|
| 3280 |
-
},
|
| 3281 |
-
{
|
| 3282 |
-
"epoch": 0.936,
|
| 3283 |
-
"grad_norm": 10.367242813110352,
|
| 3284 |
-
"learning_rate": 4.744775803921475e-06,
|
| 3285 |
-
"loss": 9.039401054382324,
|
| 3286 |
-
"step": 468
|
| 3287 |
-
},
|
| 3288 |
-
{
|
| 3289 |
-
"epoch": 0.938,
|
| 3290 |
-
"grad_norm": 8.92366886138916,
|
| 3291 |
-
"learning_rate": 4.4626287169296846e-06,
|
| 3292 |
-
"loss": 9.68399715423584,
|
| 3293 |
-
"step": 469
|
| 3294 |
-
},
|
| 3295 |
-
{
|
| 3296 |
-
"epoch": 0.94,
|
| 3297 |
-
"grad_norm": 15.21575927734375,
|
| 3298 |
-
"learning_rate": 4.189035068011071e-06,
|
| 3299 |
-
"loss": 10.225751876831055,
|
| 3300 |
-
"step": 470
|
| 3301 |
-
},
|
| 3302 |
-
{
|
| 3303 |
-
"epoch": 0.942,
|
| 3304 |
-
"grad_norm": 11.650805473327637,
|
| 3305 |
-
"learning_rate": 3.924006825038129e-06,
|
| 3306 |
-
"loss": 9.687941551208496,
|
| 3307 |
-
"step": 471
|
| 3308 |
-
},
|
| 3309 |
-
{
|
| 3310 |
-
"epoch": 0.944,
|
| 3311 |
-
"grad_norm": 10.030238151550293,
|
| 3312 |
-
"learning_rate": 3.6675555812047956e-06,
|
| 3313 |
-
"loss": 9.854727745056152,
|
| 3314 |
-
"step": 472
|
| 3315 |
-
},
|
| 3316 |
-
{
|
| 3317 |
-
"epoch": 0.946,
|
| 3318 |
-
"grad_norm": 11.019728660583496,
|
| 3319 |
-
"learning_rate": 3.4196925545192604e-06,
|
| 3320 |
-
"loss": 9.399672508239746,
|
| 3321 |
-
"step": 473
|
| 3322 |
-
},
|
| 3323 |
-
{
|
| 3324 |
-
"epoch": 0.948,
|
| 3325 |
-
"grad_norm": 12.790410041809082,
|
| 3326 |
-
"learning_rate": 3.1804285873132668e-06,
|
| 3327 |
-
"loss": 9.853860855102539,
|
| 3328 |
-
"step": 474
|
| 3329 |
-
},
|
| 3330 |
-
{
|
| 3331 |
-
"epoch": 0.95,
|
| 3332 |
-
"grad_norm": 12.762582778930664,
|
| 3333 |
-
"learning_rate": 2.9497741457678695e-06,
|
| 3334 |
-
"loss": 9.703636169433594,
|
| 3335 |
-
"step": 475
|
| 3336 |
-
},
|
| 3337 |
-
{
|
| 3338 |
-
"epoch": 0.952,
|
| 3339 |
-
"grad_norm": 15.26113224029541,
|
| 3340 |
-
"learning_rate": 2.7277393194555358e-06,
|
| 3341 |
-
"loss": 10.118289947509766,
|
| 3342 |
-
"step": 476
|
| 3343 |
-
},
|
| 3344 |
-
{
|
| 3345 |
-
"epoch": 0.954,
|
| 3346 |
-
"grad_norm": 25.00994873046875,
|
| 3347 |
-
"learning_rate": 2.514333820898851e-06,
|
| 3348 |
-
"loss": 10.311869621276855,
|
| 3349 |
-
"step": 477
|
| 3350 |
-
},
|
| 3351 |
-
{
|
| 3352 |
-
"epoch": 0.956,
|
| 3353 |
-
"grad_norm": 22.177448272705078,
|
| 3354 |
-
"learning_rate": 2.3095669851457056e-06,
|
| 3355 |
-
"loss": 10.860949516296387,
|
| 3356 |
-
"step": 478
|
| 3357 |
-
},
|
| 3358 |
-
{
|
| 3359 |
-
"epoch": 0.958,
|
| 3360 |
-
"grad_norm": 28.116384506225586,
|
| 3361 |
-
"learning_rate": 2.113447769360799e-06,
|
| 3362 |
-
"loss": 11.024401664733887,
|
| 3363 |
-
"step": 479
|
| 3364 |
-
},
|
| 3365 |
-
{
|
| 3366 |
-
"epoch": 0.96,
|
| 3367 |
-
"grad_norm": 23.95313262939453,
|
| 3368 |
-
"learning_rate": 1.925984752434018e-06,
|
| 3369 |
-
"loss": 11.532613754272461,
|
| 3370 |
-
"step": 480
|
| 3371 |
-
},
|
| 3372 |
-
{
|
| 3373 |
-
"epoch": 0.962,
|
| 3374 |
-
"grad_norm": 25.587406158447266,
|
| 3375 |
-
"learning_rate": 1.7471861346049612e-06,
|
| 3376 |
-
"loss": 10.99028205871582,
|
| 3377 |
-
"step": 481
|
| 3378 |
-
},
|
| 3379 |
-
{
|
| 3380 |
-
"epoch": 0.964,
|
| 3381 |
-
"grad_norm": 32.761314392089844,
|
| 3382 |
-
"learning_rate": 1.577059737104447e-06,
|
| 3383 |
-
"loss": 11.86240005493164,
|
| 3384 |
-
"step": 482
|
| 3385 |
-
},
|
| 3386 |
-
{
|
| 3387 |
-
"epoch": 0.966,
|
| 3388 |
-
"grad_norm": 23.345834732055664,
|
| 3389 |
-
"learning_rate": 1.4156130018121882e-06,
|
| 3390 |
-
"loss": 10.85324764251709,
|
| 3391 |
-
"step": 483
|
| 3392 |
-
},
|
| 3393 |
-
{
|
| 3394 |
-
"epoch": 0.968,
|
| 3395 |
-
"grad_norm": 21.13359832763672,
|
| 3396 |
-
"learning_rate": 1.2628529909314512e-06,
|
| 3397 |
-
"loss": 10.59592056274414,
|
| 3398 |
-
"step": 484
|
| 3399 |
-
},
|
| 3400 |
-
{
|
| 3401 |
-
"epoch": 0.97,
|
| 3402 |
-
"grad_norm": 19.342931747436523,
|
| 3403 |
-
"learning_rate": 1.1187863866799709e-06,
|
| 3404 |
-
"loss": 10.746866226196289,
|
| 3405 |
-
"step": 485
|
| 3406 |
-
},
|
| 3407 |
-
{
|
| 3408 |
-
"epoch": 0.972,
|
| 3409 |
-
"grad_norm": 25.432170867919922,
|
| 3410 |
-
"learning_rate": 9.834194909977169e-07,
|
| 3411 |
-
"loss": 10.924625396728516,
|
| 3412 |
-
"step": 486
|
| 3413 |
-
},
|
| 3414 |
-
{
|
| 3415 |
-
"epoch": 0.974,
|
| 3416 |
-
"grad_norm": 25.994157791137695,
|
| 3417 |
-
"learning_rate": 8.567582252712481e-07,
|
| 3418 |
-
"loss": 11.445934295654297,
|
| 3419 |
-
"step": 487
|
| 3420 |
-
},
|
| 3421 |
-
{
|
| 3422 |
-
"epoch": 0.976,
|
| 3423 |
-
"grad_norm": 50.760746002197266,
|
| 3424 |
-
"learning_rate": 7.388081300746308e-07,
|
| 3425 |
-
"loss": 12.419108390808105,
|
| 3426 |
-
"step": 488
|
| 3427 |
-
},
|
| 3428 |
-
{
|
| 3429 |
-
"epoch": 0.978,
|
| 3430 |
-
"grad_norm": 30.54327964782715,
|
| 3431 |
-
"learning_rate": 6.295743649271435e-07,
|
| 3432 |
-
"loss": 11.308867454528809,
|
| 3433 |
-
"step": 489
|
| 3434 |
-
},
|
| 3435 |
-
{
|
| 3436 |
-
"epoch": 0.98,
|
| 3437 |
-
"grad_norm": 43.5800895690918,
|
| 3438 |
-
"learning_rate": 5.290617080675686e-07,
|
| 3439 |
-
"loss": 12.132267951965332,
|
| 3440 |
-
"step": 490
|
| 3441 |
-
},
|
| 3442 |
-
{
|
| 3443 |
-
"epoch": 0.982,
|
| 3444 |
-
"grad_norm": 45.79883575439453,
|
| 3445 |
-
"learning_rate": 4.3727455624509307e-07,
|
| 3446 |
-
"loss": 12.2587308883667,
|
| 3447 |
-
"step": 491
|
| 3448 |
-
},
|
| 3449 |
-
{
|
| 3450 |
-
"epoch": 0.984,
|
| 3451 |
-
"grad_norm": 27.356428146362305,
|
| 3452 |
-
"learning_rate": 3.5421692452710656e-07,
|
| 3453 |
-
"loss": 12.494772911071777,
|
| 3454 |
-
"step": 492
|
| 3455 |
-
},
|
| 3456 |
-
{
|
| 3457 |
-
"epoch": 0.986,
|
| 3458 |
-
"grad_norm": 20.774913787841797,
|
| 3459 |
-
"learning_rate": 2.798924461234753e-07,
|
| 3460 |
-
"loss": 11.75953197479248,
|
| 3461 |
-
"step": 493
|
| 3462 |
-
},
|
| 3463 |
-
{
|
| 3464 |
-
"epoch": 0.988,
|
| 3465 |
-
"grad_norm": 53.61131286621094,
|
| 3466 |
-
"learning_rate": 2.1430437222769163e-07,
|
| 3467 |
-
"loss": 12.5416898727417,
|
| 3468 |
-
"step": 494
|
| 3469 |
-
},
|
| 3470 |
-
{
|
| 3471 |
-
"epoch": 0.99,
|
| 3472 |
-
"grad_norm": 37.316715240478516,
|
| 3473 |
-
"learning_rate": 1.5745557187456515e-07,
|
| 3474 |
-
"loss": 12.78895378112793,
|
| 3475 |
-
"step": 495
|
| 3476 |
-
},
|
| 3477 |
-
{
|
| 3478 |
-
"epoch": 0.992,
|
| 3479 |
-
"grad_norm": 33.00537109375,
|
| 3480 |
-
"learning_rate": 1.0934853181479022e-07,
|
| 3481 |
-
"loss": 12.693473815917969,
|
| 3482 |
-
"step": 496
|
| 3483 |
-
},
|
| 3484 |
-
{
|
| 3485 |
-
"epoch": 0.994,
|
| 3486 |
-
"grad_norm": 23.76810073852539,
|
| 3487 |
-
"learning_rate": 6.998535640614368e-08,
|
| 3488 |
-
"loss": 11.896931648254395,
|
| 3489 |
-
"step": 497
|
| 3490 |
-
},
|
| 3491 |
-
{
|
| 3492 |
-
"epoch": 0.996,
|
| 3493 |
-
"grad_norm": 46.77400207519531,
|
| 3494 |
-
"learning_rate": 3.9367767521447664e-08,
|
| 3495 |
-
"loss": 12.127659797668457,
|
| 3496 |
-
"step": 498
|
| 3497 |
-
},
|
| 3498 |
-
{
|
| 3499 |
-
"epoch": 0.998,
|
| 3500 |
-
"grad_norm": 54.48703384399414,
|
| 3501 |
-
"learning_rate": 1.7497104473207514e-08,
|
| 3502 |
-
"loss": 12.056303977966309,
|
| 3503 |
-
"step": 499
|
| 3504 |
},
|
| 3505 |
{
|
| 3506 |
-
"epoch":
|
| 3507 |
-
"grad_norm":
|
| 3508 |
-
"learning_rate":
|
| 3509 |
-
"loss":
|
| 3510 |
"step": 500
|
| 3511 |
},
|
| 3512 |
{
|
| 3513 |
-
"epoch":
|
| 3514 |
-
"eval_accuracy": 0.
|
| 3515 |
-
"eval_loss":
|
| 3516 |
-
"eval_runtime":
|
| 3517 |
-
"eval_samples_per_second":
|
| 3518 |
-
"eval_steps_per_second":
|
| 3519 |
"step": 500
|
| 3520 |
}
|
| 3521 |
],
|
| 3522 |
-
"logging_steps":
|
| 3523 |
-
"max_steps":
|
| 3524 |
"num_input_tokens_seen": 0,
|
| 3525 |
"num_train_epochs": 9223372036854775807,
|
| 3526 |
-
"save_steps":
|
| 3527 |
"stateful_callbacks": {
|
| 3528 |
"TrainerControl": {
|
| 3529 |
"args": {
|
|
@@ -3531,7 +102,7 @@
|
|
| 3531 |
"should_evaluate": false,
|
| 3532 |
"should_log": false,
|
| 3533 |
"should_save": true,
|
| 3534 |
-
"should_training_stop":
|
| 3535 |
},
|
| 3536 |
"attributes": {}
|
| 3537 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
"global_step": 500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.1,
|
| 14 |
+
"grad_norm": 2.3951845169067383,
|
| 15 |
+
"learning_rate": 0.0003973800426880847,
|
| 16 |
+
"loss": 9.800797119140626,
|
| 17 |
+
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
},
|
| 19 |
{
|
| 20 |
+
"epoch": 0.1,
|
| 21 |
+
"eval_accuracy": 0.09641291585127201,
|
| 22 |
+
"eval_loss": 9.091407775878906,
|
| 23 |
+
"eval_runtime": 39.6773,
|
| 24 |
+
"eval_samples_per_second": 25.203,
|
| 25 |
+
"eval_steps_per_second": 25.203,
|
| 26 |
+
"step": 100
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.2,
|
| 30 |
+
"grad_norm": 2.5533459186553955,
|
| 31 |
+
"learning_rate": 0.0003762085737488283,
|
| 32 |
+
"loss": 8.910618286132813,
|
| 33 |
+
"step": 200
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.2,
|
| 37 |
+
"eval_accuracy": 0.09622896281800392,
|
| 38 |
+
"eval_loss": 8.752147674560547,
|
| 39 |
+
"eval_runtime": 33.1575,
|
| 40 |
+
"eval_samples_per_second": 30.159,
|
| 41 |
+
"eval_steps_per_second": 30.159,
|
| 42 |
+
"step": 200
|
| 43 |
},
|
| 44 |
{
|
| 45 |
+
"epoch": 0.3,
|
| 46 |
+
"grad_norm": 2.921639919281006,
|
| 47 |
+
"learning_rate": 0.00033594217168615465,
|
| 48 |
+
"loss": 8.53966796875,
|
| 49 |
+
"step": 300
|
| 50 |
},
|
| 51 |
{
|
| 52 |
+
"epoch": 0.3,
|
| 53 |
+
"eval_accuracy": 0.0970450097847358,
|
| 54 |
+
"eval_loss": 8.573400497436523,
|
| 55 |
+
"eval_runtime": 31.8348,
|
| 56 |
+
"eval_samples_per_second": 31.412,
|
| 57 |
+
"eval_steps_per_second": 31.412,
|
| 58 |
+
"step": 300
|
| 59 |
},
|
| 60 |
{
|
| 61 |
+
"epoch": 0.4,
|
| 62 |
+
"grad_norm": 2.1571640968322754,
|
| 63 |
+
"learning_rate": 0.00028094432596115747,
|
| 64 |
+
"loss": 8.399310302734374,
|
| 65 |
"step": 400
|
| 66 |
},
|
| 67 |
{
|
| 68 |
+
"epoch": 0.4,
|
| 69 |
+
"eval_accuracy": 0.10219373776908024,
|
| 70 |
+
"eval_loss": 8.501660346984863,
|
| 71 |
+
"eval_runtime": 32.4742,
|
| 72 |
+
"eval_samples_per_second": 30.794,
|
| 73 |
+
"eval_steps_per_second": 30.794,
|
| 74 |
+
"step": 400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.5,
|
| 78 |
+
"grad_norm": 2.42037034034729,
|
| 79 |
+
"learning_rate": 0.00021717490653764342,
|
| 80 |
+
"loss": 8.401516723632813,
|
| 81 |
"step": 500
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.5,
|
| 85 |
+
"eval_accuracy": 0.10070058708414872,
|
| 86 |
+
"eval_loss": 8.452370643615723,
|
| 87 |
+
"eval_runtime": 32.4058,
|
| 88 |
+
"eval_samples_per_second": 30.859,
|
| 89 |
+
"eval_steps_per_second": 30.859,
|
| 90 |
"step": 500
|
| 91 |
}
|
| 92 |
],
|
| 93 |
+
"logging_steps": 100,
|
| 94 |
+
"max_steps": 1000,
|
| 95 |
"num_input_tokens_seen": 0,
|
| 96 |
"num_train_epochs": 9223372036854775807,
|
| 97 |
+
"save_steps": 500,
|
| 98 |
"stateful_callbacks": {
|
| 99 |
"TrainerControl": {
|
| 100 |
"args": {
|
|
|
|
| 102 |
"should_evaluate": false,
|
| 103 |
"should_log": false,
|
| 104 |
"should_save": true,
|
| 105 |
+
"should_training_stop": false
|
| 106 |
},
|
| 107 |
"attributes": {}
|
| 108 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5329
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f0098da0a3455c5880575320ae51291e35e8d0961f89a145f0520da235cee9b
|
| 3 |
size 5329
|