prune 5 old checkpoints
Browse files- slots/0/checkpoint-4078/config.json +0 -24
- slots/0/checkpoint-4078/trainer_state.json +0 -1462
- slots/12/checkpoint-4488/config.json +0 -24
- slots/12/checkpoint-4488/trainer_state.json +0 -1609
- slots/16/checkpoint-1731/config.json +0 -24
- slots/16/checkpoint-1731/trainer_state.json +0 -643
- slots/4/checkpoint-1539/config.json +0 -24
- slots/4/checkpoint-1539/trainer_state.json +0 -573
- slots/8/checkpoint-2230/config.json +0 -24
- slots/8/checkpoint-2230/trainer_state.json +0 -818
slots/0/checkpoint-4078/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/0/checkpoint-4078/trainer_state.json
DELETED
|
@@ -1,1462 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.545200832240113,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 4078,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001336931908386741,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 129.80032348632812,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.002673863816773482,
|
| 21 |
-
"grad_norm": 63.64365768432617,
|
| 22 |
-
"learning_rate": 4.999995392022967e-05,
|
| 23 |
-
"loss": 63.88374408922697,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.005347727633546964,
|
| 28 |
-
"grad_norm": 24.627853393554688,
|
| 29 |
-
"learning_rate": 4.999976672145381e-05,
|
| 30 |
-
"loss": 12.65963363647461,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008021591450320446,
|
| 35 |
-
"grad_norm": 14.29983901977539,
|
| 36 |
-
"learning_rate": 4.999943552476422e-05,
|
| 37 |
-
"loss": 5.90204963684082,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.010695455267093928,
|
| 42 |
-
"grad_norm": 15.690323829650879,
|
| 43 |
-
"learning_rate": 4.999896033206858e-05,
|
| 44 |
-
"loss": 3.9918922424316405,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.01336931908386741,
|
| 49 |
-
"grad_norm": 31.583160400390625,
|
| 50 |
-
"learning_rate": 4.999834114610398e-05,
|
| 51 |
-
"loss": 2.9675426483154297,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.01604318290064089,
|
| 56 |
-
"grad_norm": 13.034649848937988,
|
| 57 |
-
"learning_rate": 4.999757797043691e-05,
|
| 58 |
-
"loss": 2.725296401977539,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.018717046717414372,
|
| 63 |
-
"grad_norm": 8.362203598022461,
|
| 64 |
-
"learning_rate": 4.999667080946324e-05,
|
| 65 |
-
"loss": 2.2478992462158205,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.021390910534187856,
|
| 70 |
-
"grad_norm": 8.726786613464355,
|
| 71 |
-
"learning_rate": 4.999561966840821e-05,
|
| 72 |
-
"loss": 1.8447845458984375,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.024064774350961337,
|
| 77 |
-
"grad_norm": 10.092752456665039,
|
| 78 |
-
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
-
"loss": 1.5611843109130858,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02673863816773482,
|
| 84 |
-
"grad_norm": 9.090085983276367,
|
| 85 |
-
"learning_rate": 4.999308547110146e-05,
|
| 86 |
-
"loss": 1.520334815979004,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.029412501984508302,
|
| 91 |
-
"grad_norm": 9.668124198913574,
|
| 92 |
-
"learning_rate": 4.999160242944665e-05,
|
| 93 |
-
"loss": 1.2818055152893066,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.03208636580128178,
|
| 98 |
-
"grad_norm": 9.182533264160156,
|
| 99 |
-
"learning_rate": 4.998997543690418e-05,
|
| 100 |
-
"loss": 1.0428407669067383,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.03476022961805526,
|
| 105 |
-
"grad_norm": 5.745838165283203,
|
| 106 |
-
"learning_rate": 4.998820450284549e-05,
|
| 107 |
-
"loss": 1.2343652725219727,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.037434093434828744,
|
| 112 |
-
"grad_norm": 8.651643753051758,
|
| 113 |
-
"learning_rate": 4.99862896374711e-05,
|
| 114 |
-
"loss": 0.8859601020812988,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04010795725160223,
|
| 119 |
-
"grad_norm": 10.765266418457031,
|
| 120 |
-
"learning_rate": 4.998423085181056e-05,
|
| 121 |
-
"loss": 0.989600658416748,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04278182106837571,
|
| 126 |
-
"grad_norm": 6.092499256134033,
|
| 127 |
-
"learning_rate": 4.998202815772245e-05,
|
| 128 |
-
"loss": 0.7189463615417481,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.04545568488514919,
|
| 133 |
-
"grad_norm": 6.352876663208008,
|
| 134 |
-
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
-
"loss": 0.7489545345306396,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.048129548701922674,
|
| 140 |
-
"grad_norm": 4.620656490325928,
|
| 141 |
-
"learning_rate": 4.997719109584209e-05,
|
| 142 |
-
"loss": 0.7381401538848877,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.050803412518696155,
|
| 147 |
-
"grad_norm": 7.796917915344238,
|
| 148 |
-
"learning_rate": 4.997455675591119e-05,
|
| 149 |
-
"loss": 0.5687405109405518,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.05347727633546964,
|
| 154 |
-
"grad_norm": 2.837172508239746,
|
| 155 |
-
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
-
"loss": 0.5686865329742432,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.05615114015224312,
|
| 161 |
-
"grad_norm": 3.3103690147399902,
|
| 162 |
-
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
-
"loss": 0.625730562210083,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.058825003969016604,
|
| 168 |
-
"grad_norm": 3.5682132244110107,
|
| 169 |
-
"learning_rate": 4.99657906847257e-05,
|
| 170 |
-
"loss": 0.6125466346740722,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.061498867785790085,
|
| 175 |
-
"grad_norm": 5.63640832901001,
|
| 176 |
-
"learning_rate": 4.996258103330218e-05,
|
| 177 |
-
"loss": 0.6182214260101319,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.06417273160256357,
|
| 182 |
-
"grad_norm": 4.698945999145508,
|
| 183 |
-
"learning_rate": 4.995922759815339e-05,
|
| 184 |
-
"loss": 0.43828091621398924,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.06684659541933705,
|
| 189 |
-
"grad_norm": 2.1976189613342285,
|
| 190 |
-
"learning_rate": 4.995573039859501e-05,
|
| 191 |
-
"loss": 0.4459230899810791,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.06952045923611053,
|
| 196 |
-
"grad_norm": 3.8809523582458496,
|
| 197 |
-
"learning_rate": 4.995208945477081e-05,
|
| 198 |
-
"loss": 0.3821882963180542,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.07219432305288401,
|
| 203 |
-
"grad_norm": 3.75144100189209,
|
| 204 |
-
"learning_rate": 4.994830478765251e-05,
|
| 205 |
-
"loss": 0.5800807476043701,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.07486818686965749,
|
| 210 |
-
"grad_norm": 3.0038585662841797,
|
| 211 |
-
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
-
"loss": 0.3928264617919922,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.07754205068643098,
|
| 217 |
-
"grad_norm": 3.614591598510742,
|
| 218 |
-
"learning_rate": 4.994030437155961e-05,
|
| 219 |
-
"loss": 0.48637890815734863,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.08021591450320446,
|
| 224 |
-
"grad_norm": 4.143443584442139,
|
| 225 |
-
"learning_rate": 4.993608866866718e-05,
|
| 226 |
-
"loss": 0.3650153160095215,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.08288977831997794,
|
| 231 |
-
"grad_norm": 6.692712783813477,
|
| 232 |
-
"learning_rate": 4.993172933464471e-05,
|
| 233 |
-
"loss": 0.3677916288375854,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.08556364213675142,
|
| 238 |
-
"grad_norm": 8.383441925048828,
|
| 239 |
-
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
-
"loss": 0.3399480104446411,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.0882375059535249,
|
| 245 |
-
"grad_norm": 5.566338062286377,
|
| 246 |
-
"learning_rate": 4.992257987447532e-05,
|
| 247 |
-
"loss": 0.28104052543640134,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.09091136977029839,
|
| 252 |
-
"grad_norm": 3.1196420192718506,
|
| 253 |
-
"learning_rate": 4.991778980102904e-05,
|
| 254 |
-
"loss": 0.351950478553772,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.09358523358707187,
|
| 259 |
-
"grad_norm": 3.47979736328125,
|
| 260 |
-
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
-
"loss": 0.27501535415649414,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.09625909740384535,
|
| 266 |
-
"grad_norm": 5.446717262268066,
|
| 267 |
-
"learning_rate": 4.990777910536653e-05,
|
| 268 |
-
"loss": 0.2651593923568726,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.09893296122061883,
|
| 273 |
-
"grad_norm": 7.6145339012146,
|
| 274 |
-
"learning_rate": 4.990255854081161e-05,
|
| 275 |
-
"loss": 0.35140380859375,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.10160682503739231,
|
| 280 |
-
"grad_norm": 8.445616722106934,
|
| 281 |
-
"learning_rate": 4.989719453825918e-05,
|
| 282 |
-
"loss": 0.2961219072341919,
|
| 283 |
-
"step": 760
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.10428068885416579,
|
| 287 |
-
"grad_norm": 6.339537620544434,
|
| 288 |
-
"learning_rate": 4.9891687128605744e-05,
|
| 289 |
-
"loss": 0.24962289333343507,
|
| 290 |
-
"step": 780
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.10695455267093928,
|
| 294 |
-
"grad_norm": 3.3369436264038086,
|
| 295 |
-
"learning_rate": 4.988603634357383e-05,
|
| 296 |
-
"loss": 0.2124847412109375,
|
| 297 |
-
"step": 800
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.10962841648771277,
|
| 301 |
-
"grad_norm": 2.2909045219421387,
|
| 302 |
-
"learning_rate": 4.988024221571177e-05,
|
| 303 |
-
"loss": 0.24679112434387207,
|
| 304 |
-
"step": 820
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.11230228030448625,
|
| 308 |
-
"grad_norm": 3.1149911880493164,
|
| 309 |
-
"learning_rate": 4.9874304778393574e-05,
|
| 310 |
-
"loss": 0.22161397933959961,
|
| 311 |
-
"step": 840
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.11497614412125973,
|
| 315 |
-
"grad_norm": 14.802160263061523,
|
| 316 |
-
"learning_rate": 4.9868224065818706e-05,
|
| 317 |
-
"loss": 0.2623537302017212,
|
| 318 |
-
"step": 860
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.11765000793803321,
|
| 322 |
-
"grad_norm": 5.586325168609619,
|
| 323 |
-
"learning_rate": 4.98620001130119e-05,
|
| 324 |
-
"loss": 0.3560942649841309,
|
| 325 |
-
"step": 880
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.12032387175480669,
|
| 329 |
-
"grad_norm": 3.390017032623291,
|
| 330 |
-
"learning_rate": 4.9855632955822916e-05,
|
| 331 |
-
"loss": 0.16934787034988402,
|
| 332 |
-
"step": 900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.12299773557158017,
|
| 336 |
-
"grad_norm": 6.070940971374512,
|
| 337 |
-
"learning_rate": 4.984912263092641e-05,
|
| 338 |
-
"loss": 0.2131197214126587,
|
| 339 |
-
"step": 920
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.12567159938835365,
|
| 343 |
-
"grad_norm": 1.4912281036376953,
|
| 344 |
-
"learning_rate": 4.984246917582166e-05,
|
| 345 |
-
"loss": 0.25128653049468996,
|
| 346 |
-
"step": 940
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.12834546320512713,
|
| 350 |
-
"grad_norm": 7.000472545623779,
|
| 351 |
-
"learning_rate": 4.9835672628832366e-05,
|
| 352 |
-
"loss": 0.2653592586517334,
|
| 353 |
-
"step": 960
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.1310193270219006,
|
| 357 |
-
"grad_norm": 5.427223205566406,
|
| 358 |
-
"learning_rate": 4.9828733029106434e-05,
|
| 359 |
-
"loss": 0.1653295636177063,
|
| 360 |
-
"step": 980
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.1336931908386741,
|
| 364 |
-
"grad_norm": 1.9502102136611938,
|
| 365 |
-
"learning_rate": 4.982165041661575e-05,
|
| 366 |
-
"loss": 0.2250870943069458,
|
| 367 |
-
"step": 1000
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.13636705465544757,
|
| 371 |
-
"grad_norm": 0.6216259598731995,
|
| 372 |
-
"learning_rate": 4.981442483215595e-05,
|
| 373 |
-
"loss": 0.18943849802017212,
|
| 374 |
-
"step": 1020
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.13904091847222105,
|
| 378 |
-
"grad_norm": 2.3363687992095947,
|
| 379 |
-
"learning_rate": 4.98070563173462e-05,
|
| 380 |
-
"loss": 0.1673592209815979,
|
| 381 |
-
"step": 1040
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.14171478228899453,
|
| 385 |
-
"grad_norm": 1.040717601776123,
|
| 386 |
-
"learning_rate": 4.979954491462892e-05,
|
| 387 |
-
"loss": 0.2113173007965088,
|
| 388 |
-
"step": 1060
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.14438864610576801,
|
| 392 |
-
"grad_norm": 2.735522747039795,
|
| 393 |
-
"learning_rate": 4.979189066726955e-05,
|
| 394 |
-
"loss": 0.17504971027374266,
|
| 395 |
-
"step": 1080
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.1470625099225415,
|
| 399 |
-
"grad_norm": 4.701151371002197,
|
| 400 |
-
"learning_rate": 4.978409361935636e-05,
|
| 401 |
-
"loss": 0.15881222486495972,
|
| 402 |
-
"step": 1100
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.14973637373931498,
|
| 406 |
-
"grad_norm": 2.735919237136841,
|
| 407 |
-
"learning_rate": 4.9776153815800075e-05,
|
| 408 |
-
"loss": 0.14044179916381835,
|
| 409 |
-
"step": 1120
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.15241023755608848,
|
| 413 |
-
"grad_norm": 3.5479538440704346,
|
| 414 |
-
"learning_rate": 4.976807130233375e-05,
|
| 415 |
-
"loss": 0.18565714359283447,
|
| 416 |
-
"step": 1140
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.15508410137286197,
|
| 420 |
-
"grad_norm": 3.2167458534240723,
|
| 421 |
-
"learning_rate": 4.975984612551243e-05,
|
| 422 |
-
"loss": 0.13236271142959594,
|
| 423 |
-
"step": 1160
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.15775796518963545,
|
| 427 |
-
"grad_norm": 1.0206760168075562,
|
| 428 |
-
"learning_rate": 4.975147833271288e-05,
|
| 429 |
-
"loss": 0.19124728441238403,
|
| 430 |
-
"step": 1180
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.16043182900640893,
|
| 434 |
-
"grad_norm": 4.194457530975342,
|
| 435 |
-
"learning_rate": 4.9742967972133335e-05,
|
| 436 |
-
"loss": 0.144741427898407,
|
| 437 |
-
"step": 1200
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.1631056928231824,
|
| 441 |
-
"grad_norm": 3.0225746631622314,
|
| 442 |
-
"learning_rate": 4.973431509279323e-05,
|
| 443 |
-
"loss": 0.1374324679374695,
|
| 444 |
-
"step": 1220
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.1657795566399559,
|
| 448 |
-
"grad_norm": 4.243523120880127,
|
| 449 |
-
"learning_rate": 4.972551974453287e-05,
|
| 450 |
-
"loss": 0.13663809299468993,
|
| 451 |
-
"step": 1240
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.16845342045672937,
|
| 455 |
-
"grad_norm": 2.4990086555480957,
|
| 456 |
-
"learning_rate": 4.971658197801322e-05,
|
| 457 |
-
"loss": 0.16817957162857056,
|
| 458 |
-
"step": 1260
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.17112728427350285,
|
| 462 |
-
"grad_norm": 4.983982563018799,
|
| 463 |
-
"learning_rate": 4.9707501844715554e-05,
|
| 464 |
-
"loss": 0.13795313835144044,
|
| 465 |
-
"step": 1280
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.17380114809027633,
|
| 469 |
-
"grad_norm": 3.6780316829681396,
|
| 470 |
-
"learning_rate": 4.969827939694115e-05,
|
| 471 |
-
"loss": 0.1637880802154541,
|
| 472 |
-
"step": 1300
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.1764750119070498,
|
| 476 |
-
"grad_norm": 0.7950732707977295,
|
| 477 |
-
"learning_rate": 4.968891468781105e-05,
|
| 478 |
-
"loss": 0.10979138612747193,
|
| 479 |
-
"step": 1320
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.1791488757238233,
|
| 483 |
-
"grad_norm": 1.2414121627807617,
|
| 484 |
-
"learning_rate": 4.967940777126569e-05,
|
| 485 |
-
"loss": 0.13692171573638917,
|
| 486 |
-
"step": 1340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.18182273954059677,
|
| 490 |
-
"grad_norm": 2.1383633613586426,
|
| 491 |
-
"learning_rate": 4.9669758702064636e-05,
|
| 492 |
-
"loss": 0.07821698188781738,
|
| 493 |
-
"step": 1360
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.18449660335737025,
|
| 497 |
-
"grad_norm": 5.061275959014893,
|
| 498 |
-
"learning_rate": 4.965996753578623e-05,
|
| 499 |
-
"loss": 0.19053516387939454,
|
| 500 |
-
"step": 1380
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.18717046717414373,
|
| 504 |
-
"grad_norm": 6.151792049407959,
|
| 505 |
-
"learning_rate": 4.9650034328827305e-05,
|
| 506 |
-
"loss": 0.11360721588134766,
|
| 507 |
-
"step": 1400
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.18984433099091721,
|
| 511 |
-
"grad_norm": 1.0604305267333984,
|
| 512 |
-
"learning_rate": 4.963995913840284e-05,
|
| 513 |
-
"loss": 0.13138024806976317,
|
| 514 |
-
"step": 1420
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.1925181948076907,
|
| 518 |
-
"grad_norm": 1.7159489393234253,
|
| 519 |
-
"learning_rate": 4.9629742022545623e-05,
|
| 520 |
-
"loss": 0.08657677173614502,
|
| 521 |
-
"step": 1440
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.19519205862446418,
|
| 525 |
-
"grad_norm": 2.4207754135131836,
|
| 526 |
-
"learning_rate": 4.961938304010595e-05,
|
| 527 |
-
"loss": 0.10309149026870727,
|
| 528 |
-
"step": 1460
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.19786592244123766,
|
| 532 |
-
"grad_norm": 1.532060146331787,
|
| 533 |
-
"learning_rate": 4.9608882250751245e-05,
|
| 534 |
-
"loss": 0.13628544807434081,
|
| 535 |
-
"step": 1480
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.20053978625801114,
|
| 539 |
-
"grad_norm": 6.409943580627441,
|
| 540 |
-
"learning_rate": 4.959823971496574e-05,
|
| 541 |
-
"loss": 0.10584845542907714,
|
| 542 |
-
"step": 1500
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.20321365007478462,
|
| 546 |
-
"grad_norm": 2.452012538909912,
|
| 547 |
-
"learning_rate": 4.9587455494050136e-05,
|
| 548 |
-
"loss": 0.06506187915802002,
|
| 549 |
-
"step": 1520
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.2058875138915581,
|
| 553 |
-
"grad_norm": 5.3016533851623535,
|
| 554 |
-
"learning_rate": 4.9576529650121214e-05,
|
| 555 |
-
"loss": 0.11848526000976563,
|
| 556 |
-
"step": 1540
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.20856137770833158,
|
| 560 |
-
"grad_norm": 4.341775894165039,
|
| 561 |
-
"learning_rate": 4.956546224611152e-05,
|
| 562 |
-
"loss": 0.11318533420562744,
|
| 563 |
-
"step": 1560
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.21123524152510506,
|
| 567 |
-
"grad_norm": 1.9056169986724854,
|
| 568 |
-
"learning_rate": 4.9554253345768965e-05,
|
| 569 |
-
"loss": 0.12768398523330687,
|
| 570 |
-
"step": 1580
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.21390910534187857,
|
| 574 |
-
"grad_norm": 1.8939746618270874,
|
| 575 |
-
"learning_rate": 4.9542903013656486e-05,
|
| 576 |
-
"loss": 0.10782338380813598,
|
| 577 |
-
"step": 1600
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.21658296915865205,
|
| 581 |
-
"grad_norm": 8.53671932220459,
|
| 582 |
-
"learning_rate": 4.9531411315151654e-05,
|
| 583 |
-
"loss": 0.1733921766281128,
|
| 584 |
-
"step": 1620
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.21925683297542553,
|
| 588 |
-
"grad_norm": 2.0152978897094727,
|
| 589 |
-
"learning_rate": 4.951977831644632e-05,
|
| 590 |
-
"loss": 0.11197054386138916,
|
| 591 |
-
"step": 1640
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.221930696792199,
|
| 595 |
-
"grad_norm": 3.8422367572784424,
|
| 596 |
-
"learning_rate": 4.95080040845462e-05,
|
| 597 |
-
"loss": 0.11441781520843505,
|
| 598 |
-
"step": 1660
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.2246045606089725,
|
| 602 |
-
"grad_norm": 1.819858193397522,
|
| 603 |
-
"learning_rate": 4.949608868727053e-05,
|
| 604 |
-
"loss": 0.11403474807739258,
|
| 605 |
-
"step": 1680
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.22727842442574597,
|
| 609 |
-
"grad_norm": 7.45100212097168,
|
| 610 |
-
"learning_rate": 4.948403219325163e-05,
|
| 611 |
-
"loss": 0.13117753267288207,
|
| 612 |
-
"step": 1700
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.22995228824251945,
|
| 616 |
-
"grad_norm": 0.6526040434837341,
|
| 617 |
-
"learning_rate": 4.947183467193456e-05,
|
| 618 |
-
"loss": 0.07524924874305725,
|
| 619 |
-
"step": 1720
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.23262615205929293,
|
| 623 |
-
"grad_norm": 3.814746856689453,
|
| 624 |
-
"learning_rate": 4.945949619357668e-05,
|
| 625 |
-
"loss": 0.07659345269203185,
|
| 626 |
-
"step": 1740
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.23530001587606642,
|
| 630 |
-
"grad_norm": 2.373124122619629,
|
| 631 |
-
"learning_rate": 4.944701682924726e-05,
|
| 632 |
-
"loss": 0.1147496223449707,
|
| 633 |
-
"step": 1760
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.2379738796928399,
|
| 637 |
-
"grad_norm": 0.11161285638809204,
|
| 638 |
-
"learning_rate": 4.943439665082707e-05,
|
| 639 |
-
"loss": 0.07256829738616943,
|
| 640 |
-
"step": 1780
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.24064774350961338,
|
| 644 |
-
"grad_norm": 0.45990192890167236,
|
| 645 |
-
"learning_rate": 4.942163573100794e-05,
|
| 646 |
-
"loss": 0.07726740837097168,
|
| 647 |
-
"step": 1800
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.24332160732638686,
|
| 651 |
-
"grad_norm": 4.2301926612854,
|
| 652 |
-
"learning_rate": 4.940873414329242e-05,
|
| 653 |
-
"loss": 0.09349535703659058,
|
| 654 |
-
"step": 1820
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.24599547114316034,
|
| 658 |
-
"grad_norm": 2.442178726196289,
|
| 659 |
-
"learning_rate": 4.939569196199325e-05,
|
| 660 |
-
"loss": 0.12413722276687622,
|
| 661 |
-
"step": 1840
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.24866933495993382,
|
| 665 |
-
"grad_norm": 2.523683786392212,
|
| 666 |
-
"learning_rate": 4.938250926223302e-05,
|
| 667 |
-
"loss": 0.08566288352012634,
|
| 668 |
-
"step": 1860
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.2513431987767073,
|
| 672 |
-
"grad_norm": 3.511075258255005,
|
| 673 |
-
"learning_rate": 4.936918611994368e-05,
|
| 674 |
-
"loss": 0.08007702231407166,
|
| 675 |
-
"step": 1880
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.2540170625934808,
|
| 679 |
-
"grad_norm": 6.254627704620361,
|
| 680 |
-
"learning_rate": 4.935572261186614e-05,
|
| 681 |
-
"loss": 0.10983954668045044,
|
| 682 |
-
"step": 1900
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 0.25669092641025426,
|
| 686 |
-
"grad_norm": 1.5211899280548096,
|
| 687 |
-
"learning_rate": 4.934211881554981e-05,
|
| 688 |
-
"loss": 0.09120344519615173,
|
| 689 |
-
"step": 1920
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 0.25936479022702774,
|
| 693 |
-
"grad_norm": 2.5893588066101074,
|
| 694 |
-
"learning_rate": 4.932837480935214e-05,
|
| 695 |
-
"loss": 0.08754412531852722,
|
| 696 |
-
"step": 1940
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 0.2620386540438012,
|
| 700 |
-
"grad_norm": 6.878556251525879,
|
| 701 |
-
"learning_rate": 4.931449067243821e-05,
|
| 702 |
-
"loss": 0.08636274933815002,
|
| 703 |
-
"step": 1960
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 0.2647125178605747,
|
| 707 |
-
"grad_norm": 2.9078798294067383,
|
| 708 |
-
"learning_rate": 4.9300466484780226e-05,
|
| 709 |
-
"loss": 0.09582929015159607,
|
| 710 |
-
"step": 1980
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.2673863816773482,
|
| 714 |
-
"grad_norm": 3.391852855682373,
|
| 715 |
-
"learning_rate": 4.92863023271571e-05,
|
| 716 |
-
"loss": 0.0850919783115387,
|
| 717 |
-
"step": 2000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.27006024549412166,
|
| 721 |
-
"grad_norm": 5.522103309631348,
|
| 722 |
-
"learning_rate": 4.927199828115395e-05,
|
| 723 |
-
"loss": 0.050999772548675534,
|
| 724 |
-
"step": 2020
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 0.27273410931089515,
|
| 728 |
-
"grad_norm": 0.90350741147995,
|
| 729 |
-
"learning_rate": 4.925755442916167e-05,
|
| 730 |
-
"loss": 0.10100446939468384,
|
| 731 |
-
"step": 2040
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 0.2754079731276686,
|
| 735 |
-
"grad_norm": 1.602030634880066,
|
| 736 |
-
"learning_rate": 4.924297085437641e-05,
|
| 737 |
-
"loss": 0.0468633770942688,
|
| 738 |
-
"step": 2060
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 0.2780818369444421,
|
| 742 |
-
"grad_norm": 1.5823460817337036,
|
| 743 |
-
"learning_rate": 4.922824764079913e-05,
|
| 744 |
-
"loss": 0.06786358952522278,
|
| 745 |
-
"step": 2080
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.2807557007612156,
|
| 749 |
-
"grad_norm": 1.6624343395233154,
|
| 750 |
-
"learning_rate": 4.92133848732351e-05,
|
| 751 |
-
"loss": 0.05772828459739685,
|
| 752 |
-
"step": 2100
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 0.28342956457798907,
|
| 756 |
-
"grad_norm": 0.947078287601471,
|
| 757 |
-
"learning_rate": 4.9198382637293424e-05,
|
| 758 |
-
"loss": 0.08012173175811768,
|
| 759 |
-
"step": 2120
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.28610342839476255,
|
| 763 |
-
"grad_norm": 0.2919924259185791,
|
| 764 |
-
"learning_rate": 4.918324101938653e-05,
|
| 765 |
-
"loss": 0.1208539366722107,
|
| 766 |
-
"step": 2140
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.28877729221153603,
|
| 770 |
-
"grad_norm": 9.258247375488281,
|
| 771 |
-
"learning_rate": 4.916796010672969e-05,
|
| 772 |
-
"loss": 0.10037034749984741,
|
| 773 |
-
"step": 2160
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.2914511560283095,
|
| 777 |
-
"grad_norm": 4.0920491218566895,
|
| 778 |
-
"learning_rate": 4.915253998734051e-05,
|
| 779 |
-
"loss": 0.061488878726959226,
|
| 780 |
-
"step": 2180
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 0.294125019845083,
|
| 784 |
-
"grad_norm": 6.1126627922058105,
|
| 785 |
-
"learning_rate": 4.913698075003841e-05,
|
| 786 |
-
"loss": 0.0862967312335968,
|
| 787 |
-
"step": 2200
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.29679888366185647,
|
| 791 |
-
"grad_norm": 2.585484743118286,
|
| 792 |
-
"learning_rate": 4.912128248444414e-05,
|
| 793 |
-
"loss": 0.05393874645233154,
|
| 794 |
-
"step": 2220
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 0.29947274747862995,
|
| 798 |
-
"grad_norm": 6.944481372833252,
|
| 799 |
-
"learning_rate": 4.9105445280979256e-05,
|
| 800 |
-
"loss": 0.08570566773414612,
|
| 801 |
-
"step": 2240
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 0.30214661129540343,
|
| 805 |
-
"grad_norm": 1.3824089765548706,
|
| 806 |
-
"learning_rate": 4.908946923086556e-05,
|
| 807 |
-
"loss": 0.09689127206802368,
|
| 808 |
-
"step": 2260
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"epoch": 0.30482047511217697,
|
| 812 |
-
"grad_norm": 3.4861342906951904,
|
| 813 |
-
"learning_rate": 4.907335442612464e-05,
|
| 814 |
-
"loss": 0.12550976276397705,
|
| 815 |
-
"step": 2280
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 0.30749433892895045,
|
| 819 |
-
"grad_norm": 3.668980121612549,
|
| 820 |
-
"learning_rate": 4.905710095957728e-05,
|
| 821 |
-
"loss": 0.09089353680610657,
|
| 822 |
-
"step": 2300
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.31016820274572393,
|
| 826 |
-
"grad_norm": 1.093095064163208,
|
| 827 |
-
"learning_rate": 4.904070892484298e-05,
|
| 828 |
-
"loss": 0.03925192356109619,
|
| 829 |
-
"step": 2320
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 0.3128420665624974,
|
| 833 |
-
"grad_norm": 0.8169485926628113,
|
| 834 |
-
"learning_rate": 4.9024178416339364e-05,
|
| 835 |
-
"loss": 0.0979581356048584,
|
| 836 |
-
"step": 2340
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 0.3155159303792709,
|
| 840 |
-
"grad_norm": 1.892451286315918,
|
| 841 |
-
"learning_rate": 4.900750952928166e-05,
|
| 842 |
-
"loss": 0.05913209915161133,
|
| 843 |
-
"step": 2360
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 0.3181897941960444,
|
| 847 |
-
"grad_norm": 0.24644255638122559,
|
| 848 |
-
"learning_rate": 4.8990702359682184e-05,
|
| 849 |
-
"loss": 0.06815173625946044,
|
| 850 |
-
"step": 2380
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 0.32086365801281785,
|
| 854 |
-
"grad_norm": 2.1861305236816406,
|
| 855 |
-
"learning_rate": 4.897375700434972e-05,
|
| 856 |
-
"loss": 0.04142785966396332,
|
| 857 |
-
"step": 2400
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 0.32353752182959133,
|
| 861 |
-
"grad_norm": 2.6643004417419434,
|
| 862 |
-
"learning_rate": 4.8956673560889013e-05,
|
| 863 |
-
"loss": 0.05177200436592102,
|
| 864 |
-
"step": 2420
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 0.3262113856463648,
|
| 868 |
-
"grad_norm": 2.588113784790039,
|
| 869 |
-
"learning_rate": 4.8939452127700195e-05,
|
| 870 |
-
"loss": 0.05783546566963196,
|
| 871 |
-
"step": 2440
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.3288852494631383,
|
| 875 |
-
"grad_norm": 2.419644594192505,
|
| 876 |
-
"learning_rate": 4.8922092803978203e-05,
|
| 877 |
-
"loss": 0.08906854391098022,
|
| 878 |
-
"step": 2460
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 0.3315591132799118,
|
| 882 |
-
"grad_norm": 0.16949939727783203,
|
| 883 |
-
"learning_rate": 4.890459568971223e-05,
|
| 884 |
-
"loss": 0.10305211544036866,
|
| 885 |
-
"step": 2480
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 0.33423297709668526,
|
| 889 |
-
"grad_norm": 0.10032984614372253,
|
| 890 |
-
"learning_rate": 4.8886960885685126e-05,
|
| 891 |
-
"loss": 0.06348527669906616,
|
| 892 |
-
"step": 2500
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 0.33690684091345874,
|
| 896 |
-
"grad_norm": 3.3658738136291504,
|
| 897 |
-
"learning_rate": 4.8869188493472854e-05,
|
| 898 |
-
"loss": 0.06826075911521912,
|
| 899 |
-
"step": 2520
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 0.3395807047302322,
|
| 903 |
-
"grad_norm": 0.8656186461448669,
|
| 904 |
-
"learning_rate": 4.885127861544386e-05,
|
| 905 |
-
"loss": 0.05929765701293945,
|
| 906 |
-
"step": 2540
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 0.3422545685470057,
|
| 910 |
-
"grad_norm": 0.1492065042257309,
|
| 911 |
-
"learning_rate": 4.8833231354758496e-05,
|
| 912 |
-
"loss": 0.09429731965065002,
|
| 913 |
-
"step": 2560
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 0.3449284323637792,
|
| 917 |
-
"grad_norm": 0.6010928153991699,
|
| 918 |
-
"learning_rate": 4.881504681536846e-05,
|
| 919 |
-
"loss": 0.06262240409851075,
|
| 920 |
-
"step": 2580
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.34760229618055266,
|
| 924 |
-
"grad_norm": 1.6506450176239014,
|
| 925 |
-
"learning_rate": 4.879672510201616e-05,
|
| 926 |
-
"loss": 0.061688083410263064,
|
| 927 |
-
"step": 2600
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.35027615999732614,
|
| 931 |
-
"grad_norm": 0.2703142464160919,
|
| 932 |
-
"learning_rate": 4.877826632023412e-05,
|
| 933 |
-
"loss": 0.06175137162208557,
|
| 934 |
-
"step": 2620
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 0.3529500238140996,
|
| 938 |
-
"grad_norm": 3.1056365966796875,
|
| 939 |
-
"learning_rate": 4.875967057634437e-05,
|
| 940 |
-
"loss": 0.07828506827354431,
|
| 941 |
-
"step": 2640
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 0.3556238876308731,
|
| 945 |
-
"grad_norm": 0.28790283203125,
|
| 946 |
-
"learning_rate": 4.874093797745784e-05,
|
| 947 |
-
"loss": 0.11355981826782227,
|
| 948 |
-
"step": 2660
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 0.3582977514476466,
|
| 952 |
-
"grad_norm": 2.3372068405151367,
|
| 953 |
-
"learning_rate": 4.8722068631473746e-05,
|
| 954 |
-
"loss": 0.048267871141433716,
|
| 955 |
-
"step": 2680
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 0.36097161526442006,
|
| 959 |
-
"grad_norm": 0.12767371535301208,
|
| 960 |
-
"learning_rate": 4.8703062647078976e-05,
|
| 961 |
-
"loss": 0.04319801032543182,
|
| 962 |
-
"step": 2700
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 0.36364547908119355,
|
| 966 |
-
"grad_norm": 0.5145738124847412,
|
| 967 |
-
"learning_rate": 4.868392013374741e-05,
|
| 968 |
-
"loss": 0.0773090660572052,
|
| 969 |
-
"step": 2720
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 0.366319342897967,
|
| 973 |
-
"grad_norm": 0.8518500328063965,
|
| 974 |
-
"learning_rate": 4.866464120173937e-05,
|
| 975 |
-
"loss": 0.05149460434913635,
|
| 976 |
-
"step": 2740
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 0.3689932067147405,
|
| 980 |
-
"grad_norm": 3.6726584434509277,
|
| 981 |
-
"learning_rate": 4.8645225962100924e-05,
|
| 982 |
-
"loss": 0.06896821856498718,
|
| 983 |
-
"step": 2760
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.371667070531514,
|
| 987 |
-
"grad_norm": 1.5626497268676758,
|
| 988 |
-
"learning_rate": 4.862567452666329e-05,
|
| 989 |
-
"loss": 0.047730174660682675,
|
| 990 |
-
"step": 2780
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 0.37434093434828747,
|
| 994 |
-
"grad_norm": 6.562028884887695,
|
| 995 |
-
"learning_rate": 4.8605987008042144e-05,
|
| 996 |
-
"loss": 0.07060698866844177,
|
| 997 |
-
"step": 2800
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 0.37701479816506095,
|
| 1001 |
-
"grad_norm": 0.7631726861000061,
|
| 1002 |
-
"learning_rate": 4.8586163519637005e-05,
|
| 1003 |
-
"loss": 0.04944324493408203,
|
| 1004 |
-
"step": 2820
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 0.37968866198183443,
|
| 1008 |
-
"grad_norm": 1.6982293128967285,
|
| 1009 |
-
"learning_rate": 4.8566204175630595e-05,
|
| 1010 |
-
"loss": 0.03000348210334778,
|
| 1011 |
-
"step": 2840
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 0.3823625257986079,
|
| 1015 |
-
"grad_norm": 0.6487429141998291,
|
| 1016 |
-
"learning_rate": 4.854610909098812e-05,
|
| 1017 |
-
"loss": 0.06691416501998901,
|
| 1018 |
-
"step": 2860
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 0.3850363896153814,
|
| 1022 |
-
"grad_norm": 0.7648892402648926,
|
| 1023 |
-
"learning_rate": 4.852587838145668e-05,
|
| 1024 |
-
"loss": 0.05529783964157105,
|
| 1025 |
-
"step": 2880
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 0.38771025343215487,
|
| 1029 |
-
"grad_norm": 0.11601298302412033,
|
| 1030 |
-
"learning_rate": 4.850551216356457e-05,
|
| 1031 |
-
"loss": 0.07780832052230835,
|
| 1032 |
-
"step": 2900
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.39038411724892835,
|
| 1036 |
-
"grad_norm": 0.9443137645721436,
|
| 1037 |
-
"learning_rate": 4.8485010554620594e-05,
|
| 1038 |
-
"loss": 0.08007023930549621,
|
| 1039 |
-
"step": 2920
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 0.39305798106570183,
|
| 1043 |
-
"grad_norm": 0.8828252553939819,
|
| 1044 |
-
"learning_rate": 4.846437367271341e-05,
|
| 1045 |
-
"loss": 0.03541453182697296,
|
| 1046 |
-
"step": 2940
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 0.3957318448824753,
|
| 1050 |
-
"grad_norm": 0.21668888628482819,
|
| 1051 |
-
"learning_rate": 4.844360163671083e-05,
|
| 1052 |
-
"loss": 0.08354364633560181,
|
| 1053 |
-
"step": 2960
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 0.3984057086992488,
|
| 1057 |
-
"grad_norm": 0.6840483546257019,
|
| 1058 |
-
"learning_rate": 4.8422694566259194e-05,
|
| 1059 |
-
"loss": 0.045807772874832155,
|
| 1060 |
-
"step": 2980
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 0.4010795725160223,
|
| 1064 |
-
"grad_norm": 1.2754698991775513,
|
| 1065 |
-
"learning_rate": 4.8401652581782584e-05,
|
| 1066 |
-
"loss": 0.053487342596054074,
|
| 1067 |
-
"step": 3000
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 0.40375343633279576,
|
| 1071 |
-
"grad_norm": 0.19012756645679474,
|
| 1072 |
-
"learning_rate": 4.838047580448222e-05,
|
| 1073 |
-
"loss": 0.05881953239440918,
|
| 1074 |
-
"step": 3020
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.40642730014956924,
|
| 1078 |
-
"grad_norm": 2.1057698726654053,
|
| 1079 |
-
"learning_rate": 4.835916435633569e-05,
|
| 1080 |
-
"loss": 0.031065690517425536,
|
| 1081 |
-
"step": 3040
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 0.4091011639663427,
|
| 1085 |
-
"grad_norm": 4.188559055328369,
|
| 1086 |
-
"learning_rate": 4.833771836009633e-05,
|
| 1087 |
-
"loss": 0.07205432653427124,
|
| 1088 |
-
"step": 3060
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.4117750277831162,
|
| 1092 |
-
"grad_norm": 6.975829124450684,
|
| 1093 |
-
"learning_rate": 4.831613793929242e-05,
|
| 1094 |
-
"loss": 0.04953635036945343,
|
| 1095 |
-
"step": 3080
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 0.4144488915998897,
|
| 1099 |
-
"grad_norm": 4.725269317626953,
|
| 1100 |
-
"learning_rate": 4.8294423218226546e-05,
|
| 1101 |
-
"loss": 0.05965519547462463,
|
| 1102 |
-
"step": 3100
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 0.41712275541666316,
|
| 1106 |
-
"grad_norm": 1.7124755382537842,
|
| 1107 |
-
"learning_rate": 4.827257432197486e-05,
|
| 1108 |
-
"loss": 0.039625433087348935,
|
| 1109 |
-
"step": 3120
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 0.41979661923343664,
|
| 1113 |
-
"grad_norm": 2.6687324047088623,
|
| 1114 |
-
"learning_rate": 4.825059137638636e-05,
|
| 1115 |
-
"loss": 0.05020809769630432,
|
| 1116 |
-
"step": 3140
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 0.4224704830502101,
|
| 1120 |
-
"grad_norm": 1.111640214920044,
|
| 1121 |
-
"learning_rate": 4.822847450808215e-05,
|
| 1122 |
-
"loss": 0.04404452443122864,
|
| 1123 |
-
"step": 3160
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 0.42514434686698366,
|
| 1127 |
-
"grad_norm": 0.2128070890903473,
|
| 1128 |
-
"learning_rate": 4.8206223844454744e-05,
|
| 1129 |
-
"loss": 0.08283355236053466,
|
| 1130 |
-
"step": 3180
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.42781821068375714,
|
| 1134 |
-
"grad_norm": 0.10757248103618622,
|
| 1135 |
-
"learning_rate": 4.818383951366729e-05,
|
| 1136 |
-
"loss": 0.08568671345710754,
|
| 1137 |
-
"step": 3200
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.4304920745005306,
|
| 1141 |
-
"grad_norm": 0.08344592899084091,
|
| 1142 |
-
"learning_rate": 4.816132164465289e-05,
|
| 1143 |
-
"loss": 0.0426956832408905,
|
| 1144 |
-
"step": 3220
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.4331659383173041,
|
| 1148 |
-
"grad_norm": 0.5657751560211182,
|
| 1149 |
-
"learning_rate": 4.813867036711378e-05,
|
| 1150 |
-
"loss": 0.04971776902675629,
|
| 1151 |
-
"step": 3240
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.4358398021340776,
|
| 1155 |
-
"grad_norm": 2.1529288291931152,
|
| 1156 |
-
"learning_rate": 4.8115885811520654e-05,
|
| 1157 |
-
"loss": 0.025386181473731995,
|
| 1158 |
-
"step": 3260
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.43851366595085106,
|
| 1162 |
-
"grad_norm": 4.228519916534424,
|
| 1163 |
-
"learning_rate": 4.809296810911188e-05,
|
| 1164 |
-
"loss": 0.06401395201683044,
|
| 1165 |
-
"step": 3280
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.44118752976762454,
|
| 1169 |
-
"grad_norm": 6.770420551300049,
|
| 1170 |
-
"learning_rate": 4.806991739189274e-05,
|
| 1171 |
-
"loss": 0.16425553560256959,
|
| 1172 |
-
"step": 3300
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.443861393584398,
|
| 1176 |
-
"grad_norm": 0.5303187370300293,
|
| 1177 |
-
"learning_rate": 4.804673379263467e-05,
|
| 1178 |
-
"loss": 0.045900467038154605,
|
| 1179 |
-
"step": 3320
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.4465352574011715,
|
| 1183 |
-
"grad_norm": 0.221473827958107,
|
| 1184 |
-
"learning_rate": 4.802341744487453e-05,
|
| 1185 |
-
"loss": 0.07529735565185547,
|
| 1186 |
-
"step": 3340
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.449209121217945,
|
| 1190 |
-
"grad_norm": 3.48736834526062,
|
| 1191 |
-
"learning_rate": 4.799996848291378e-05,
|
| 1192 |
-
"loss": 0.062433135509490964,
|
| 1193 |
-
"step": 3360
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.45188298503471847,
|
| 1197 |
-
"grad_norm": 2.650038242340088,
|
| 1198 |
-
"learning_rate": 4.797638704181774e-05,
|
| 1199 |
-
"loss": 0.03762982189655304,
|
| 1200 |
-
"step": 3380
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 0.45455684885149195,
|
| 1204 |
-
"grad_norm": 3.159665584564209,
|
| 1205 |
-
"learning_rate": 4.795267325741483e-05,
|
| 1206 |
-
"loss": 0.04745924174785614,
|
| 1207 |
-
"step": 3400
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 0.4572307126682654,
|
| 1211 |
-
"grad_norm": 0.8763885498046875,
|
| 1212 |
-
"learning_rate": 4.7928827266295715e-05,
|
| 1213 |
-
"loss": 0.07380253076553345,
|
| 1214 |
-
"step": 3420
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.4599045764850389,
|
| 1218 |
-
"grad_norm": 0.1779366433620453,
|
| 1219 |
-
"learning_rate": 4.790484920581262e-05,
|
| 1220 |
-
"loss": 0.045916372537612916,
|
| 1221 |
-
"step": 3440
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 0.4625784403018124,
|
| 1225 |
-
"grad_norm": 1.1228729486465454,
|
| 1226 |
-
"learning_rate": 4.7880739214078454e-05,
|
| 1227 |
-
"loss": 0.04461723566055298,
|
| 1228 |
-
"step": 3460
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 0.46525230411858587,
|
| 1232 |
-
"grad_norm": 0.1629919707775116,
|
| 1233 |
-
"learning_rate": 4.785649742996605e-05,
|
| 1234 |
-
"loss": 0.017159442603588104,
|
| 1235 |
-
"step": 3480
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 0.46792616793535935,
|
| 1239 |
-
"grad_norm": 3.583951473236084,
|
| 1240 |
-
"learning_rate": 4.783212399310737e-05,
|
| 1241 |
-
"loss": 0.047145146131515506,
|
| 1242 |
-
"step": 3500
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 0.47060003175213283,
|
| 1246 |
-
"grad_norm": 0.9766237139701843,
|
| 1247 |
-
"learning_rate": 4.780761904389267e-05,
|
| 1248 |
-
"loss": 0.050229442119598386,
|
| 1249 |
-
"step": 3520
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 0.4732738955689063,
|
| 1253 |
-
"grad_norm": 0.05617872253060341,
|
| 1254 |
-
"learning_rate": 4.778298272346976e-05,
|
| 1255 |
-
"loss": 0.024862812459468843,
|
| 1256 |
-
"step": 3540
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 0.4759477593856798,
|
| 1260 |
-
"grad_norm": 1.3586453199386597,
|
| 1261 |
-
"learning_rate": 4.775821517374308e-05,
|
| 1262 |
-
"loss": 0.02117772251367569,
|
| 1263 |
-
"step": 3560
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 0.4786216232024533,
|
| 1267 |
-
"grad_norm": 1.2116742134094238,
|
| 1268 |
-
"learning_rate": 4.7733316537373006e-05,
|
| 1269 |
-
"loss": 0.03060794174671173,
|
| 1270 |
-
"step": 3580
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 0.48129548701922675,
|
| 1274 |
-
"grad_norm": 0.39403238892555237,
|
| 1275 |
-
"learning_rate": 4.770828695777493e-05,
|
| 1276 |
-
"loss": 0.05482668280601501,
|
| 1277 |
-
"step": 3600
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 0.48396935083600023,
|
| 1281 |
-
"grad_norm": 0.9248486161231995,
|
| 1282 |
-
"learning_rate": 4.7683126579118495e-05,
|
| 1283 |
-
"loss": 0.03612814247608185,
|
| 1284 |
-
"step": 3620
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 0.4866432146527737,
|
| 1288 |
-
"grad_norm": 0.1624649167060852,
|
| 1289 |
-
"learning_rate": 4.7657835546326736e-05,
|
| 1290 |
-
"loss": 0.04334873259067536,
|
| 1291 |
-
"step": 3640
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 0.4893170784695472,
|
| 1295 |
-
"grad_norm": 0.5321119427680969,
|
| 1296 |
-
"learning_rate": 4.763241400507524e-05,
|
| 1297 |
-
"loss": 0.0461233913898468,
|
| 1298 |
-
"step": 3660
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 0.4919909422863207,
|
| 1302 |
-
"grad_norm": 0.34861093759536743,
|
| 1303 |
-
"learning_rate": 4.760686210179133e-05,
|
| 1304 |
-
"loss": 0.024829554557800292,
|
| 1305 |
-
"step": 3680
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 0.49466480610309416,
|
| 1309 |
-
"grad_norm": 1.2561241388320923,
|
| 1310 |
-
"learning_rate": 4.758117998365322e-05,
|
| 1311 |
-
"loss": 0.03157005608081818,
|
| 1312 |
-
"step": 3700
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 0.49733866991986764,
|
| 1316 |
-
"grad_norm": 0.8691341280937195,
|
| 1317 |
-
"learning_rate": 4.7555367798589146e-05,
|
| 1318 |
-
"loss": 0.04310203492641449,
|
| 1319 |
-
"step": 3720
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 0.5000125337366411,
|
| 1323 |
-
"grad_norm": 0.3134572505950928,
|
| 1324 |
-
"learning_rate": 4.752942569527653e-05,
|
| 1325 |
-
"loss": 0.03796039223670959,
|
| 1326 |
-
"step": 3740
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 0.5026863975534146,
|
| 1330 |
-
"grad_norm": 2.3359289169311523,
|
| 1331 |
-
"learning_rate": 4.75033538231411e-05,
|
| 1332 |
-
"loss": 0.055599170923233035,
|
| 1333 |
-
"step": 3760
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 0.5053602613701881,
|
| 1337 |
-
"grad_norm": 7.426175594329834,
|
| 1338 |
-
"learning_rate": 4.747715233235608e-05,
|
| 1339 |
-
"loss": 0.054436272382736205,
|
| 1340 |
-
"step": 3780
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 0.5080341251869616,
|
| 1344 |
-
"grad_norm": 0.5940203070640564,
|
| 1345 |
-
"learning_rate": 4.745082137384128e-05,
|
| 1346 |
-
"loss": 0.03682814538478851,
|
| 1347 |
-
"step": 3800
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 0.510707989003735,
|
| 1351 |
-
"grad_norm": 0.22821389138698578,
|
| 1352 |
-
"learning_rate": 4.7424361099262225e-05,
|
| 1353 |
-
"loss": 0.051123309135437014,
|
| 1354 |
-
"step": 3820
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 0.5133818528205085,
|
| 1358 |
-
"grad_norm": 8.20633602142334,
|
| 1359 |
-
"learning_rate": 4.739777166102932e-05,
|
| 1360 |
-
"loss": 0.0704378604888916,
|
| 1361 |
-
"step": 3840
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 0.516055716637282,
|
| 1365 |
-
"grad_norm": 3.023848533630371,
|
| 1366 |
-
"learning_rate": 4.737105321229694e-05,
|
| 1367 |
-
"loss": 0.03368058800697327,
|
| 1368 |
-
"step": 3860
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 0.5187295804540555,
|
| 1372 |
-
"grad_norm": 0.07666649669408798,
|
| 1373 |
-
"learning_rate": 4.7344205906962555e-05,
|
| 1374 |
-
"loss": 0.03665303289890289,
|
| 1375 |
-
"step": 3880
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 0.521403444270829,
|
| 1379 |
-
"grad_norm": 0.7571629881858826,
|
| 1380 |
-
"learning_rate": 4.731722989966585e-05,
|
| 1381 |
-
"loss": 0.058415502309799194,
|
| 1382 |
-
"step": 3900
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 0.5240773080876024,
|
| 1386 |
-
"grad_norm": 3.2599120140075684,
|
| 1387 |
-
"learning_rate": 4.7290125345787816e-05,
|
| 1388 |
-
"loss": 0.07323018908500671,
|
| 1389 |
-
"step": 3920
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 0.5267511719043759,
|
| 1393 |
-
"grad_norm": 0.28930988907814026,
|
| 1394 |
-
"learning_rate": 4.7262892401449886e-05,
|
| 1395 |
-
"loss": 0.054371267557144165,
|
| 1396 |
-
"step": 3940
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 0.5294250357211494,
|
| 1400 |
-
"grad_norm": 2.2296454906463623,
|
| 1401 |
-
"learning_rate": 4.7235531223513004e-05,
|
| 1402 |
-
"loss": 0.040819621086120604,
|
| 1403 |
-
"step": 3960
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 0.5320988995379229,
|
| 1407 |
-
"grad_norm": 0.11608211696147919,
|
| 1408 |
-
"learning_rate": 4.720804196957675e-05,
|
| 1409 |
-
"loss": 0.05215579271316528,
|
| 1410 |
-
"step": 3980
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 0.5347727633546964,
|
| 1414 |
-
"grad_norm": 1.1587547063827515,
|
| 1415 |
-
"learning_rate": 4.7180424797978415e-05,
|
| 1416 |
-
"loss": 0.026277875900268553,
|
| 1417 |
-
"step": 4000
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 0.5374466271714698,
|
| 1421 |
-
"grad_norm": 0.06253435462713242,
|
| 1422 |
-
"learning_rate": 4.7152679867792074e-05,
|
| 1423 |
-
"loss": 0.02574407756328583,
|
| 1424 |
-
"step": 4020
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 0.5401204909882433,
|
| 1428 |
-
"grad_norm": 1.3441458940505981,
|
| 1429 |
-
"learning_rate": 4.71248073388277e-05,
|
| 1430 |
-
"loss": 0.05538107752799988,
|
| 1431 |
-
"step": 4040
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 0.5427943548050168,
|
| 1435 |
-
"grad_norm": 0.48076340556144714,
|
| 1436 |
-
"learning_rate": 4.7096807371630236e-05,
|
| 1437 |
-
"loss": 0.047986540198326114,
|
| 1438 |
-
"step": 4060
|
| 1439 |
-
}
|
| 1440 |
-
],
|
| 1441 |
-
"logging_steps": 20,
|
| 1442 |
-
"max_steps": 26180,
|
| 1443 |
-
"num_input_tokens_seen": 0,
|
| 1444 |
-
"num_train_epochs": 4,
|
| 1445 |
-
"save_steps": 1000000000,
|
| 1446 |
-
"stateful_callbacks": {
|
| 1447 |
-
"TrainerControl": {
|
| 1448 |
-
"args": {
|
| 1449 |
-
"should_epoch_stop": false,
|
| 1450 |
-
"should_evaluate": false,
|
| 1451 |
-
"should_log": false,
|
| 1452 |
-
"should_save": true,
|
| 1453 |
-
"should_training_stop": false
|
| 1454 |
-
},
|
| 1455 |
-
"attributes": {}
|
| 1456 |
-
}
|
| 1457 |
-
},
|
| 1458 |
-
"total_flos": 1.5640394138763264e+16,
|
| 1459 |
-
"train_batch_size": 1,
|
| 1460 |
-
"trial_name": null,
|
| 1461 |
-
"trial_params": null
|
| 1462 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/12/checkpoint-4488/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/12/checkpoint-4488/trainer_state.json
DELETED
|
@@ -1,1609 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.8400201208028044,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 4488,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001336931908386741,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 129.80032348632812,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.002673863816773482,
|
| 21 |
-
"grad_norm": 63.64365768432617,
|
| 22 |
-
"learning_rate": 4.999995392022967e-05,
|
| 23 |
-
"loss": 63.88374408922697,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.005347727633546964,
|
| 28 |
-
"grad_norm": 24.627853393554688,
|
| 29 |
-
"learning_rate": 4.999976672145381e-05,
|
| 30 |
-
"loss": 12.65963363647461,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008021591450320446,
|
| 35 |
-
"grad_norm": 14.29983901977539,
|
| 36 |
-
"learning_rate": 4.999943552476422e-05,
|
| 37 |
-
"loss": 5.90204963684082,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.010695455267093928,
|
| 42 |
-
"grad_norm": 15.690323829650879,
|
| 43 |
-
"learning_rate": 4.999896033206858e-05,
|
| 44 |
-
"loss": 3.9918922424316405,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.01336931908386741,
|
| 49 |
-
"grad_norm": 31.583160400390625,
|
| 50 |
-
"learning_rate": 4.999834114610398e-05,
|
| 51 |
-
"loss": 2.9675426483154297,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.01604318290064089,
|
| 56 |
-
"grad_norm": 13.034649848937988,
|
| 57 |
-
"learning_rate": 4.999757797043691e-05,
|
| 58 |
-
"loss": 2.725296401977539,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.018717046717414372,
|
| 63 |
-
"grad_norm": 8.362203598022461,
|
| 64 |
-
"learning_rate": 4.999667080946324e-05,
|
| 65 |
-
"loss": 2.2478992462158205,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.021390910534187856,
|
| 70 |
-
"grad_norm": 8.726786613464355,
|
| 71 |
-
"learning_rate": 4.999561966840821e-05,
|
| 72 |
-
"loss": 1.8447845458984375,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.024064774350961337,
|
| 77 |
-
"grad_norm": 10.092752456665039,
|
| 78 |
-
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
-
"loss": 1.5611843109130858,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02673863816773482,
|
| 84 |
-
"grad_norm": 9.090085983276367,
|
| 85 |
-
"learning_rate": 4.999308547110146e-05,
|
| 86 |
-
"loss": 1.520334815979004,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.029412501984508302,
|
| 91 |
-
"grad_norm": 9.668124198913574,
|
| 92 |
-
"learning_rate": 4.999160242944665e-05,
|
| 93 |
-
"loss": 1.2818055152893066,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.03208636580128178,
|
| 98 |
-
"grad_norm": 9.182533264160156,
|
| 99 |
-
"learning_rate": 4.998997543690418e-05,
|
| 100 |
-
"loss": 1.0428407669067383,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.03476022961805526,
|
| 105 |
-
"grad_norm": 5.745838165283203,
|
| 106 |
-
"learning_rate": 4.998820450284549e-05,
|
| 107 |
-
"loss": 1.2343652725219727,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.037434093434828744,
|
| 112 |
-
"grad_norm": 8.651643753051758,
|
| 113 |
-
"learning_rate": 4.99862896374711e-05,
|
| 114 |
-
"loss": 0.8859601020812988,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04010795725160223,
|
| 119 |
-
"grad_norm": 10.765266418457031,
|
| 120 |
-
"learning_rate": 4.998423085181056e-05,
|
| 121 |
-
"loss": 0.989600658416748,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04278182106837571,
|
| 126 |
-
"grad_norm": 6.092499256134033,
|
| 127 |
-
"learning_rate": 4.998202815772245e-05,
|
| 128 |
-
"loss": 0.7189463615417481,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.04545568488514919,
|
| 133 |
-
"grad_norm": 6.352876663208008,
|
| 134 |
-
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
-
"loss": 0.7489545345306396,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.048129548701922674,
|
| 140 |
-
"grad_norm": 4.620656490325928,
|
| 141 |
-
"learning_rate": 4.997719109584209e-05,
|
| 142 |
-
"loss": 0.7381401538848877,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.050803412518696155,
|
| 147 |
-
"grad_norm": 7.796917915344238,
|
| 148 |
-
"learning_rate": 4.997455675591119e-05,
|
| 149 |
-
"loss": 0.5687405109405518,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.05347727633546964,
|
| 154 |
-
"grad_norm": 2.837172508239746,
|
| 155 |
-
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
-
"loss": 0.5686865329742432,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.05615114015224312,
|
| 161 |
-
"grad_norm": 3.3103690147399902,
|
| 162 |
-
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
-
"loss": 0.625730562210083,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.058825003969016604,
|
| 168 |
-
"grad_norm": 3.5682132244110107,
|
| 169 |
-
"learning_rate": 4.99657906847257e-05,
|
| 170 |
-
"loss": 0.6125466346740722,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.061498867785790085,
|
| 175 |
-
"grad_norm": 5.63640832901001,
|
| 176 |
-
"learning_rate": 4.996258103330218e-05,
|
| 177 |
-
"loss": 0.6182214260101319,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.06417273160256357,
|
| 182 |
-
"grad_norm": 4.698945999145508,
|
| 183 |
-
"learning_rate": 4.995922759815339e-05,
|
| 184 |
-
"loss": 0.43828091621398924,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.06684659541933705,
|
| 189 |
-
"grad_norm": 2.1976189613342285,
|
| 190 |
-
"learning_rate": 4.995573039859501e-05,
|
| 191 |
-
"loss": 0.4459230899810791,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.06952045923611053,
|
| 196 |
-
"grad_norm": 3.8809523582458496,
|
| 197 |
-
"learning_rate": 4.995208945477081e-05,
|
| 198 |
-
"loss": 0.3821882963180542,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.07219432305288401,
|
| 203 |
-
"grad_norm": 3.75144100189209,
|
| 204 |
-
"learning_rate": 4.994830478765251e-05,
|
| 205 |
-
"loss": 0.5800807476043701,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.07486818686965749,
|
| 210 |
-
"grad_norm": 3.0038585662841797,
|
| 211 |
-
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
-
"loss": 0.3928264617919922,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.07754205068643098,
|
| 217 |
-
"grad_norm": 3.614591598510742,
|
| 218 |
-
"learning_rate": 4.994030437155961e-05,
|
| 219 |
-
"loss": 0.48637890815734863,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.08021591450320446,
|
| 224 |
-
"grad_norm": 4.143443584442139,
|
| 225 |
-
"learning_rate": 4.993608866866718e-05,
|
| 226 |
-
"loss": 0.3650153160095215,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.08288977831997794,
|
| 231 |
-
"grad_norm": 6.692712783813477,
|
| 232 |
-
"learning_rate": 4.993172933464471e-05,
|
| 233 |
-
"loss": 0.3677916288375854,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.08556364213675142,
|
| 238 |
-
"grad_norm": 8.383441925048828,
|
| 239 |
-
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
-
"loss": 0.3399480104446411,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.0882375059535249,
|
| 245 |
-
"grad_norm": 5.566338062286377,
|
| 246 |
-
"learning_rate": 4.992257987447532e-05,
|
| 247 |
-
"loss": 0.28104052543640134,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.09091136977029839,
|
| 252 |
-
"grad_norm": 3.1196420192718506,
|
| 253 |
-
"learning_rate": 4.991778980102904e-05,
|
| 254 |
-
"loss": 0.351950478553772,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.09358523358707187,
|
| 259 |
-
"grad_norm": 3.47979736328125,
|
| 260 |
-
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
-
"loss": 0.27501535415649414,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.09625909740384535,
|
| 266 |
-
"grad_norm": 5.446717262268066,
|
| 267 |
-
"learning_rate": 4.990777910536653e-05,
|
| 268 |
-
"loss": 0.2651593923568726,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.09893296122061883,
|
| 273 |
-
"grad_norm": 7.6145339012146,
|
| 274 |
-
"learning_rate": 4.990255854081161e-05,
|
| 275 |
-
"loss": 0.35140380859375,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.10160682503739231,
|
| 280 |
-
"grad_norm": 8.445616722106934,
|
| 281 |
-
"learning_rate": 4.989719453825918e-05,
|
| 282 |
-
"loss": 0.2961219072341919,
|
| 283 |
-
"step": 760
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.10428068885416579,
|
| 287 |
-
"grad_norm": 6.339537620544434,
|
| 288 |
-
"learning_rate": 4.9891687128605744e-05,
|
| 289 |
-
"loss": 0.24962289333343507,
|
| 290 |
-
"step": 780
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.10695455267093928,
|
| 294 |
-
"grad_norm": 3.3369436264038086,
|
| 295 |
-
"learning_rate": 4.988603634357383e-05,
|
| 296 |
-
"loss": 0.2124847412109375,
|
| 297 |
-
"step": 800
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.10962841648771277,
|
| 301 |
-
"grad_norm": 2.2909045219421387,
|
| 302 |
-
"learning_rate": 4.988024221571177e-05,
|
| 303 |
-
"loss": 0.24679112434387207,
|
| 304 |
-
"step": 820
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.11230228030448625,
|
| 308 |
-
"grad_norm": 3.1149911880493164,
|
| 309 |
-
"learning_rate": 4.9874304778393574e-05,
|
| 310 |
-
"loss": 0.22161397933959961,
|
| 311 |
-
"step": 840
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.11497614412125973,
|
| 315 |
-
"grad_norm": 14.802160263061523,
|
| 316 |
-
"learning_rate": 4.9868224065818706e-05,
|
| 317 |
-
"loss": 0.2623537302017212,
|
| 318 |
-
"step": 860
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.11765000793803321,
|
| 322 |
-
"grad_norm": 5.586325168609619,
|
| 323 |
-
"learning_rate": 4.98620001130119e-05,
|
| 324 |
-
"loss": 0.3560942649841309,
|
| 325 |
-
"step": 880
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.12032387175480669,
|
| 329 |
-
"grad_norm": 3.390017032623291,
|
| 330 |
-
"learning_rate": 4.9855632955822916e-05,
|
| 331 |
-
"loss": 0.16934787034988402,
|
| 332 |
-
"step": 900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.12299773557158017,
|
| 336 |
-
"grad_norm": 6.070940971374512,
|
| 337 |
-
"learning_rate": 4.984912263092641e-05,
|
| 338 |
-
"loss": 0.2131197214126587,
|
| 339 |
-
"step": 920
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.12567159938835365,
|
| 343 |
-
"grad_norm": 1.4912281036376953,
|
| 344 |
-
"learning_rate": 4.984246917582166e-05,
|
| 345 |
-
"loss": 0.25128653049468996,
|
| 346 |
-
"step": 940
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.12834546320512713,
|
| 350 |
-
"grad_norm": 7.000472545623779,
|
| 351 |
-
"learning_rate": 4.9835672628832366e-05,
|
| 352 |
-
"loss": 0.2653592586517334,
|
| 353 |
-
"step": 960
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.1310193270219006,
|
| 357 |
-
"grad_norm": 5.427223205566406,
|
| 358 |
-
"learning_rate": 4.9828733029106434e-05,
|
| 359 |
-
"loss": 0.1653295636177063,
|
| 360 |
-
"step": 980
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.1336931908386741,
|
| 364 |
-
"grad_norm": 1.9502102136611938,
|
| 365 |
-
"learning_rate": 4.982165041661575e-05,
|
| 366 |
-
"loss": 0.2250870943069458,
|
| 367 |
-
"step": 1000
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.13636705465544757,
|
| 371 |
-
"grad_norm": 0.6216259598731995,
|
| 372 |
-
"learning_rate": 4.981442483215595e-05,
|
| 373 |
-
"loss": 0.18943849802017212,
|
| 374 |
-
"step": 1020
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.13904091847222105,
|
| 378 |
-
"grad_norm": 2.3363687992095947,
|
| 379 |
-
"learning_rate": 4.98070563173462e-05,
|
| 380 |
-
"loss": 0.1673592209815979,
|
| 381 |
-
"step": 1040
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.14171478228899453,
|
| 385 |
-
"grad_norm": 1.040717601776123,
|
| 386 |
-
"learning_rate": 4.979954491462892e-05,
|
| 387 |
-
"loss": 0.2113173007965088,
|
| 388 |
-
"step": 1060
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.14438864610576801,
|
| 392 |
-
"grad_norm": 2.735522747039795,
|
| 393 |
-
"learning_rate": 4.979189066726955e-05,
|
| 394 |
-
"loss": 0.17504971027374266,
|
| 395 |
-
"step": 1080
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.1470625099225415,
|
| 399 |
-
"grad_norm": 4.701151371002197,
|
| 400 |
-
"learning_rate": 4.978409361935636e-05,
|
| 401 |
-
"loss": 0.15881222486495972,
|
| 402 |
-
"step": 1100
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.14973637373931498,
|
| 406 |
-
"grad_norm": 2.735919237136841,
|
| 407 |
-
"learning_rate": 4.9776153815800075e-05,
|
| 408 |
-
"loss": 0.14044179916381835,
|
| 409 |
-
"step": 1120
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.15241023755608848,
|
| 413 |
-
"grad_norm": 3.5479538440704346,
|
| 414 |
-
"learning_rate": 4.976807130233375e-05,
|
| 415 |
-
"loss": 0.18565714359283447,
|
| 416 |
-
"step": 1140
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.15508410137286197,
|
| 420 |
-
"grad_norm": 3.2167458534240723,
|
| 421 |
-
"learning_rate": 4.975984612551243e-05,
|
| 422 |
-
"loss": 0.13236271142959594,
|
| 423 |
-
"step": 1160
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.15775796518963545,
|
| 427 |
-
"grad_norm": 1.0206760168075562,
|
| 428 |
-
"learning_rate": 4.975147833271288e-05,
|
| 429 |
-
"loss": 0.19124728441238403,
|
| 430 |
-
"step": 1180
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.16043182900640893,
|
| 434 |
-
"grad_norm": 4.194457530975342,
|
| 435 |
-
"learning_rate": 4.9742967972133335e-05,
|
| 436 |
-
"loss": 0.144741427898407,
|
| 437 |
-
"step": 1200
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.1631056928231824,
|
| 441 |
-
"grad_norm": 3.0225746631622314,
|
| 442 |
-
"learning_rate": 4.973431509279323e-05,
|
| 443 |
-
"loss": 0.1374324679374695,
|
| 444 |
-
"step": 1220
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.1657795566399559,
|
| 448 |
-
"grad_norm": 4.243523120880127,
|
| 449 |
-
"learning_rate": 4.972551974453287e-05,
|
| 450 |
-
"loss": 0.13663809299468993,
|
| 451 |
-
"step": 1240
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.16845342045672937,
|
| 455 |
-
"grad_norm": 2.4990086555480957,
|
| 456 |
-
"learning_rate": 4.971658197801322e-05,
|
| 457 |
-
"loss": 0.16817957162857056,
|
| 458 |
-
"step": 1260
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.17112728427350285,
|
| 462 |
-
"grad_norm": 4.983982563018799,
|
| 463 |
-
"learning_rate": 4.9707501844715554e-05,
|
| 464 |
-
"loss": 0.13795313835144044,
|
| 465 |
-
"step": 1280
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.17380114809027633,
|
| 469 |
-
"grad_norm": 3.6780316829681396,
|
| 470 |
-
"learning_rate": 4.969827939694115e-05,
|
| 471 |
-
"loss": 0.1637880802154541,
|
| 472 |
-
"step": 1300
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.1764750119070498,
|
| 476 |
-
"grad_norm": 0.7950732707977295,
|
| 477 |
-
"learning_rate": 4.968891468781105e-05,
|
| 478 |
-
"loss": 0.10979138612747193,
|
| 479 |
-
"step": 1320
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.1791488757238233,
|
| 483 |
-
"grad_norm": 1.2414121627807617,
|
| 484 |
-
"learning_rate": 4.967940777126569e-05,
|
| 485 |
-
"loss": 0.13692171573638917,
|
| 486 |
-
"step": 1340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.18182273954059677,
|
| 490 |
-
"grad_norm": 2.1383633613586426,
|
| 491 |
-
"learning_rate": 4.9669758702064636e-05,
|
| 492 |
-
"loss": 0.07821698188781738,
|
| 493 |
-
"step": 1360
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.18449660335737025,
|
| 497 |
-
"grad_norm": 5.061275959014893,
|
| 498 |
-
"learning_rate": 4.965996753578623e-05,
|
| 499 |
-
"loss": 0.19053516387939454,
|
| 500 |
-
"step": 1380
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.18717046717414373,
|
| 504 |
-
"grad_norm": 6.151792049407959,
|
| 505 |
-
"learning_rate": 4.9650034328827305e-05,
|
| 506 |
-
"loss": 0.11360721588134766,
|
| 507 |
-
"step": 1400
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.18984433099091721,
|
| 511 |
-
"grad_norm": 1.0604305267333984,
|
| 512 |
-
"learning_rate": 4.963995913840284e-05,
|
| 513 |
-
"loss": 0.13138024806976317,
|
| 514 |
-
"step": 1420
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.1925181948076907,
|
| 518 |
-
"grad_norm": 1.7159489393234253,
|
| 519 |
-
"learning_rate": 4.9629742022545623e-05,
|
| 520 |
-
"loss": 0.08657677173614502,
|
| 521 |
-
"step": 1440
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.19519205862446418,
|
| 525 |
-
"grad_norm": 2.4207754135131836,
|
| 526 |
-
"learning_rate": 4.961938304010595e-05,
|
| 527 |
-
"loss": 0.10309149026870727,
|
| 528 |
-
"step": 1460
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.19786592244123766,
|
| 532 |
-
"grad_norm": 1.532060146331787,
|
| 533 |
-
"learning_rate": 4.9608882250751245e-05,
|
| 534 |
-
"loss": 0.13628544807434081,
|
| 535 |
-
"step": 1480
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.20053978625801114,
|
| 539 |
-
"grad_norm": 6.409943580627441,
|
| 540 |
-
"learning_rate": 4.959823971496574e-05,
|
| 541 |
-
"loss": 0.10584845542907714,
|
| 542 |
-
"step": 1500
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.20321365007478462,
|
| 546 |
-
"grad_norm": 2.452012538909912,
|
| 547 |
-
"learning_rate": 4.9587455494050136e-05,
|
| 548 |
-
"loss": 0.06506187915802002,
|
| 549 |
-
"step": 1520
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.2058875138915581,
|
| 553 |
-
"grad_norm": 5.3016533851623535,
|
| 554 |
-
"learning_rate": 4.9576529650121214e-05,
|
| 555 |
-
"loss": 0.11848526000976563,
|
| 556 |
-
"step": 1540
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.20856137770833158,
|
| 560 |
-
"grad_norm": 4.341775894165039,
|
| 561 |
-
"learning_rate": 4.956546224611152e-05,
|
| 562 |
-
"loss": 0.11318533420562744,
|
| 563 |
-
"step": 1560
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.21123524152510506,
|
| 567 |
-
"grad_norm": 1.9056169986724854,
|
| 568 |
-
"learning_rate": 4.9554253345768965e-05,
|
| 569 |
-
"loss": 0.12768398523330687,
|
| 570 |
-
"step": 1580
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.21390910534187857,
|
| 574 |
-
"grad_norm": 1.8939746618270874,
|
| 575 |
-
"learning_rate": 4.9542903013656486e-05,
|
| 576 |
-
"loss": 0.10782338380813598,
|
| 577 |
-
"step": 1600
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.21658296915865205,
|
| 581 |
-
"grad_norm": 8.53671932220459,
|
| 582 |
-
"learning_rate": 4.9531411315151654e-05,
|
| 583 |
-
"loss": 0.1733921766281128,
|
| 584 |
-
"step": 1620
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.21925683297542553,
|
| 588 |
-
"grad_norm": 2.0152978897094727,
|
| 589 |
-
"learning_rate": 4.951977831644632e-05,
|
| 590 |
-
"loss": 0.11197054386138916,
|
| 591 |
-
"step": 1640
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.221930696792199,
|
| 595 |
-
"grad_norm": 3.8422367572784424,
|
| 596 |
-
"learning_rate": 4.95080040845462e-05,
|
| 597 |
-
"loss": 0.11441781520843505,
|
| 598 |
-
"step": 1660
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.2246045606089725,
|
| 602 |
-
"grad_norm": 1.819858193397522,
|
| 603 |
-
"learning_rate": 4.949608868727053e-05,
|
| 604 |
-
"loss": 0.11403474807739258,
|
| 605 |
-
"step": 1680
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.22727842442574597,
|
| 609 |
-
"grad_norm": 7.45100212097168,
|
| 610 |
-
"learning_rate": 4.948403219325163e-05,
|
| 611 |
-
"loss": 0.13117753267288207,
|
| 612 |
-
"step": 1700
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.22995228824251945,
|
| 616 |
-
"grad_norm": 0.6526040434837341,
|
| 617 |
-
"learning_rate": 4.947183467193456e-05,
|
| 618 |
-
"loss": 0.07524924874305725,
|
| 619 |
-
"step": 1720
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.23262615205929293,
|
| 623 |
-
"grad_norm": 3.814746856689453,
|
| 624 |
-
"learning_rate": 4.945949619357668e-05,
|
| 625 |
-
"loss": 0.07659345269203185,
|
| 626 |
-
"step": 1740
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.23530001587606642,
|
| 630 |
-
"grad_norm": 2.373124122619629,
|
| 631 |
-
"learning_rate": 4.944701682924726e-05,
|
| 632 |
-
"loss": 0.1147496223449707,
|
| 633 |
-
"step": 1760
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.2379738796928399,
|
| 637 |
-
"grad_norm": 0.11161285638809204,
|
| 638 |
-
"learning_rate": 4.943439665082707e-05,
|
| 639 |
-
"loss": 0.07256829738616943,
|
| 640 |
-
"step": 1780
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.24064774350961338,
|
| 644 |
-
"grad_norm": 0.45990192890167236,
|
| 645 |
-
"learning_rate": 4.942163573100794e-05,
|
| 646 |
-
"loss": 0.07726740837097168,
|
| 647 |
-
"step": 1800
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.24332160732638686,
|
| 651 |
-
"grad_norm": 4.2301926612854,
|
| 652 |
-
"learning_rate": 4.940873414329242e-05,
|
| 653 |
-
"loss": 0.09349535703659058,
|
| 654 |
-
"step": 1820
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.24599547114316034,
|
| 658 |
-
"grad_norm": 2.442178726196289,
|
| 659 |
-
"learning_rate": 4.939569196199325e-05,
|
| 660 |
-
"loss": 0.12413722276687622,
|
| 661 |
-
"step": 1840
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.24866933495993382,
|
| 665 |
-
"grad_norm": 2.523683786392212,
|
| 666 |
-
"learning_rate": 4.938250926223302e-05,
|
| 667 |
-
"loss": 0.08566288352012634,
|
| 668 |
-
"step": 1860
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.2513431987767073,
|
| 672 |
-
"grad_norm": 3.511075258255005,
|
| 673 |
-
"learning_rate": 4.936918611994368e-05,
|
| 674 |
-
"loss": 0.08007702231407166,
|
| 675 |
-
"step": 1880
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.2540170625934808,
|
| 679 |
-
"grad_norm": 6.254627704620361,
|
| 680 |
-
"learning_rate": 4.935572261186614e-05,
|
| 681 |
-
"loss": 0.10983954668045044,
|
| 682 |
-
"step": 1900
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 0.25669092641025426,
|
| 686 |
-
"grad_norm": 1.5211899280548096,
|
| 687 |
-
"learning_rate": 4.934211881554981e-05,
|
| 688 |
-
"loss": 0.09120344519615173,
|
| 689 |
-
"step": 1920
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 0.25936479022702774,
|
| 693 |
-
"grad_norm": 2.5893588066101074,
|
| 694 |
-
"learning_rate": 4.932837480935214e-05,
|
| 695 |
-
"loss": 0.08754412531852722,
|
| 696 |
-
"step": 1940
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 0.2620386540438012,
|
| 700 |
-
"grad_norm": 6.878556251525879,
|
| 701 |
-
"learning_rate": 4.931449067243821e-05,
|
| 702 |
-
"loss": 0.08636274933815002,
|
| 703 |
-
"step": 1960
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 0.2647125178605747,
|
| 707 |
-
"grad_norm": 2.9078798294067383,
|
| 708 |
-
"learning_rate": 4.9300466484780226e-05,
|
| 709 |
-
"loss": 0.09582929015159607,
|
| 710 |
-
"step": 1980
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.2673863816773482,
|
| 714 |
-
"grad_norm": 3.391852855682373,
|
| 715 |
-
"learning_rate": 4.92863023271571e-05,
|
| 716 |
-
"loss": 0.0850919783115387,
|
| 717 |
-
"step": 2000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.27006024549412166,
|
| 721 |
-
"grad_norm": 5.522103309631348,
|
| 722 |
-
"learning_rate": 4.927199828115395e-05,
|
| 723 |
-
"loss": 0.050999772548675534,
|
| 724 |
-
"step": 2020
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 0.27273410931089515,
|
| 728 |
-
"grad_norm": 0.90350741147995,
|
| 729 |
-
"learning_rate": 4.925755442916167e-05,
|
| 730 |
-
"loss": 0.10100446939468384,
|
| 731 |
-
"step": 2040
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 0.2754079731276686,
|
| 735 |
-
"grad_norm": 1.602030634880066,
|
| 736 |
-
"learning_rate": 4.924297085437641e-05,
|
| 737 |
-
"loss": 0.0468633770942688,
|
| 738 |
-
"step": 2060
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 0.2780818369444421,
|
| 742 |
-
"grad_norm": 1.5823460817337036,
|
| 743 |
-
"learning_rate": 4.922824764079913e-05,
|
| 744 |
-
"loss": 0.06786358952522278,
|
| 745 |
-
"step": 2080
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.2807557007612156,
|
| 749 |
-
"grad_norm": 1.6624343395233154,
|
| 750 |
-
"learning_rate": 4.92133848732351e-05,
|
| 751 |
-
"loss": 0.05772828459739685,
|
| 752 |
-
"step": 2100
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 0.28342956457798907,
|
| 756 |
-
"grad_norm": 0.947078287601471,
|
| 757 |
-
"learning_rate": 4.9198382637293424e-05,
|
| 758 |
-
"loss": 0.08012173175811768,
|
| 759 |
-
"step": 2120
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.28610342839476255,
|
| 763 |
-
"grad_norm": 0.2919924259185791,
|
| 764 |
-
"learning_rate": 4.918324101938653e-05,
|
| 765 |
-
"loss": 0.1208539366722107,
|
| 766 |
-
"step": 2140
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.28877729221153603,
|
| 770 |
-
"grad_norm": 9.258247375488281,
|
| 771 |
-
"learning_rate": 4.916796010672969e-05,
|
| 772 |
-
"loss": 0.10037034749984741,
|
| 773 |
-
"step": 2160
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.2914511560283095,
|
| 777 |
-
"grad_norm": 4.0920491218566895,
|
| 778 |
-
"learning_rate": 4.915253998734051e-05,
|
| 779 |
-
"loss": 0.061488878726959226,
|
| 780 |
-
"step": 2180
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 0.294125019845083,
|
| 784 |
-
"grad_norm": 6.1126627922058105,
|
| 785 |
-
"learning_rate": 4.913698075003841e-05,
|
| 786 |
-
"loss": 0.0862967312335968,
|
| 787 |
-
"step": 2200
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.29679888366185647,
|
| 791 |
-
"grad_norm": 2.585484743118286,
|
| 792 |
-
"learning_rate": 4.912128248444414e-05,
|
| 793 |
-
"loss": 0.05393874645233154,
|
| 794 |
-
"step": 2220
|
| 795 |
-
},
|
| 796 |
-
{
|
| 797 |
-
"epoch": 0.29947274747862995,
|
| 798 |
-
"grad_norm": 6.944481372833252,
|
| 799 |
-
"learning_rate": 4.9105445280979256e-05,
|
| 800 |
-
"loss": 0.08570566773414612,
|
| 801 |
-
"step": 2240
|
| 802 |
-
},
|
| 803 |
-
{
|
| 804 |
-
"epoch": 0.30214661129540343,
|
| 805 |
-
"grad_norm": 1.3824089765548706,
|
| 806 |
-
"learning_rate": 4.908946923086556e-05,
|
| 807 |
-
"loss": 0.09689127206802368,
|
| 808 |
-
"step": 2260
|
| 809 |
-
},
|
| 810 |
-
{
|
| 811 |
-
"epoch": 0.30482047511217697,
|
| 812 |
-
"grad_norm": 3.4861342906951904,
|
| 813 |
-
"learning_rate": 4.907335442612464e-05,
|
| 814 |
-
"loss": 0.12550976276397705,
|
| 815 |
-
"step": 2280
|
| 816 |
-
},
|
| 817 |
-
{
|
| 818 |
-
"epoch": 0.30749433892895045,
|
| 819 |
-
"grad_norm": 3.668980121612549,
|
| 820 |
-
"learning_rate": 4.905710095957728e-05,
|
| 821 |
-
"loss": 0.09089353680610657,
|
| 822 |
-
"step": 2300
|
| 823 |
-
},
|
| 824 |
-
{
|
| 825 |
-
"epoch": 0.31016820274572393,
|
| 826 |
-
"grad_norm": 1.093095064163208,
|
| 827 |
-
"learning_rate": 4.904070892484298e-05,
|
| 828 |
-
"loss": 0.03925192356109619,
|
| 829 |
-
"step": 2320
|
| 830 |
-
},
|
| 831 |
-
{
|
| 832 |
-
"epoch": 0.3128420665624974,
|
| 833 |
-
"grad_norm": 0.8169485926628113,
|
| 834 |
-
"learning_rate": 4.9024178416339364e-05,
|
| 835 |
-
"loss": 0.0979581356048584,
|
| 836 |
-
"step": 2340
|
| 837 |
-
},
|
| 838 |
-
{
|
| 839 |
-
"epoch": 0.3155159303792709,
|
| 840 |
-
"grad_norm": 1.892451286315918,
|
| 841 |
-
"learning_rate": 4.900750952928166e-05,
|
| 842 |
-
"loss": 0.05913209915161133,
|
| 843 |
-
"step": 2360
|
| 844 |
-
},
|
| 845 |
-
{
|
| 846 |
-
"epoch": 0.3181897941960444,
|
| 847 |
-
"grad_norm": 0.24644255638122559,
|
| 848 |
-
"learning_rate": 4.8990702359682184e-05,
|
| 849 |
-
"loss": 0.06815173625946044,
|
| 850 |
-
"step": 2380
|
| 851 |
-
},
|
| 852 |
-
{
|
| 853 |
-
"epoch": 0.32086365801281785,
|
| 854 |
-
"grad_norm": 2.1861305236816406,
|
| 855 |
-
"learning_rate": 4.897375700434972e-05,
|
| 856 |
-
"loss": 0.04142785966396332,
|
| 857 |
-
"step": 2400
|
| 858 |
-
},
|
| 859 |
-
{
|
| 860 |
-
"epoch": 0.32353752182959133,
|
| 861 |
-
"grad_norm": 2.6643004417419434,
|
| 862 |
-
"learning_rate": 4.8956673560889013e-05,
|
| 863 |
-
"loss": 0.05177200436592102,
|
| 864 |
-
"step": 2420
|
| 865 |
-
},
|
| 866 |
-
{
|
| 867 |
-
"epoch": 0.3262113856463648,
|
| 868 |
-
"grad_norm": 2.588113784790039,
|
| 869 |
-
"learning_rate": 4.8939452127700195e-05,
|
| 870 |
-
"loss": 0.05783546566963196,
|
| 871 |
-
"step": 2440
|
| 872 |
-
},
|
| 873 |
-
{
|
| 874 |
-
"epoch": 0.3288852494631383,
|
| 875 |
-
"grad_norm": 2.419644594192505,
|
| 876 |
-
"learning_rate": 4.8922092803978203e-05,
|
| 877 |
-
"loss": 0.08906854391098022,
|
| 878 |
-
"step": 2460
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 0.3315591132799118,
|
| 882 |
-
"grad_norm": 0.16949939727783203,
|
| 883 |
-
"learning_rate": 4.890459568971223e-05,
|
| 884 |
-
"loss": 0.10305211544036866,
|
| 885 |
-
"step": 2480
|
| 886 |
-
},
|
| 887 |
-
{
|
| 888 |
-
"epoch": 0.33423297709668526,
|
| 889 |
-
"grad_norm": 0.10032984614372253,
|
| 890 |
-
"learning_rate": 4.8886960885685126e-05,
|
| 891 |
-
"loss": 0.06348527669906616,
|
| 892 |
-
"step": 2500
|
| 893 |
-
},
|
| 894 |
-
{
|
| 895 |
-
"epoch": 0.33690684091345874,
|
| 896 |
-
"grad_norm": 3.3658738136291504,
|
| 897 |
-
"learning_rate": 4.8869188493472854e-05,
|
| 898 |
-
"loss": 0.06826075911521912,
|
| 899 |
-
"step": 2520
|
| 900 |
-
},
|
| 901 |
-
{
|
| 902 |
-
"epoch": 0.3395807047302322,
|
| 903 |
-
"grad_norm": 0.8656186461448669,
|
| 904 |
-
"learning_rate": 4.885127861544386e-05,
|
| 905 |
-
"loss": 0.05929765701293945,
|
| 906 |
-
"step": 2540
|
| 907 |
-
},
|
| 908 |
-
{
|
| 909 |
-
"epoch": 0.3422545685470057,
|
| 910 |
-
"grad_norm": 0.1492065042257309,
|
| 911 |
-
"learning_rate": 4.8833231354758496e-05,
|
| 912 |
-
"loss": 0.09429731965065002,
|
| 913 |
-
"step": 2560
|
| 914 |
-
},
|
| 915 |
-
{
|
| 916 |
-
"epoch": 0.3449284323637792,
|
| 917 |
-
"grad_norm": 0.6010928153991699,
|
| 918 |
-
"learning_rate": 4.881504681536846e-05,
|
| 919 |
-
"loss": 0.06262240409851075,
|
| 920 |
-
"step": 2580
|
| 921 |
-
},
|
| 922 |
-
{
|
| 923 |
-
"epoch": 0.34760229618055266,
|
| 924 |
-
"grad_norm": 1.6506450176239014,
|
| 925 |
-
"learning_rate": 4.879672510201616e-05,
|
| 926 |
-
"loss": 0.061688083410263064,
|
| 927 |
-
"step": 2600
|
| 928 |
-
},
|
| 929 |
-
{
|
| 930 |
-
"epoch": 0.35027615999732614,
|
| 931 |
-
"grad_norm": 0.2703142464160919,
|
| 932 |
-
"learning_rate": 4.877826632023412e-05,
|
| 933 |
-
"loss": 0.06175137162208557,
|
| 934 |
-
"step": 2620
|
| 935 |
-
},
|
| 936 |
-
{
|
| 937 |
-
"epoch": 0.3529500238140996,
|
| 938 |
-
"grad_norm": 3.1056365966796875,
|
| 939 |
-
"learning_rate": 4.875967057634437e-05,
|
| 940 |
-
"loss": 0.07828506827354431,
|
| 941 |
-
"step": 2640
|
| 942 |
-
},
|
| 943 |
-
{
|
| 944 |
-
"epoch": 0.3556238876308731,
|
| 945 |
-
"grad_norm": 0.28790283203125,
|
| 946 |
-
"learning_rate": 4.874093797745784e-05,
|
| 947 |
-
"loss": 0.11355981826782227,
|
| 948 |
-
"step": 2660
|
| 949 |
-
},
|
| 950 |
-
{
|
| 951 |
-
"epoch": 0.3582977514476466,
|
| 952 |
-
"grad_norm": 2.3372068405151367,
|
| 953 |
-
"learning_rate": 4.8722068631473746e-05,
|
| 954 |
-
"loss": 0.048267871141433716,
|
| 955 |
-
"step": 2680
|
| 956 |
-
},
|
| 957 |
-
{
|
| 958 |
-
"epoch": 0.36097161526442006,
|
| 959 |
-
"grad_norm": 0.12767371535301208,
|
| 960 |
-
"learning_rate": 4.8703062647078976e-05,
|
| 961 |
-
"loss": 0.04319801032543182,
|
| 962 |
-
"step": 2700
|
| 963 |
-
},
|
| 964 |
-
{
|
| 965 |
-
"epoch": 0.36364547908119355,
|
| 966 |
-
"grad_norm": 0.5145738124847412,
|
| 967 |
-
"learning_rate": 4.868392013374741e-05,
|
| 968 |
-
"loss": 0.0773090660572052,
|
| 969 |
-
"step": 2720
|
| 970 |
-
},
|
| 971 |
-
{
|
| 972 |
-
"epoch": 0.366319342897967,
|
| 973 |
-
"grad_norm": 0.8518500328063965,
|
| 974 |
-
"learning_rate": 4.866464120173937e-05,
|
| 975 |
-
"loss": 0.05149460434913635,
|
| 976 |
-
"step": 2740
|
| 977 |
-
},
|
| 978 |
-
{
|
| 979 |
-
"epoch": 0.3689932067147405,
|
| 980 |
-
"grad_norm": 3.6726584434509277,
|
| 981 |
-
"learning_rate": 4.8645225962100924e-05,
|
| 982 |
-
"loss": 0.06896821856498718,
|
| 983 |
-
"step": 2760
|
| 984 |
-
},
|
| 985 |
-
{
|
| 986 |
-
"epoch": 0.371667070531514,
|
| 987 |
-
"grad_norm": 1.5626497268676758,
|
| 988 |
-
"learning_rate": 4.862567452666329e-05,
|
| 989 |
-
"loss": 0.047730174660682675,
|
| 990 |
-
"step": 2780
|
| 991 |
-
},
|
| 992 |
-
{
|
| 993 |
-
"epoch": 0.37434093434828747,
|
| 994 |
-
"grad_norm": 6.562028884887695,
|
| 995 |
-
"learning_rate": 4.8605987008042144e-05,
|
| 996 |
-
"loss": 0.07060698866844177,
|
| 997 |
-
"step": 2800
|
| 998 |
-
},
|
| 999 |
-
{
|
| 1000 |
-
"epoch": 0.37701479816506095,
|
| 1001 |
-
"grad_norm": 0.7631726861000061,
|
| 1002 |
-
"learning_rate": 4.8586163519637005e-05,
|
| 1003 |
-
"loss": 0.04944324493408203,
|
| 1004 |
-
"step": 2820
|
| 1005 |
-
},
|
| 1006 |
-
{
|
| 1007 |
-
"epoch": 0.37968866198183443,
|
| 1008 |
-
"grad_norm": 1.6982293128967285,
|
| 1009 |
-
"learning_rate": 4.8566204175630595e-05,
|
| 1010 |
-
"loss": 0.03000348210334778,
|
| 1011 |
-
"step": 2840
|
| 1012 |
-
},
|
| 1013 |
-
{
|
| 1014 |
-
"epoch": 0.3823625257986079,
|
| 1015 |
-
"grad_norm": 0.6487429141998291,
|
| 1016 |
-
"learning_rate": 4.854610909098812e-05,
|
| 1017 |
-
"loss": 0.06691416501998901,
|
| 1018 |
-
"step": 2860
|
| 1019 |
-
},
|
| 1020 |
-
{
|
| 1021 |
-
"epoch": 0.3850363896153814,
|
| 1022 |
-
"grad_norm": 0.7648892402648926,
|
| 1023 |
-
"learning_rate": 4.852587838145668e-05,
|
| 1024 |
-
"loss": 0.05529783964157105,
|
| 1025 |
-
"step": 2880
|
| 1026 |
-
},
|
| 1027 |
-
{
|
| 1028 |
-
"epoch": 0.38771025343215487,
|
| 1029 |
-
"grad_norm": 0.11601298302412033,
|
| 1030 |
-
"learning_rate": 4.850551216356457e-05,
|
| 1031 |
-
"loss": 0.07780832052230835,
|
| 1032 |
-
"step": 2900
|
| 1033 |
-
},
|
| 1034 |
-
{
|
| 1035 |
-
"epoch": 0.39038411724892835,
|
| 1036 |
-
"grad_norm": 0.9443137645721436,
|
| 1037 |
-
"learning_rate": 4.8485010554620594e-05,
|
| 1038 |
-
"loss": 0.08007023930549621,
|
| 1039 |
-
"step": 2920
|
| 1040 |
-
},
|
| 1041 |
-
{
|
| 1042 |
-
"epoch": 0.39305798106570183,
|
| 1043 |
-
"grad_norm": 0.8828252553939819,
|
| 1044 |
-
"learning_rate": 4.846437367271341e-05,
|
| 1045 |
-
"loss": 0.03541453182697296,
|
| 1046 |
-
"step": 2940
|
| 1047 |
-
},
|
| 1048 |
-
{
|
| 1049 |
-
"epoch": 0.3957318448824753,
|
| 1050 |
-
"grad_norm": 0.21668888628482819,
|
| 1051 |
-
"learning_rate": 4.844360163671083e-05,
|
| 1052 |
-
"loss": 0.08354364633560181,
|
| 1053 |
-
"step": 2960
|
| 1054 |
-
},
|
| 1055 |
-
{
|
| 1056 |
-
"epoch": 0.3984057086992488,
|
| 1057 |
-
"grad_norm": 0.6840483546257019,
|
| 1058 |
-
"learning_rate": 4.8422694566259194e-05,
|
| 1059 |
-
"loss": 0.045807772874832155,
|
| 1060 |
-
"step": 2980
|
| 1061 |
-
},
|
| 1062 |
-
{
|
| 1063 |
-
"epoch": 0.4010795725160223,
|
| 1064 |
-
"grad_norm": 1.2754698991775513,
|
| 1065 |
-
"learning_rate": 4.8401652581782584e-05,
|
| 1066 |
-
"loss": 0.053487342596054074,
|
| 1067 |
-
"step": 3000
|
| 1068 |
-
},
|
| 1069 |
-
{
|
| 1070 |
-
"epoch": 0.40375343633279576,
|
| 1071 |
-
"grad_norm": 0.19012756645679474,
|
| 1072 |
-
"learning_rate": 4.838047580448222e-05,
|
| 1073 |
-
"loss": 0.05881953239440918,
|
| 1074 |
-
"step": 3020
|
| 1075 |
-
},
|
| 1076 |
-
{
|
| 1077 |
-
"epoch": 0.40642730014956924,
|
| 1078 |
-
"grad_norm": 2.1057698726654053,
|
| 1079 |
-
"learning_rate": 4.835916435633569e-05,
|
| 1080 |
-
"loss": 0.031065690517425536,
|
| 1081 |
-
"step": 3040
|
| 1082 |
-
},
|
| 1083 |
-
{
|
| 1084 |
-
"epoch": 0.4091011639663427,
|
| 1085 |
-
"grad_norm": 4.188559055328369,
|
| 1086 |
-
"learning_rate": 4.833771836009633e-05,
|
| 1087 |
-
"loss": 0.07205432653427124,
|
| 1088 |
-
"step": 3060
|
| 1089 |
-
},
|
| 1090 |
-
{
|
| 1091 |
-
"epoch": 0.4117750277831162,
|
| 1092 |
-
"grad_norm": 6.975829124450684,
|
| 1093 |
-
"learning_rate": 4.831613793929242e-05,
|
| 1094 |
-
"loss": 0.04953635036945343,
|
| 1095 |
-
"step": 3080
|
| 1096 |
-
},
|
| 1097 |
-
{
|
| 1098 |
-
"epoch": 0.4144488915998897,
|
| 1099 |
-
"grad_norm": 4.725269317626953,
|
| 1100 |
-
"learning_rate": 4.8294423218226546e-05,
|
| 1101 |
-
"loss": 0.05965519547462463,
|
| 1102 |
-
"step": 3100
|
| 1103 |
-
},
|
| 1104 |
-
{
|
| 1105 |
-
"epoch": 0.41712275541666316,
|
| 1106 |
-
"grad_norm": 1.7124755382537842,
|
| 1107 |
-
"learning_rate": 4.827257432197486e-05,
|
| 1108 |
-
"loss": 0.039625433087348935,
|
| 1109 |
-
"step": 3120
|
| 1110 |
-
},
|
| 1111 |
-
{
|
| 1112 |
-
"epoch": 0.41979661923343664,
|
| 1113 |
-
"grad_norm": 2.6687324047088623,
|
| 1114 |
-
"learning_rate": 4.825059137638636e-05,
|
| 1115 |
-
"loss": 0.05020809769630432,
|
| 1116 |
-
"step": 3140
|
| 1117 |
-
},
|
| 1118 |
-
{
|
| 1119 |
-
"epoch": 0.4224704830502101,
|
| 1120 |
-
"grad_norm": 1.111640214920044,
|
| 1121 |
-
"learning_rate": 4.822847450808215e-05,
|
| 1122 |
-
"loss": 0.04404452443122864,
|
| 1123 |
-
"step": 3160
|
| 1124 |
-
},
|
| 1125 |
-
{
|
| 1126 |
-
"epoch": 0.42514434686698366,
|
| 1127 |
-
"grad_norm": 0.2128070890903473,
|
| 1128 |
-
"learning_rate": 4.8206223844454744e-05,
|
| 1129 |
-
"loss": 0.08283355236053466,
|
| 1130 |
-
"step": 3180
|
| 1131 |
-
},
|
| 1132 |
-
{
|
| 1133 |
-
"epoch": 0.42781821068375714,
|
| 1134 |
-
"grad_norm": 0.10757248103618622,
|
| 1135 |
-
"learning_rate": 4.818383951366729e-05,
|
| 1136 |
-
"loss": 0.08568671345710754,
|
| 1137 |
-
"step": 3200
|
| 1138 |
-
},
|
| 1139 |
-
{
|
| 1140 |
-
"epoch": 0.4304920745005306,
|
| 1141 |
-
"grad_norm": 0.08344592899084091,
|
| 1142 |
-
"learning_rate": 4.816132164465289e-05,
|
| 1143 |
-
"loss": 0.0426956832408905,
|
| 1144 |
-
"step": 3220
|
| 1145 |
-
},
|
| 1146 |
-
{
|
| 1147 |
-
"epoch": 0.4331659383173041,
|
| 1148 |
-
"grad_norm": 0.5657751560211182,
|
| 1149 |
-
"learning_rate": 4.813867036711378e-05,
|
| 1150 |
-
"loss": 0.04971776902675629,
|
| 1151 |
-
"step": 3240
|
| 1152 |
-
},
|
| 1153 |
-
{
|
| 1154 |
-
"epoch": 0.4358398021340776,
|
| 1155 |
-
"grad_norm": 2.1529288291931152,
|
| 1156 |
-
"learning_rate": 4.8115885811520654e-05,
|
| 1157 |
-
"loss": 0.025386181473731995,
|
| 1158 |
-
"step": 3260
|
| 1159 |
-
},
|
| 1160 |
-
{
|
| 1161 |
-
"epoch": 0.43851366595085106,
|
| 1162 |
-
"grad_norm": 4.228519916534424,
|
| 1163 |
-
"learning_rate": 4.809296810911188e-05,
|
| 1164 |
-
"loss": 0.06401395201683044,
|
| 1165 |
-
"step": 3280
|
| 1166 |
-
},
|
| 1167 |
-
{
|
| 1168 |
-
"epoch": 0.44118752976762454,
|
| 1169 |
-
"grad_norm": 6.770420551300049,
|
| 1170 |
-
"learning_rate": 4.806991739189274e-05,
|
| 1171 |
-
"loss": 0.16425553560256959,
|
| 1172 |
-
"step": 3300
|
| 1173 |
-
},
|
| 1174 |
-
{
|
| 1175 |
-
"epoch": 0.443861393584398,
|
| 1176 |
-
"grad_norm": 0.5303187370300293,
|
| 1177 |
-
"learning_rate": 4.804673379263467e-05,
|
| 1178 |
-
"loss": 0.045900467038154605,
|
| 1179 |
-
"step": 3320
|
| 1180 |
-
},
|
| 1181 |
-
{
|
| 1182 |
-
"epoch": 0.4465352574011715,
|
| 1183 |
-
"grad_norm": 0.221473827958107,
|
| 1184 |
-
"learning_rate": 4.802341744487453e-05,
|
| 1185 |
-
"loss": 0.07529735565185547,
|
| 1186 |
-
"step": 3340
|
| 1187 |
-
},
|
| 1188 |
-
{
|
| 1189 |
-
"epoch": 0.449209121217945,
|
| 1190 |
-
"grad_norm": 3.48736834526062,
|
| 1191 |
-
"learning_rate": 4.799996848291378e-05,
|
| 1192 |
-
"loss": 0.062433135509490964,
|
| 1193 |
-
"step": 3360
|
| 1194 |
-
},
|
| 1195 |
-
{
|
| 1196 |
-
"epoch": 0.45188298503471847,
|
| 1197 |
-
"grad_norm": 2.650038242340088,
|
| 1198 |
-
"learning_rate": 4.797638704181774e-05,
|
| 1199 |
-
"loss": 0.03762982189655304,
|
| 1200 |
-
"step": 3380
|
| 1201 |
-
},
|
| 1202 |
-
{
|
| 1203 |
-
"epoch": 0.45455684885149195,
|
| 1204 |
-
"grad_norm": 3.159665584564209,
|
| 1205 |
-
"learning_rate": 4.795267325741483e-05,
|
| 1206 |
-
"loss": 0.04745924174785614,
|
| 1207 |
-
"step": 3400
|
| 1208 |
-
},
|
| 1209 |
-
{
|
| 1210 |
-
"epoch": 0.4572307126682654,
|
| 1211 |
-
"grad_norm": 0.8763885498046875,
|
| 1212 |
-
"learning_rate": 4.7928827266295715e-05,
|
| 1213 |
-
"loss": 0.07380253076553345,
|
| 1214 |
-
"step": 3420
|
| 1215 |
-
},
|
| 1216 |
-
{
|
| 1217 |
-
"epoch": 0.4599045764850389,
|
| 1218 |
-
"grad_norm": 0.1779366433620453,
|
| 1219 |
-
"learning_rate": 4.790484920581262e-05,
|
| 1220 |
-
"loss": 0.045916372537612916,
|
| 1221 |
-
"step": 3440
|
| 1222 |
-
},
|
| 1223 |
-
{
|
| 1224 |
-
"epoch": 0.4625784403018124,
|
| 1225 |
-
"grad_norm": 1.1228729486465454,
|
| 1226 |
-
"learning_rate": 4.7880739214078454e-05,
|
| 1227 |
-
"loss": 0.04461723566055298,
|
| 1228 |
-
"step": 3460
|
| 1229 |
-
},
|
| 1230 |
-
{
|
| 1231 |
-
"epoch": 0.46525230411858587,
|
| 1232 |
-
"grad_norm": 0.1629919707775116,
|
| 1233 |
-
"learning_rate": 4.785649742996605e-05,
|
| 1234 |
-
"loss": 0.017159442603588104,
|
| 1235 |
-
"step": 3480
|
| 1236 |
-
},
|
| 1237 |
-
{
|
| 1238 |
-
"epoch": 0.46792616793535935,
|
| 1239 |
-
"grad_norm": 3.583951473236084,
|
| 1240 |
-
"learning_rate": 4.783212399310737e-05,
|
| 1241 |
-
"loss": 0.047145146131515506,
|
| 1242 |
-
"step": 3500
|
| 1243 |
-
},
|
| 1244 |
-
{
|
| 1245 |
-
"epoch": 0.47060003175213283,
|
| 1246 |
-
"grad_norm": 0.9766237139701843,
|
| 1247 |
-
"learning_rate": 4.780761904389267e-05,
|
| 1248 |
-
"loss": 0.050229442119598386,
|
| 1249 |
-
"step": 3520
|
| 1250 |
-
},
|
| 1251 |
-
{
|
| 1252 |
-
"epoch": 0.4732738955689063,
|
| 1253 |
-
"grad_norm": 0.05617872253060341,
|
| 1254 |
-
"learning_rate": 4.778298272346976e-05,
|
| 1255 |
-
"loss": 0.024862812459468843,
|
| 1256 |
-
"step": 3540
|
| 1257 |
-
},
|
| 1258 |
-
{
|
| 1259 |
-
"epoch": 0.4759477593856798,
|
| 1260 |
-
"grad_norm": 1.3586453199386597,
|
| 1261 |
-
"learning_rate": 4.775821517374308e-05,
|
| 1262 |
-
"loss": 0.02117772251367569,
|
| 1263 |
-
"step": 3560
|
| 1264 |
-
},
|
| 1265 |
-
{
|
| 1266 |
-
"epoch": 0.4786216232024533,
|
| 1267 |
-
"grad_norm": 1.2116742134094238,
|
| 1268 |
-
"learning_rate": 4.7733316537373006e-05,
|
| 1269 |
-
"loss": 0.03060794174671173,
|
| 1270 |
-
"step": 3580
|
| 1271 |
-
},
|
| 1272 |
-
{
|
| 1273 |
-
"epoch": 0.48129548701922675,
|
| 1274 |
-
"grad_norm": 0.39403238892555237,
|
| 1275 |
-
"learning_rate": 4.770828695777493e-05,
|
| 1276 |
-
"loss": 0.05482668280601501,
|
| 1277 |
-
"step": 3600
|
| 1278 |
-
},
|
| 1279 |
-
{
|
| 1280 |
-
"epoch": 0.48396935083600023,
|
| 1281 |
-
"grad_norm": 0.9248486161231995,
|
| 1282 |
-
"learning_rate": 4.7683126579118495e-05,
|
| 1283 |
-
"loss": 0.03612814247608185,
|
| 1284 |
-
"step": 3620
|
| 1285 |
-
},
|
| 1286 |
-
{
|
| 1287 |
-
"epoch": 0.4866432146527737,
|
| 1288 |
-
"grad_norm": 0.1624649167060852,
|
| 1289 |
-
"learning_rate": 4.7657835546326736e-05,
|
| 1290 |
-
"loss": 0.04334873259067536,
|
| 1291 |
-
"step": 3640
|
| 1292 |
-
},
|
| 1293 |
-
{
|
| 1294 |
-
"epoch": 0.4893170784695472,
|
| 1295 |
-
"grad_norm": 0.5321119427680969,
|
| 1296 |
-
"learning_rate": 4.763241400507524e-05,
|
| 1297 |
-
"loss": 0.0461233913898468,
|
| 1298 |
-
"step": 3660
|
| 1299 |
-
},
|
| 1300 |
-
{
|
| 1301 |
-
"epoch": 0.4919909422863207,
|
| 1302 |
-
"grad_norm": 0.34861093759536743,
|
| 1303 |
-
"learning_rate": 4.760686210179133e-05,
|
| 1304 |
-
"loss": 0.024829554557800292,
|
| 1305 |
-
"step": 3680
|
| 1306 |
-
},
|
| 1307 |
-
{
|
| 1308 |
-
"epoch": 0.49466480610309416,
|
| 1309 |
-
"grad_norm": 1.2561241388320923,
|
| 1310 |
-
"learning_rate": 4.758117998365322e-05,
|
| 1311 |
-
"loss": 0.03157005608081818,
|
| 1312 |
-
"step": 3700
|
| 1313 |
-
},
|
| 1314 |
-
{
|
| 1315 |
-
"epoch": 0.49733866991986764,
|
| 1316 |
-
"grad_norm": 0.8691341280937195,
|
| 1317 |
-
"learning_rate": 4.7555367798589146e-05,
|
| 1318 |
-
"loss": 0.04310203492641449,
|
| 1319 |
-
"step": 3720
|
| 1320 |
-
},
|
| 1321 |
-
{
|
| 1322 |
-
"epoch": 0.5000125337366411,
|
| 1323 |
-
"grad_norm": 0.3134572505950928,
|
| 1324 |
-
"learning_rate": 4.752942569527653e-05,
|
| 1325 |
-
"loss": 0.03796039223670959,
|
| 1326 |
-
"step": 3740
|
| 1327 |
-
},
|
| 1328 |
-
{
|
| 1329 |
-
"epoch": 0.5026863975534146,
|
| 1330 |
-
"grad_norm": 2.3359289169311523,
|
| 1331 |
-
"learning_rate": 4.75033538231411e-05,
|
| 1332 |
-
"loss": 0.055599170923233035,
|
| 1333 |
-
"step": 3760
|
| 1334 |
-
},
|
| 1335 |
-
{
|
| 1336 |
-
"epoch": 0.5053602613701881,
|
| 1337 |
-
"grad_norm": 7.426175594329834,
|
| 1338 |
-
"learning_rate": 4.747715233235608e-05,
|
| 1339 |
-
"loss": 0.054436272382736205,
|
| 1340 |
-
"step": 3780
|
| 1341 |
-
},
|
| 1342 |
-
{
|
| 1343 |
-
"epoch": 0.5080341251869616,
|
| 1344 |
-
"grad_norm": 0.5940203070640564,
|
| 1345 |
-
"learning_rate": 4.745082137384128e-05,
|
| 1346 |
-
"loss": 0.03682814538478851,
|
| 1347 |
-
"step": 3800
|
| 1348 |
-
},
|
| 1349 |
-
{
|
| 1350 |
-
"epoch": 0.510707989003735,
|
| 1351 |
-
"grad_norm": 0.22821389138698578,
|
| 1352 |
-
"learning_rate": 4.7424361099262225e-05,
|
| 1353 |
-
"loss": 0.051123309135437014,
|
| 1354 |
-
"step": 3820
|
| 1355 |
-
},
|
| 1356 |
-
{
|
| 1357 |
-
"epoch": 0.5133818528205085,
|
| 1358 |
-
"grad_norm": 8.20633602142334,
|
| 1359 |
-
"learning_rate": 4.739777166102932e-05,
|
| 1360 |
-
"loss": 0.0704378604888916,
|
| 1361 |
-
"step": 3840
|
| 1362 |
-
},
|
| 1363 |
-
{
|
| 1364 |
-
"epoch": 0.516055716637282,
|
| 1365 |
-
"grad_norm": 3.023848533630371,
|
| 1366 |
-
"learning_rate": 4.737105321229694e-05,
|
| 1367 |
-
"loss": 0.03368058800697327,
|
| 1368 |
-
"step": 3860
|
| 1369 |
-
},
|
| 1370 |
-
{
|
| 1371 |
-
"epoch": 0.5187295804540555,
|
| 1372 |
-
"grad_norm": 0.07666649669408798,
|
| 1373 |
-
"learning_rate": 4.7344205906962555e-05,
|
| 1374 |
-
"loss": 0.03665303289890289,
|
| 1375 |
-
"step": 3880
|
| 1376 |
-
},
|
| 1377 |
-
{
|
| 1378 |
-
"epoch": 0.521403444270829,
|
| 1379 |
-
"grad_norm": 0.7571629881858826,
|
| 1380 |
-
"learning_rate": 4.731722989966585e-05,
|
| 1381 |
-
"loss": 0.058415502309799194,
|
| 1382 |
-
"step": 3900
|
| 1383 |
-
},
|
| 1384 |
-
{
|
| 1385 |
-
"epoch": 0.5240773080876024,
|
| 1386 |
-
"grad_norm": 3.2599120140075684,
|
| 1387 |
-
"learning_rate": 4.7290125345787816e-05,
|
| 1388 |
-
"loss": 0.07323018908500671,
|
| 1389 |
-
"step": 3920
|
| 1390 |
-
},
|
| 1391 |
-
{
|
| 1392 |
-
"epoch": 0.5267511719043759,
|
| 1393 |
-
"grad_norm": 0.28930988907814026,
|
| 1394 |
-
"learning_rate": 4.7262892401449886e-05,
|
| 1395 |
-
"loss": 0.054371267557144165,
|
| 1396 |
-
"step": 3940
|
| 1397 |
-
},
|
| 1398 |
-
{
|
| 1399 |
-
"epoch": 0.5294250357211494,
|
| 1400 |
-
"grad_norm": 2.2296454906463623,
|
| 1401 |
-
"learning_rate": 4.7235531223513004e-05,
|
| 1402 |
-
"loss": 0.040819621086120604,
|
| 1403 |
-
"step": 3960
|
| 1404 |
-
},
|
| 1405 |
-
{
|
| 1406 |
-
"epoch": 0.5320988995379229,
|
| 1407 |
-
"grad_norm": 0.11608211696147919,
|
| 1408 |
-
"learning_rate": 4.720804196957675e-05,
|
| 1409 |
-
"loss": 0.05215579271316528,
|
| 1410 |
-
"step": 3980
|
| 1411 |
-
},
|
| 1412 |
-
{
|
| 1413 |
-
"epoch": 0.5347727633546964,
|
| 1414 |
-
"grad_norm": 1.1587547063827515,
|
| 1415 |
-
"learning_rate": 4.7180424797978415e-05,
|
| 1416 |
-
"loss": 0.026277875900268553,
|
| 1417 |
-
"step": 4000
|
| 1418 |
-
},
|
| 1419 |
-
{
|
| 1420 |
-
"epoch": 0.5374466271714698,
|
| 1421 |
-
"grad_norm": 0.06253435462713242,
|
| 1422 |
-
"learning_rate": 4.7152679867792074e-05,
|
| 1423 |
-
"loss": 0.02574407756328583,
|
| 1424 |
-
"step": 4020
|
| 1425 |
-
},
|
| 1426 |
-
{
|
| 1427 |
-
"epoch": 0.5401204909882433,
|
| 1428 |
-
"grad_norm": 1.3441458940505981,
|
| 1429 |
-
"learning_rate": 4.71248073388277e-05,
|
| 1430 |
-
"loss": 0.05538107752799988,
|
| 1431 |
-
"step": 4040
|
| 1432 |
-
},
|
| 1433 |
-
{
|
| 1434 |
-
"epoch": 0.5427943548050168,
|
| 1435 |
-
"grad_norm": 0.48076340556144714,
|
| 1436 |
-
"learning_rate": 4.7096807371630236e-05,
|
| 1437 |
-
"loss": 0.047986540198326114,
|
| 1438 |
-
"step": 4060
|
| 1439 |
-
},
|
| 1440 |
-
{
|
| 1441 |
-
"epoch": 0.5454682186217903,
|
| 1442 |
-
"grad_norm": 0.5924936532974243,
|
| 1443 |
-
"learning_rate": 4.706868012747867e-05,
|
| 1444 |
-
"loss": 0.05463914275169372,
|
| 1445 |
-
"step": 4080
|
| 1446 |
-
},
|
| 1447 |
-
{
|
| 1448 |
-
"epoch": 0.7673980604481948,
|
| 1449 |
-
"grad_norm": 0.4317747950553894,
|
| 1450 |
-
"learning_rate": 4.431151627307268e-05,
|
| 1451 |
-
"loss": 0.020150287946065267,
|
| 1452 |
-
"step": 4100
|
| 1453 |
-
},
|
| 1454 |
-
{
|
| 1455 |
-
"epoch": 0.7711414656211129,
|
| 1456 |
-
"grad_norm": 0.7817056179046631,
|
| 1457 |
-
"learning_rate": 4.425806509248848e-05,
|
| 1458 |
-
"loss": 0.029157504439353943,
|
| 1459 |
-
"step": 4120
|
| 1460 |
-
},
|
| 1461 |
-
{
|
| 1462 |
-
"epoch": 0.7748848707940308,
|
| 1463 |
-
"grad_norm": 4.8055219650268555,
|
| 1464 |
-
"learning_rate": 4.420439652052499e-05,
|
| 1465 |
-
"loss": 0.05370526909828186,
|
| 1466 |
-
"step": 4140
|
| 1467 |
-
},
|
| 1468 |
-
{
|
| 1469 |
-
"epoch": 0.7786282759669488,
|
| 1470 |
-
"grad_norm": 0.5026630759239197,
|
| 1471 |
-
"learning_rate": 4.415051116301072e-05,
|
| 1472 |
-
"loss": 0.03046708106994629,
|
| 1473 |
-
"step": 4160
|
| 1474 |
-
},
|
| 1475 |
-
{
|
| 1476 |
-
"epoch": 0.7823716811398669,
|
| 1477 |
-
"grad_norm": 0.10916770249605179,
|
| 1478 |
-
"learning_rate": 4.409640962822132e-05,
|
| 1479 |
-
"loss": 0.024332405626773836,
|
| 1480 |
-
"step": 4180
|
| 1481 |
-
},
|
| 1482 |
-
{
|
| 1483 |
-
"epoch": 0.7861150863127849,
|
| 1484 |
-
"grad_norm": 0.559330403804779,
|
| 1485 |
-
"learning_rate": 4.404209252687275e-05,
|
| 1486 |
-
"loss": 0.016460689902305602,
|
| 1487 |
-
"step": 4200
|
| 1488 |
-
},
|
| 1489 |
-
{
|
| 1490 |
-
"epoch": 0.789858491485703,
|
| 1491 |
-
"grad_norm": 1.0136655569076538,
|
| 1492 |
-
"learning_rate": 4.398756047211431e-05,
|
| 1493 |
-
"loss": 0.03609788417816162,
|
| 1494 |
-
"step": 4220
|
| 1495 |
-
},
|
| 1496 |
-
{
|
| 1497 |
-
"epoch": 0.7936018966586209,
|
| 1498 |
-
"grad_norm": 1.597655177116394,
|
| 1499 |
-
"learning_rate": 4.39328140795218e-05,
|
| 1500 |
-
"loss": 0.03543405830860138,
|
| 1501 |
-
"step": 4240
|
| 1502 |
-
},
|
| 1503 |
-
{
|
| 1504 |
-
"epoch": 0.797345301831539,
|
| 1505 |
-
"grad_norm": 0.21154026687145233,
|
| 1506 |
-
"learning_rate": 4.387785396709052e-05,
|
| 1507 |
-
"loss": 0.06029776334762573,
|
| 1508 |
-
"step": 4260
|
| 1509 |
-
},
|
| 1510 |
-
{
|
| 1511 |
-
"epoch": 0.801088707004457,
|
| 1512 |
-
"grad_norm": 1.3103065490722656,
|
| 1513 |
-
"learning_rate": 4.382268075522831e-05,
|
| 1514 |
-
"loss": 0.05260368585586548,
|
| 1515 |
-
"step": 4280
|
| 1516 |
-
},
|
| 1517 |
-
{
|
| 1518 |
-
"epoch": 0.804832112177375,
|
| 1519 |
-
"grad_norm": 0.07966366410255432,
|
| 1520 |
-
"learning_rate": 4.3767295066748564e-05,
|
| 1521 |
-
"loss": 0.02691380977630615,
|
| 1522 |
-
"step": 4300
|
| 1523 |
-
},
|
| 1524 |
-
{
|
| 1525 |
-
"epoch": 0.8085755173502931,
|
| 1526 |
-
"grad_norm": 0.17526337504386902,
|
| 1527 |
-
"learning_rate": 4.371169752686316e-05,
|
| 1528 |
-
"loss": 0.02926369905471802,
|
| 1529 |
-
"step": 4320
|
| 1530 |
-
},
|
| 1531 |
-
{
|
| 1532 |
-
"epoch": 0.812318922523211,
|
| 1533 |
-
"grad_norm": 1.6024236679077148,
|
| 1534 |
-
"learning_rate": 4.3655888763175436e-05,
|
| 1535 |
-
"loss": 0.09227399826049805,
|
| 1536 |
-
"step": 4340
|
| 1537 |
-
},
|
| 1538 |
-
{
|
| 1539 |
-
"epoch": 0.8160623276961291,
|
| 1540 |
-
"grad_norm": 2.4870424270629883,
|
| 1541 |
-
"learning_rate": 4.3599869405673085e-05,
|
| 1542 |
-
"loss": 0.05842729210853577,
|
| 1543 |
-
"step": 4360
|
| 1544 |
-
},
|
| 1545 |
-
{
|
| 1546 |
-
"epoch": 0.8198057328690471,
|
| 1547 |
-
"grad_norm": 0.27151742577552795,
|
| 1548 |
-
"learning_rate": 4.354364008672106e-05,
|
| 1549 |
-
"loss": 0.03142921030521393,
|
| 1550 |
-
"step": 4380
|
| 1551 |
-
},
|
| 1552 |
-
{
|
| 1553 |
-
"epoch": 0.8235491380419652,
|
| 1554 |
-
"grad_norm": 0.06985723227262497,
|
| 1555 |
-
"learning_rate": 4.3487201441054435e-05,
|
| 1556 |
-
"loss": 0.021850427985191344,
|
| 1557 |
-
"step": 4400
|
| 1558 |
-
},
|
| 1559 |
-
{
|
| 1560 |
-
"epoch": 0.8272925432148831,
|
| 1561 |
-
"grad_norm": 0.9921057224273682,
|
| 1562 |
-
"learning_rate": 4.343055410577122e-05,
|
| 1563 |
-
"loss": 0.03694712221622467,
|
| 1564 |
-
"step": 4420
|
| 1565 |
-
},
|
| 1566 |
-
{
|
| 1567 |
-
"epoch": 0.8310359483878011,
|
| 1568 |
-
"grad_norm": 3.330061912536621,
|
| 1569 |
-
"learning_rate": 4.3373698720325176e-05,
|
| 1570 |
-
"loss": 0.04195646345615387,
|
| 1571 |
-
"step": 4440
|
| 1572 |
-
},
|
| 1573 |
-
{
|
| 1574 |
-
"epoch": 0.8347793535607192,
|
| 1575 |
-
"grad_norm": 4.173893928527832,
|
| 1576 |
-
"learning_rate": 4.331663592651862e-05,
|
| 1577 |
-
"loss": 0.025657615065574645,
|
| 1578 |
-
"step": 4460
|
| 1579 |
-
},
|
| 1580 |
-
{
|
| 1581 |
-
"epoch": 0.8385227587336372,
|
| 1582 |
-
"grad_norm": 0.514607846736908,
|
| 1583 |
-
"learning_rate": 4.3259366368495167e-05,
|
| 1584 |
-
"loss": 0.018329595029354096,
|
| 1585 |
-
"step": 4480
|
| 1586 |
-
}
|
| 1587 |
-
],
|
| 1588 |
-
"logging_steps": 20,
|
| 1589 |
-
"max_steps": 18701,
|
| 1590 |
-
"num_input_tokens_seen": 0,
|
| 1591 |
-
"num_train_epochs": 4,
|
| 1592 |
-
"save_steps": 1000000000,
|
| 1593 |
-
"stateful_callbacks": {
|
| 1594 |
-
"TrainerControl": {
|
| 1595 |
-
"args": {
|
| 1596 |
-
"should_epoch_stop": false,
|
| 1597 |
-
"should_evaluate": false,
|
| 1598 |
-
"should_log": false,
|
| 1599 |
-
"should_save": true,
|
| 1600 |
-
"should_training_stop": false
|
| 1601 |
-
},
|
| 1602 |
-
"attributes": {}
|
| 1603 |
-
}
|
| 1604 |
-
},
|
| 1605 |
-
"total_flos": 1.7212871234617344e+16,
|
| 1606 |
-
"train_batch_size": 1,
|
| 1607 |
-
"trial_name": null,
|
| 1608 |
-
"trial_params": null
|
| 1609 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/16/checkpoint-1731/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/16/checkpoint-1731/trainer_state.json
DELETED
|
@@ -1,643 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.23142291334174486,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 1731,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001336931908386741,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 150.77503967285156,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.002673863816773482,
|
| 21 |
-
"grad_norm": 65.5747299194336,
|
| 22 |
-
"learning_rate": 4.999995392022967e-05,
|
| 23 |
-
"loss": 65.57097424958882,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.005347727633546964,
|
| 28 |
-
"grad_norm": 29.183452606201172,
|
| 29 |
-
"learning_rate": 4.999976672145381e-05,
|
| 30 |
-
"loss": 12.781006622314454,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008021591450320446,
|
| 35 |
-
"grad_norm": 22.14234161376953,
|
| 36 |
-
"learning_rate": 4.999943552476422e-05,
|
| 37 |
-
"loss": 7.058085632324219,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.010695455267093928,
|
| 42 |
-
"grad_norm": 25.116792678833008,
|
| 43 |
-
"learning_rate": 4.999896033206858e-05,
|
| 44 |
-
"loss": 5.17547492980957,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.01336931908386741,
|
| 49 |
-
"grad_norm": 17.95206642150879,
|
| 50 |
-
"learning_rate": 4.999834114610398e-05,
|
| 51 |
-
"loss": 3.6641620635986327,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.01604318290064089,
|
| 56 |
-
"grad_norm": 13.571253776550293,
|
| 57 |
-
"learning_rate": 4.999757797043691e-05,
|
| 58 |
-
"loss": 2.9684438705444336,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.018717046717414372,
|
| 63 |
-
"grad_norm": 11.740418434143066,
|
| 64 |
-
"learning_rate": 4.999667080946324e-05,
|
| 65 |
-
"loss": 2.658764457702637,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.021390910534187856,
|
| 70 |
-
"grad_norm": 8.911559104919434,
|
| 71 |
-
"learning_rate": 4.999561966840821e-05,
|
| 72 |
-
"loss": 2.0916315078735352,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.024064774350961337,
|
| 77 |
-
"grad_norm": 7.613410472869873,
|
| 78 |
-
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
-
"loss": 1.730122184753418,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02673863816773482,
|
| 84 |
-
"grad_norm": 10.990550994873047,
|
| 85 |
-
"learning_rate": 4.999308547110146e-05,
|
| 86 |
-
"loss": 1.689904022216797,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.029412501984508302,
|
| 91 |
-
"grad_norm": 13.414973258972168,
|
| 92 |
-
"learning_rate": 4.999160242944665e-05,
|
| 93 |
-
"loss": 1.331161880493164,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.03208636580128178,
|
| 98 |
-
"grad_norm": 9.434646606445312,
|
| 99 |
-
"learning_rate": 4.998997543690418e-05,
|
| 100 |
-
"loss": 1.285347843170166,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.03476022961805526,
|
| 105 |
-
"grad_norm": 15.048566818237305,
|
| 106 |
-
"learning_rate": 4.998820450284549e-05,
|
| 107 |
-
"loss": 1.1024582862854004,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.037434093434828744,
|
| 112 |
-
"grad_norm": 7.001116752624512,
|
| 113 |
-
"learning_rate": 4.99862896374711e-05,
|
| 114 |
-
"loss": 1.106878662109375,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04010795725160223,
|
| 119 |
-
"grad_norm": 10.976016998291016,
|
| 120 |
-
"learning_rate": 4.998423085181056e-05,
|
| 121 |
-
"loss": 1.0516901969909669,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04278182106837571,
|
| 126 |
-
"grad_norm": 10.106513977050781,
|
| 127 |
-
"learning_rate": 4.998202815772245e-05,
|
| 128 |
-
"loss": 0.8515177726745605,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.04545568488514919,
|
| 133 |
-
"grad_norm": 3.9024946689605713,
|
| 134 |
-
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
-
"loss": 0.851468849182129,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.048129548701922674,
|
| 140 |
-
"grad_norm": 7.649858474731445,
|
| 141 |
-
"learning_rate": 4.997719109584209e-05,
|
| 142 |
-
"loss": 0.8077478408813477,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.050803412518696155,
|
| 147 |
-
"grad_norm": 11.689629554748535,
|
| 148 |
-
"learning_rate": 4.997455675591119e-05,
|
| 149 |
-
"loss": 0.6951910495758057,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.05347727633546964,
|
| 154 |
-
"grad_norm": 12.950127601623535,
|
| 155 |
-
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
-
"loss": 0.6042420387268066,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.05615114015224312,
|
| 161 |
-
"grad_norm": 6.282811641693115,
|
| 162 |
-
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
-
"loss": 0.6595364093780518,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.058825003969016604,
|
| 168 |
-
"grad_norm": 3.553151845932007,
|
| 169 |
-
"learning_rate": 4.99657906847257e-05,
|
| 170 |
-
"loss": 0.71713285446167,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.061498867785790085,
|
| 175 |
-
"grad_norm": 5.715928554534912,
|
| 176 |
-
"learning_rate": 4.996258103330218e-05,
|
| 177 |
-
"loss": 0.5080134391784668,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.06417273160256357,
|
| 182 |
-
"grad_norm": 8.924579620361328,
|
| 183 |
-
"learning_rate": 4.995922759815339e-05,
|
| 184 |
-
"loss": 0.4877929210662842,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.06684659541933705,
|
| 189 |
-
"grad_norm": 4.2473649978637695,
|
| 190 |
-
"learning_rate": 4.995573039859501e-05,
|
| 191 |
-
"loss": 0.5352929592132568,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.06952045923611053,
|
| 196 |
-
"grad_norm": 5.880599021911621,
|
| 197 |
-
"learning_rate": 4.995208945477081e-05,
|
| 198 |
-
"loss": 0.520254373550415,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.07219432305288401,
|
| 203 |
-
"grad_norm": 8.42399787902832,
|
| 204 |
-
"learning_rate": 4.994830478765251e-05,
|
| 205 |
-
"loss": 0.5289288520812988,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.07486818686965749,
|
| 210 |
-
"grad_norm": 5.284440040588379,
|
| 211 |
-
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
-
"loss": 0.4858880043029785,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.07754205068643098,
|
| 217 |
-
"grad_norm": 4.662903308868408,
|
| 218 |
-
"learning_rate": 4.994030437155961e-05,
|
| 219 |
-
"loss": 0.4253716468811035,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.08021591450320446,
|
| 224 |
-
"grad_norm": 2.9051289558410645,
|
| 225 |
-
"learning_rate": 4.993608866866718e-05,
|
| 226 |
-
"loss": 0.32270302772521975,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.08288977831997794,
|
| 231 |
-
"grad_norm": 2.804189920425415,
|
| 232 |
-
"learning_rate": 4.993172933464471e-05,
|
| 233 |
-
"loss": 0.329010009765625,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.08556364213675142,
|
| 238 |
-
"grad_norm": 1.455569863319397,
|
| 239 |
-
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
-
"loss": 0.35200178623199463,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.0882375059535249,
|
| 245 |
-
"grad_norm": 11.18322467803955,
|
| 246 |
-
"learning_rate": 4.992257987447532e-05,
|
| 247 |
-
"loss": 0.30103607177734376,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.09091136977029839,
|
| 252 |
-
"grad_norm": 6.4625773429870605,
|
| 253 |
-
"learning_rate": 4.991778980102904e-05,
|
| 254 |
-
"loss": 0.3078582525253296,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.09358523358707187,
|
| 259 |
-
"grad_norm": 11.98585033416748,
|
| 260 |
-
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
-
"loss": 0.30774364471435545,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.09625909740384535,
|
| 266 |
-
"grad_norm": 10.583540916442871,
|
| 267 |
-
"learning_rate": 4.990777910536653e-05,
|
| 268 |
-
"loss": 0.30326850414276124,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.09893296122061883,
|
| 273 |
-
"grad_norm": 6.04772424697876,
|
| 274 |
-
"learning_rate": 4.990255854081161e-05,
|
| 275 |
-
"loss": 0.37041378021240234,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.10160682503739231,
|
| 280 |
-
"grad_norm": 7.155930519104004,
|
| 281 |
-
"learning_rate": 4.989719453825918e-05,
|
| 282 |
-
"loss": 0.33830211162567136,
|
| 283 |
-
"step": 760
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.10428068885416579,
|
| 287 |
-
"grad_norm": 6.669356346130371,
|
| 288 |
-
"learning_rate": 4.9891687128605744e-05,
|
| 289 |
-
"loss": 0.3225187540054321,
|
| 290 |
-
"step": 780
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.10695455267093928,
|
| 294 |
-
"grad_norm": 2.8060507774353027,
|
| 295 |
-
"learning_rate": 4.988603634357383e-05,
|
| 296 |
-
"loss": 0.27278661727905273,
|
| 297 |
-
"step": 800
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.10962841648771277,
|
| 301 |
-
"grad_norm": 4.8071699142456055,
|
| 302 |
-
"learning_rate": 4.988024221571177e-05,
|
| 303 |
-
"loss": 0.2937103986740112,
|
| 304 |
-
"step": 820
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.11230228030448625,
|
| 308 |
-
"grad_norm": 5.857818603515625,
|
| 309 |
-
"learning_rate": 4.9874304778393574e-05,
|
| 310 |
-
"loss": 0.17092934846878052,
|
| 311 |
-
"step": 840
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.11497614412125973,
|
| 315 |
-
"grad_norm": 3.4762802124023438,
|
| 316 |
-
"learning_rate": 4.9868224065818706e-05,
|
| 317 |
-
"loss": 0.1812203049659729,
|
| 318 |
-
"step": 860
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.11765000793803321,
|
| 322 |
-
"grad_norm": 6.1022186279296875,
|
| 323 |
-
"learning_rate": 4.98620001130119e-05,
|
| 324 |
-
"loss": 0.2638218879699707,
|
| 325 |
-
"step": 880
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.12032387175480669,
|
| 329 |
-
"grad_norm": 7.3963799476623535,
|
| 330 |
-
"learning_rate": 4.9855632955822916e-05,
|
| 331 |
-
"loss": 0.25242090225219727,
|
| 332 |
-
"step": 900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.12299773557158017,
|
| 336 |
-
"grad_norm": 1.6199634075164795,
|
| 337 |
-
"learning_rate": 4.984912263092641e-05,
|
| 338 |
-
"loss": 0.2603979349136353,
|
| 339 |
-
"step": 920
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.12567159938835365,
|
| 343 |
-
"grad_norm": 8.293638229370117,
|
| 344 |
-
"learning_rate": 4.984246917582166e-05,
|
| 345 |
-
"loss": 0.2023364543914795,
|
| 346 |
-
"step": 940
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.12834546320512713,
|
| 350 |
-
"grad_norm": 8.61099910736084,
|
| 351 |
-
"learning_rate": 4.9835672628832366e-05,
|
| 352 |
-
"loss": 0.16875293254852294,
|
| 353 |
-
"step": 960
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.1310193270219006,
|
| 357 |
-
"grad_norm": 0.5758953094482422,
|
| 358 |
-
"learning_rate": 4.9828733029106434e-05,
|
| 359 |
-
"loss": 0.24272210597991944,
|
| 360 |
-
"step": 980
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.1336931908386741,
|
| 364 |
-
"grad_norm": 6.830381870269775,
|
| 365 |
-
"learning_rate": 4.982165041661575e-05,
|
| 366 |
-
"loss": 0.1900513768196106,
|
| 367 |
-
"step": 1000
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.13636705465544757,
|
| 371 |
-
"grad_norm": 3.979989528656006,
|
| 372 |
-
"learning_rate": 4.981442483215595e-05,
|
| 373 |
-
"loss": 0.23804061412811278,
|
| 374 |
-
"step": 1020
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.13904091847222105,
|
| 378 |
-
"grad_norm": 3.480762004852295,
|
| 379 |
-
"learning_rate": 4.98070563173462e-05,
|
| 380 |
-
"loss": 0.19694632291793823,
|
| 381 |
-
"step": 1040
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.14171478228899453,
|
| 385 |
-
"grad_norm": 2.7838668823242188,
|
| 386 |
-
"learning_rate": 4.979954491462892e-05,
|
| 387 |
-
"loss": 0.20703573226928712,
|
| 388 |
-
"step": 1060
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.14438864610576801,
|
| 392 |
-
"grad_norm": 4.724070072174072,
|
| 393 |
-
"learning_rate": 4.979189066726955e-05,
|
| 394 |
-
"loss": 0.135595703125,
|
| 395 |
-
"step": 1080
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.1470625099225415,
|
| 399 |
-
"grad_norm": 6.755579471588135,
|
| 400 |
-
"learning_rate": 4.978409361935636e-05,
|
| 401 |
-
"loss": 0.20947463512420655,
|
| 402 |
-
"step": 1100
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.14973637373931498,
|
| 406 |
-
"grad_norm": 2.989082098007202,
|
| 407 |
-
"learning_rate": 4.9776153815800075e-05,
|
| 408 |
-
"loss": 0.15855485200881958,
|
| 409 |
-
"step": 1120
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.15241023755608848,
|
| 413 |
-
"grad_norm": 2.310991048812866,
|
| 414 |
-
"learning_rate": 4.976807130233375e-05,
|
| 415 |
-
"loss": 0.1401826858520508,
|
| 416 |
-
"step": 1140
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.15508410137286197,
|
| 420 |
-
"grad_norm": 3.802248477935791,
|
| 421 |
-
"learning_rate": 4.975984612551243e-05,
|
| 422 |
-
"loss": 0.17620396614074707,
|
| 423 |
-
"step": 1160
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.15775796518963545,
|
| 427 |
-
"grad_norm": 2.0629191398620605,
|
| 428 |
-
"learning_rate": 4.975147833271288e-05,
|
| 429 |
-
"loss": 0.21862537860870362,
|
| 430 |
-
"step": 1180
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.16043182900640893,
|
| 434 |
-
"grad_norm": 2.2114040851593018,
|
| 435 |
-
"learning_rate": 4.9742967972133335e-05,
|
| 436 |
-
"loss": 0.15507786273956298,
|
| 437 |
-
"step": 1200
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.1631056928231824,
|
| 441 |
-
"grad_norm": 10.362643241882324,
|
| 442 |
-
"learning_rate": 4.973431509279323e-05,
|
| 443 |
-
"loss": 0.15983071327209472,
|
| 444 |
-
"step": 1220
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.1657795566399559,
|
| 448 |
-
"grad_norm": 0.3497254252433777,
|
| 449 |
-
"learning_rate": 4.972551974453287e-05,
|
| 450 |
-
"loss": 0.20826480388641358,
|
| 451 |
-
"step": 1240
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.16845342045672937,
|
| 455 |
-
"grad_norm": 0.3897332549095154,
|
| 456 |
-
"learning_rate": 4.971658197801322e-05,
|
| 457 |
-
"loss": 0.12974958419799804,
|
| 458 |
-
"step": 1260
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.17112728427350285,
|
| 462 |
-
"grad_norm": 6.043999195098877,
|
| 463 |
-
"learning_rate": 4.9707501844715554e-05,
|
| 464 |
-
"loss": 0.1621025800704956,
|
| 465 |
-
"step": 1280
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.17380114809027633,
|
| 469 |
-
"grad_norm": 1.7294467687606812,
|
| 470 |
-
"learning_rate": 4.969827939694115e-05,
|
| 471 |
-
"loss": 0.17147176265716552,
|
| 472 |
-
"step": 1300
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.1764750119070498,
|
| 476 |
-
"grad_norm": 0.8541621565818787,
|
| 477 |
-
"learning_rate": 4.968891468781105e-05,
|
| 478 |
-
"loss": 0.1330248236656189,
|
| 479 |
-
"step": 1320
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.1791488757238233,
|
| 483 |
-
"grad_norm": 2.197866439819336,
|
| 484 |
-
"learning_rate": 4.967940777126569e-05,
|
| 485 |
-
"loss": 0.12178658246994019,
|
| 486 |
-
"step": 1340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.18182273954059677,
|
| 490 |
-
"grad_norm": 0.8711172342300415,
|
| 491 |
-
"learning_rate": 4.9669758702064636e-05,
|
| 492 |
-
"loss": 0.19501737356185914,
|
| 493 |
-
"step": 1360
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.18449660335737025,
|
| 497 |
-
"grad_norm": 5.356688022613525,
|
| 498 |
-
"learning_rate": 4.965996753578623e-05,
|
| 499 |
-
"loss": 0.1602308750152588,
|
| 500 |
-
"step": 1380
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.18717046717414373,
|
| 504 |
-
"grad_norm": 13.699594497680664,
|
| 505 |
-
"learning_rate": 4.9650034328827305e-05,
|
| 506 |
-
"loss": 0.12051005363464355,
|
| 507 |
-
"step": 1400
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.18984433099091721,
|
| 511 |
-
"grad_norm": 1.738416314125061,
|
| 512 |
-
"learning_rate": 4.963995913840284e-05,
|
| 513 |
-
"loss": 0.1177564263343811,
|
| 514 |
-
"step": 1420
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.1925181948076907,
|
| 518 |
-
"grad_norm": 5.713410377502441,
|
| 519 |
-
"learning_rate": 4.9629742022545623e-05,
|
| 520 |
-
"loss": 0.19124844074249267,
|
| 521 |
-
"step": 1440
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.19519205862446418,
|
| 525 |
-
"grad_norm": 4.171947479248047,
|
| 526 |
-
"learning_rate": 4.961938304010595e-05,
|
| 527 |
-
"loss": 0.12269804477691651,
|
| 528 |
-
"step": 1460
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.19786592244123766,
|
| 532 |
-
"grad_norm": 2.8535916805267334,
|
| 533 |
-
"learning_rate": 4.9608882250751245e-05,
|
| 534 |
-
"loss": 0.0892150342464447,
|
| 535 |
-
"step": 1480
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.20053978625801114,
|
| 539 |
-
"grad_norm": 0.6551859378814697,
|
| 540 |
-
"learning_rate": 4.959823971496574e-05,
|
| 541 |
-
"loss": 0.10962753295898438,
|
| 542 |
-
"step": 1500
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.20321365007478462,
|
| 546 |
-
"grad_norm": 6.525365352630615,
|
| 547 |
-
"learning_rate": 4.9587455494050136e-05,
|
| 548 |
-
"loss": 0.10057367086410522,
|
| 549 |
-
"step": 1520
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.2058875138915581,
|
| 553 |
-
"grad_norm": 4.142037391662598,
|
| 554 |
-
"learning_rate": 4.9576529650121214e-05,
|
| 555 |
-
"loss": 0.1188935399055481,
|
| 556 |
-
"step": 1540
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.20856137770833158,
|
| 560 |
-
"grad_norm": 1.3209469318389893,
|
| 561 |
-
"learning_rate": 4.956546224611152e-05,
|
| 562 |
-
"loss": 0.10144028663635254,
|
| 563 |
-
"step": 1560
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.21123524152510506,
|
| 567 |
-
"grad_norm": 2.7724432945251465,
|
| 568 |
-
"learning_rate": 4.9554253345768965e-05,
|
| 569 |
-
"loss": 0.09022048711776734,
|
| 570 |
-
"step": 1580
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.21390910534187857,
|
| 574 |
-
"grad_norm": 1.7022371292114258,
|
| 575 |
-
"learning_rate": 4.9542903013656486e-05,
|
| 576 |
-
"loss": 0.13176963329315186,
|
| 577 |
-
"step": 1600
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.21658296915865205,
|
| 581 |
-
"grad_norm": 4.1148681640625,
|
| 582 |
-
"learning_rate": 4.9531411315151654e-05,
|
| 583 |
-
"loss": 0.08941383361816406,
|
| 584 |
-
"step": 1620
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.21925683297542553,
|
| 588 |
-
"grad_norm": 1.5328305959701538,
|
| 589 |
-
"learning_rate": 4.951977831644632e-05,
|
| 590 |
-
"loss": 0.12352118492126465,
|
| 591 |
-
"step": 1640
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.221930696792199,
|
| 595 |
-
"grad_norm": 2.8182177543640137,
|
| 596 |
-
"learning_rate": 4.95080040845462e-05,
|
| 597 |
-
"loss": 0.12917338609695433,
|
| 598 |
-
"step": 1660
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.2246045606089725,
|
| 602 |
-
"grad_norm": 2.4821066856384277,
|
| 603 |
-
"learning_rate": 4.949608868727053e-05,
|
| 604 |
-
"loss": 0.12498313188552856,
|
| 605 |
-
"step": 1680
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.22727842442574597,
|
| 609 |
-
"grad_norm": 2.292102336883545,
|
| 610 |
-
"learning_rate": 4.948403219325163e-05,
|
| 611 |
-
"loss": 0.10659204721450806,
|
| 612 |
-
"step": 1700
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.22995228824251945,
|
| 616 |
-
"grad_norm": 0.4930734634399414,
|
| 617 |
-
"learning_rate": 4.947183467193456e-05,
|
| 618 |
-
"loss": 0.0782310426235199,
|
| 619 |
-
"step": 1720
|
| 620 |
-
}
|
| 621 |
-
],
|
| 622 |
-
"logging_steps": 20,
|
| 623 |
-
"max_steps": 26180,
|
| 624 |
-
"num_input_tokens_seen": 0,
|
| 625 |
-
"num_train_epochs": 4,
|
| 626 |
-
"save_steps": 1000000000,
|
| 627 |
-
"stateful_callbacks": {
|
| 628 |
-
"TrainerControl": {
|
| 629 |
-
"args": {
|
| 630 |
-
"should_epoch_stop": false,
|
| 631 |
-
"should_evaluate": false,
|
| 632 |
-
"should_log": false,
|
| 633 |
-
"should_save": true,
|
| 634 |
-
"should_training_stop": false
|
| 635 |
-
},
|
| 636 |
-
"attributes": {}
|
| 637 |
-
}
|
| 638 |
-
},
|
| 639 |
-
"total_flos": 6638921592496128.0,
|
| 640 |
-
"train_batch_size": 1,
|
| 641 |
-
"trial_name": null,
|
| 642 |
-
"trial_params": null
|
| 643 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/4/checkpoint-1539/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/4/checkpoint-1539/trainer_state.json
DELETED
|
@@ -1,573 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.20575382070071943,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 1539,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001336931908386741,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 150.77503967285156,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.002673863816773482,
|
| 21 |
-
"grad_norm": 65.5747299194336,
|
| 22 |
-
"learning_rate": 4.999995392022967e-05,
|
| 23 |
-
"loss": 65.57097424958882,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.005347727633546964,
|
| 28 |
-
"grad_norm": 29.183452606201172,
|
| 29 |
-
"learning_rate": 4.999976672145381e-05,
|
| 30 |
-
"loss": 12.781006622314454,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.008021591450320446,
|
| 35 |
-
"grad_norm": 22.14234161376953,
|
| 36 |
-
"learning_rate": 4.999943552476422e-05,
|
| 37 |
-
"loss": 7.058085632324219,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.010695455267093928,
|
| 42 |
-
"grad_norm": 25.116792678833008,
|
| 43 |
-
"learning_rate": 4.999896033206858e-05,
|
| 44 |
-
"loss": 5.17547492980957,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.01336931908386741,
|
| 49 |
-
"grad_norm": 17.95206642150879,
|
| 50 |
-
"learning_rate": 4.999834114610398e-05,
|
| 51 |
-
"loss": 3.6641620635986327,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.01604318290064089,
|
| 56 |
-
"grad_norm": 13.571253776550293,
|
| 57 |
-
"learning_rate": 4.999757797043691e-05,
|
| 58 |
-
"loss": 2.9684438705444336,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.018717046717414372,
|
| 63 |
-
"grad_norm": 11.740418434143066,
|
| 64 |
-
"learning_rate": 4.999667080946324e-05,
|
| 65 |
-
"loss": 2.658764457702637,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.021390910534187856,
|
| 70 |
-
"grad_norm": 8.911559104919434,
|
| 71 |
-
"learning_rate": 4.999561966840821e-05,
|
| 72 |
-
"loss": 2.0916315078735352,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.024064774350961337,
|
| 77 |
-
"grad_norm": 7.613410472869873,
|
| 78 |
-
"learning_rate": 4.9994424553326335e-05,
|
| 79 |
-
"loss": 1.730122184753418,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.02673863816773482,
|
| 84 |
-
"grad_norm": 10.990550994873047,
|
| 85 |
-
"learning_rate": 4.999308547110146e-05,
|
| 86 |
-
"loss": 1.689904022216797,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.029412501984508302,
|
| 91 |
-
"grad_norm": 13.414973258972168,
|
| 92 |
-
"learning_rate": 4.999160242944665e-05,
|
| 93 |
-
"loss": 1.331161880493164,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.03208636580128178,
|
| 98 |
-
"grad_norm": 9.434646606445312,
|
| 99 |
-
"learning_rate": 4.998997543690418e-05,
|
| 100 |
-
"loss": 1.285347843170166,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.03476022961805526,
|
| 105 |
-
"grad_norm": 15.048566818237305,
|
| 106 |
-
"learning_rate": 4.998820450284549e-05,
|
| 107 |
-
"loss": 1.1024582862854004,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.037434093434828744,
|
| 112 |
-
"grad_norm": 7.001116752624512,
|
| 113 |
-
"learning_rate": 4.99862896374711e-05,
|
| 114 |
-
"loss": 1.106878662109375,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04010795725160223,
|
| 119 |
-
"grad_norm": 10.976016998291016,
|
| 120 |
-
"learning_rate": 4.998423085181056e-05,
|
| 121 |
-
"loss": 1.0516901969909669,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04278182106837571,
|
| 126 |
-
"grad_norm": 10.106513977050781,
|
| 127 |
-
"learning_rate": 4.998202815772245e-05,
|
| 128 |
-
"loss": 0.8515177726745605,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.04545568488514919,
|
| 133 |
-
"grad_norm": 3.9024946689605713,
|
| 134 |
-
"learning_rate": 4.9979681567894195e-05,
|
| 135 |
-
"loss": 0.851468849182129,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.048129548701922674,
|
| 140 |
-
"grad_norm": 7.649858474731445,
|
| 141 |
-
"learning_rate": 4.997719109584209e-05,
|
| 142 |
-
"loss": 0.8077478408813477,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.050803412518696155,
|
| 147 |
-
"grad_norm": 11.689629554748535,
|
| 148 |
-
"learning_rate": 4.997455675591119e-05,
|
| 149 |
-
"loss": 0.6951910495758057,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.05347727633546964,
|
| 154 |
-
"grad_norm": 12.950127601623535,
|
| 155 |
-
"learning_rate": 4.9971778563275204e-05,
|
| 156 |
-
"loss": 0.6042420387268066,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.05615114015224312,
|
| 161 |
-
"grad_norm": 6.282811641693115,
|
| 162 |
-
"learning_rate": 4.9968856533936436e-05,
|
| 163 |
-
"loss": 0.6595364093780518,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.058825003969016604,
|
| 168 |
-
"grad_norm": 3.553151845932007,
|
| 169 |
-
"learning_rate": 4.99657906847257e-05,
|
| 170 |
-
"loss": 0.71713285446167,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.061498867785790085,
|
| 175 |
-
"grad_norm": 5.715928554534912,
|
| 176 |
-
"learning_rate": 4.996258103330218e-05,
|
| 177 |
-
"loss": 0.5080134391784668,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.06417273160256357,
|
| 182 |
-
"grad_norm": 8.924579620361328,
|
| 183 |
-
"learning_rate": 4.995922759815339e-05,
|
| 184 |
-
"loss": 0.4877929210662842,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.06684659541933705,
|
| 189 |
-
"grad_norm": 4.2473649978637695,
|
| 190 |
-
"learning_rate": 4.995573039859501e-05,
|
| 191 |
-
"loss": 0.5352929592132568,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.06952045923611053,
|
| 196 |
-
"grad_norm": 5.880599021911621,
|
| 197 |
-
"learning_rate": 4.995208945477081e-05,
|
| 198 |
-
"loss": 0.520254373550415,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.07219432305288401,
|
| 203 |
-
"grad_norm": 8.42399787902832,
|
| 204 |
-
"learning_rate": 4.994830478765251e-05,
|
| 205 |
-
"loss": 0.5289288520812988,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.07486818686965749,
|
| 210 |
-
"grad_norm": 5.284440040588379,
|
| 211 |
-
"learning_rate": 4.9944376419039684e-05,
|
| 212 |
-
"loss": 0.4858880043029785,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.07754205068643098,
|
| 217 |
-
"grad_norm": 4.662903308868408,
|
| 218 |
-
"learning_rate": 4.994030437155961e-05,
|
| 219 |
-
"loss": 0.4253716468811035,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.08021591450320446,
|
| 224 |
-
"grad_norm": 2.9051289558410645,
|
| 225 |
-
"learning_rate": 4.993608866866718e-05,
|
| 226 |
-
"loss": 0.32270302772521975,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.08288977831997794,
|
| 231 |
-
"grad_norm": 2.804189920425415,
|
| 232 |
-
"learning_rate": 4.993172933464471e-05,
|
| 233 |
-
"loss": 0.329010009765625,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.08556364213675142,
|
| 238 |
-
"grad_norm": 1.455569863319397,
|
| 239 |
-
"learning_rate": 4.9927226394601815e-05,
|
| 240 |
-
"loss": 0.35200178623199463,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.0882375059535249,
|
| 245 |
-
"grad_norm": 11.18322467803955,
|
| 246 |
-
"learning_rate": 4.992257987447532e-05,
|
| 247 |
-
"loss": 0.30103607177734376,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.09091136977029839,
|
| 252 |
-
"grad_norm": 6.4625773429870605,
|
| 253 |
-
"learning_rate": 4.991778980102904e-05,
|
| 254 |
-
"loss": 0.3078582525253296,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.09358523358707187,
|
| 259 |
-
"grad_norm": 11.98585033416748,
|
| 260 |
-
"learning_rate": 4.9912856201853644e-05,
|
| 261 |
-
"loss": 0.30774364471435545,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.09625909740384535,
|
| 266 |
-
"grad_norm": 10.583540916442871,
|
| 267 |
-
"learning_rate": 4.990777910536653e-05,
|
| 268 |
-
"loss": 0.30326850414276124,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.09893296122061883,
|
| 273 |
-
"grad_norm": 6.04772424697876,
|
| 274 |
-
"learning_rate": 4.990255854081161e-05,
|
| 275 |
-
"loss": 0.37041378021240234,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.10160682503739231,
|
| 280 |
-
"grad_norm": 7.155930519104004,
|
| 281 |
-
"learning_rate": 4.989719453825918e-05,
|
| 282 |
-
"loss": 0.33830211162567136,
|
| 283 |
-
"step": 760
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.10428068885416579,
|
| 287 |
-
"grad_norm": 6.669356346130371,
|
| 288 |
-
"learning_rate": 4.9891687128605744e-05,
|
| 289 |
-
"loss": 0.3225187540054321,
|
| 290 |
-
"step": 780
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.10695455267093928,
|
| 294 |
-
"grad_norm": 2.8060507774353027,
|
| 295 |
-
"learning_rate": 4.988603634357383e-05,
|
| 296 |
-
"loss": 0.27278661727905273,
|
| 297 |
-
"step": 800
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.10962841648771277,
|
| 301 |
-
"grad_norm": 4.8071699142456055,
|
| 302 |
-
"learning_rate": 4.988024221571177e-05,
|
| 303 |
-
"loss": 0.2937103986740112,
|
| 304 |
-
"step": 820
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.11230228030448625,
|
| 308 |
-
"grad_norm": 5.857818603515625,
|
| 309 |
-
"learning_rate": 4.9874304778393574e-05,
|
| 310 |
-
"loss": 0.17092934846878052,
|
| 311 |
-
"step": 840
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.11497614412125973,
|
| 315 |
-
"grad_norm": 3.4762802124023438,
|
| 316 |
-
"learning_rate": 4.9868224065818706e-05,
|
| 317 |
-
"loss": 0.1812203049659729,
|
| 318 |
-
"step": 860
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.11765000793803321,
|
| 322 |
-
"grad_norm": 6.1022186279296875,
|
| 323 |
-
"learning_rate": 4.98620001130119e-05,
|
| 324 |
-
"loss": 0.2638218879699707,
|
| 325 |
-
"step": 880
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.12032387175480669,
|
| 329 |
-
"grad_norm": 7.3963799476623535,
|
| 330 |
-
"learning_rate": 4.9855632955822916e-05,
|
| 331 |
-
"loss": 0.25242090225219727,
|
| 332 |
-
"step": 900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.12299773557158017,
|
| 336 |
-
"grad_norm": 1.6199634075164795,
|
| 337 |
-
"learning_rate": 4.984912263092641e-05,
|
| 338 |
-
"loss": 0.2603979349136353,
|
| 339 |
-
"step": 920
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.12567159938835365,
|
| 343 |
-
"grad_norm": 8.293638229370117,
|
| 344 |
-
"learning_rate": 4.984246917582166e-05,
|
| 345 |
-
"loss": 0.2023364543914795,
|
| 346 |
-
"step": 940
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.12834546320512713,
|
| 350 |
-
"grad_norm": 8.61099910736084,
|
| 351 |
-
"learning_rate": 4.9835672628832366e-05,
|
| 352 |
-
"loss": 0.16875293254852294,
|
| 353 |
-
"step": 960
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.1310193270219006,
|
| 357 |
-
"grad_norm": 0.5758953094482422,
|
| 358 |
-
"learning_rate": 4.9828733029106434e-05,
|
| 359 |
-
"loss": 0.24272210597991944,
|
| 360 |
-
"step": 980
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.1336931908386741,
|
| 364 |
-
"grad_norm": 6.830381870269775,
|
| 365 |
-
"learning_rate": 4.982165041661575e-05,
|
| 366 |
-
"loss": 0.1900513768196106,
|
| 367 |
-
"step": 1000
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.13636705465544757,
|
| 371 |
-
"grad_norm": 3.979989528656006,
|
| 372 |
-
"learning_rate": 4.981442483215595e-05,
|
| 373 |
-
"loss": 0.23804061412811278,
|
| 374 |
-
"step": 1020
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.13904091847222105,
|
| 378 |
-
"grad_norm": 3.480762004852295,
|
| 379 |
-
"learning_rate": 4.98070563173462e-05,
|
| 380 |
-
"loss": 0.19694632291793823,
|
| 381 |
-
"step": 1040
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.14171478228899453,
|
| 385 |
-
"grad_norm": 2.7838668823242188,
|
| 386 |
-
"learning_rate": 4.979954491462892e-05,
|
| 387 |
-
"loss": 0.20703573226928712,
|
| 388 |
-
"step": 1060
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.14438864610576801,
|
| 392 |
-
"grad_norm": 4.724070072174072,
|
| 393 |
-
"learning_rate": 4.979189066726955e-05,
|
| 394 |
-
"loss": 0.135595703125,
|
| 395 |
-
"step": 1080
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.1470625099225415,
|
| 399 |
-
"grad_norm": 6.755579471588135,
|
| 400 |
-
"learning_rate": 4.978409361935636e-05,
|
| 401 |
-
"loss": 0.20947463512420655,
|
| 402 |
-
"step": 1100
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.14973637373931498,
|
| 406 |
-
"grad_norm": 2.989082098007202,
|
| 407 |
-
"learning_rate": 4.9776153815800075e-05,
|
| 408 |
-
"loss": 0.15855485200881958,
|
| 409 |
-
"step": 1120
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.15241023755608848,
|
| 413 |
-
"grad_norm": 2.310991048812866,
|
| 414 |
-
"learning_rate": 4.976807130233375e-05,
|
| 415 |
-
"loss": 0.1401826858520508,
|
| 416 |
-
"step": 1140
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.15508410137286197,
|
| 420 |
-
"grad_norm": 3.802248477935791,
|
| 421 |
-
"learning_rate": 4.975984612551243e-05,
|
| 422 |
-
"loss": 0.17620396614074707,
|
| 423 |
-
"step": 1160
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.15775796518963545,
|
| 427 |
-
"grad_norm": 2.0629191398620605,
|
| 428 |
-
"learning_rate": 4.975147833271288e-05,
|
| 429 |
-
"loss": 0.21862537860870362,
|
| 430 |
-
"step": 1180
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.16043182900640893,
|
| 434 |
-
"grad_norm": 2.2114040851593018,
|
| 435 |
-
"learning_rate": 4.9742967972133335e-05,
|
| 436 |
-
"loss": 0.15507786273956298,
|
| 437 |
-
"step": 1200
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.1631056928231824,
|
| 441 |
-
"grad_norm": 10.362643241882324,
|
| 442 |
-
"learning_rate": 4.973431509279323e-05,
|
| 443 |
-
"loss": 0.15983071327209472,
|
| 444 |
-
"step": 1220
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.1657795566399559,
|
| 448 |
-
"grad_norm": 0.3497254252433777,
|
| 449 |
-
"learning_rate": 4.972551974453287e-05,
|
| 450 |
-
"loss": 0.20826480388641358,
|
| 451 |
-
"step": 1240
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.16845342045672937,
|
| 455 |
-
"grad_norm": 0.3897332549095154,
|
| 456 |
-
"learning_rate": 4.971658197801322e-05,
|
| 457 |
-
"loss": 0.12974958419799804,
|
| 458 |
-
"step": 1260
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.17112728427350285,
|
| 462 |
-
"grad_norm": 6.043999195098877,
|
| 463 |
-
"learning_rate": 4.9707501844715554e-05,
|
| 464 |
-
"loss": 0.1621025800704956,
|
| 465 |
-
"step": 1280
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.17380114809027633,
|
| 469 |
-
"grad_norm": 1.7294467687606812,
|
| 470 |
-
"learning_rate": 4.969827939694115e-05,
|
| 471 |
-
"loss": 0.17147176265716552,
|
| 472 |
-
"step": 1300
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.1764750119070498,
|
| 476 |
-
"grad_norm": 0.8541621565818787,
|
| 477 |
-
"learning_rate": 4.968891468781105e-05,
|
| 478 |
-
"loss": 0.1330248236656189,
|
| 479 |
-
"step": 1320
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.1791488757238233,
|
| 483 |
-
"grad_norm": 2.197866439819336,
|
| 484 |
-
"learning_rate": 4.967940777126569e-05,
|
| 485 |
-
"loss": 0.12178658246994019,
|
| 486 |
-
"step": 1340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.18182273954059677,
|
| 490 |
-
"grad_norm": 0.8711172342300415,
|
| 491 |
-
"learning_rate": 4.9669758702064636e-05,
|
| 492 |
-
"loss": 0.19501737356185914,
|
| 493 |
-
"step": 1360
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.18449660335737025,
|
| 497 |
-
"grad_norm": 5.356688022613525,
|
| 498 |
-
"learning_rate": 4.965996753578623e-05,
|
| 499 |
-
"loss": 0.1602308750152588,
|
| 500 |
-
"step": 1380
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.18717046717414373,
|
| 504 |
-
"grad_norm": 13.699594497680664,
|
| 505 |
-
"learning_rate": 4.9650034328827305e-05,
|
| 506 |
-
"loss": 0.12051005363464355,
|
| 507 |
-
"step": 1400
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.18984433099091721,
|
| 511 |
-
"grad_norm": 1.738416314125061,
|
| 512 |
-
"learning_rate": 4.963995913840284e-05,
|
| 513 |
-
"loss": 0.1177564263343811,
|
| 514 |
-
"step": 1420
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.1925181948076907,
|
| 518 |
-
"grad_norm": 5.713410377502441,
|
| 519 |
-
"learning_rate": 4.9629742022545623e-05,
|
| 520 |
-
"loss": 0.19124844074249267,
|
| 521 |
-
"step": 1440
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.19519205862446418,
|
| 525 |
-
"grad_norm": 4.171947479248047,
|
| 526 |
-
"learning_rate": 4.961938304010595e-05,
|
| 527 |
-
"loss": 0.12269804477691651,
|
| 528 |
-
"step": 1460
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.19786592244123766,
|
| 532 |
-
"grad_norm": 2.8535916805267334,
|
| 533 |
-
"learning_rate": 4.9608882250751245e-05,
|
| 534 |
-
"loss": 0.0892150342464447,
|
| 535 |
-
"step": 1480
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.20053978625801114,
|
| 539 |
-
"grad_norm": 0.6551859378814697,
|
| 540 |
-
"learning_rate": 4.959823971496574e-05,
|
| 541 |
-
"loss": 0.10962753295898438,
|
| 542 |
-
"step": 1500
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.20321365007478462,
|
| 546 |
-
"grad_norm": 6.525365352630615,
|
| 547 |
-
"learning_rate": 4.9587455494050136e-05,
|
| 548 |
-
"loss": 0.10057367086410522,
|
| 549 |
-
"step": 1520
|
| 550 |
-
}
|
| 551 |
-
],
|
| 552 |
-
"logging_steps": 20,
|
| 553 |
-
"max_steps": 26180,
|
| 554 |
-
"num_input_tokens_seen": 0,
|
| 555 |
-
"num_train_epochs": 4,
|
| 556 |
-
"save_steps": 1000000000,
|
| 557 |
-
"stateful_callbacks": {
|
| 558 |
-
"TrainerControl": {
|
| 559 |
-
"args": {
|
| 560 |
-
"should_epoch_stop": false,
|
| 561 |
-
"should_evaluate": false,
|
| 562 |
-
"should_log": false,
|
| 563 |
-
"should_save": true,
|
| 564 |
-
"should_training_stop": false
|
| 565 |
-
},
|
| 566 |
-
"attributes": {}
|
| 567 |
-
}
|
| 568 |
-
},
|
| 569 |
-
"total_flos": 5902542074437632.0,
|
| 570 |
-
"train_batch_size": 1,
|
| 571 |
-
"trial_name": null,
|
| 572 |
-
"trial_params": null
|
| 573 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/8/checkpoint-2230/config.json
DELETED
|
@@ -1,24 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"architectures": [
|
| 3 |
-
"TwinyForCausalLM"
|
| 4 |
-
],
|
| 5 |
-
"attention_dropout": 0.0,
|
| 6 |
-
"dtype": "float32",
|
| 7 |
-
"hidden_dropout": 0.0,
|
| 8 |
-
"hidden_size": 768,
|
| 9 |
-
"initializer_range": 0.02,
|
| 10 |
-
"intermediate_size": 3072,
|
| 11 |
-
"max_position_embeddings": 128,
|
| 12 |
-
"model_type": "twiny",
|
| 13 |
-
"neftune_alpha": 0.0,
|
| 14 |
-
"num_attention_heads": 12,
|
| 15 |
-
"num_hidden_layers": 3,
|
| 16 |
-
"num_key_value_heads": 3,
|
| 17 |
-
"qk_norm": true,
|
| 18 |
-
"rezero_init": 1.0,
|
| 19 |
-
"rms_norm_eps": 1e-06,
|
| 20 |
-
"rope_theta": 10000.0,
|
| 21 |
-
"transformers_version": "5.0.0",
|
| 22 |
-
"use_cache": false,
|
| 23 |
-
"vocab_size": 32000
|
| 24 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
slots/8/checkpoint-2230/trainer_state.json
DELETED
|
@@ -1,818 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.34782495670060215,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 2230,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.0001559753169060996,
|
| 14 |
-
"grad_norm": Infinity,
|
| 15 |
-
"learning_rate": 5e-05,
|
| 16 |
-
"loss": 143.68887329101562,
|
| 17 |
-
"step": 1
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.0031195063381219924,
|
| 21 |
-
"grad_norm": 52.297000885009766,
|
| 22 |
-
"learning_rate": 4.999993729149804e-05,
|
| 23 |
-
"loss": 63.87562962582237,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.006239012676243985,
|
| 28 |
-
"grad_norm": 18.09991455078125,
|
| 29 |
-
"learning_rate": 4.999968253874798e-05,
|
| 30 |
-
"loss": 10.082135009765626,
|
| 31 |
-
"step": 40
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.009358519014365976,
|
| 35 |
-
"grad_norm": 15.637947082519531,
|
| 36 |
-
"learning_rate": 4.99992318244638e-05,
|
| 37 |
-
"loss": 5.112002944946289,
|
| 38 |
-
"step": 60
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.01247802535248797,
|
| 42 |
-
"grad_norm": 12.031600952148438,
|
| 43 |
-
"learning_rate": 4.9998585152178466e-05,
|
| 44 |
-
"loss": 2.7785192489624024,
|
| 45 |
-
"step": 80
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.015597531690609961,
|
| 49 |
-
"grad_norm": 6.920973300933838,
|
| 50 |
-
"learning_rate": 4.999774252696095e-05,
|
| 51 |
-
"loss": 2.0463287353515627,
|
| 52 |
-
"step": 100
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.018717038028731953,
|
| 56 |
-
"grad_norm": 7.486307144165039,
|
| 57 |
-
"learning_rate": 4.999670395541621e-05,
|
| 58 |
-
"loss": 1.7835391998291015,
|
| 59 |
-
"step": 120
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.021836544366853946,
|
| 63 |
-
"grad_norm": 8.828413963317871,
|
| 64 |
-
"learning_rate": 4.999546944568517e-05,
|
| 65 |
-
"loss": 1.2352788925170899,
|
| 66 |
-
"step": 140
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.02495605070497594,
|
| 70 |
-
"grad_norm": 9.396143913269043,
|
| 71 |
-
"learning_rate": 4.9994039007444616e-05,
|
| 72 |
-
"loss": 1.0546810150146484,
|
| 73 |
-
"step": 160
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.02807555704309793,
|
| 77 |
-
"grad_norm": 7.957871913909912,
|
| 78 |
-
"learning_rate": 4.9992412651907096e-05,
|
| 79 |
-
"loss": 0.9882014274597168,
|
| 80 |
-
"step": 180
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.031195063381219922,
|
| 84 |
-
"grad_norm": 4.318814277648926,
|
| 85 |
-
"learning_rate": 4.999059039182092e-05,
|
| 86 |
-
"loss": 0.7930902957916259,
|
| 87 |
-
"step": 200
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.034314569719341916,
|
| 91 |
-
"grad_norm": 5.154116630554199,
|
| 92 |
-
"learning_rate": 4.998857224146998e-05,
|
| 93 |
-
"loss": 0.7149968624114991,
|
| 94 |
-
"step": 220
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.037434076057463905,
|
| 98 |
-
"grad_norm": 11.299652099609375,
|
| 99 |
-
"learning_rate": 4.998635821667367e-05,
|
| 100 |
-
"loss": 0.5597743034362793,
|
| 101 |
-
"step": 240
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.040553582395585895,
|
| 105 |
-
"grad_norm": 4.086544036865234,
|
| 106 |
-
"learning_rate": 4.9983948334786756e-05,
|
| 107 |
-
"loss": 0.5503968715667724,
|
| 108 |
-
"step": 260
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.04367308873370789,
|
| 112 |
-
"grad_norm": 4.867542743682861,
|
| 113 |
-
"learning_rate": 4.998134261469924e-05,
|
| 114 |
-
"loss": 0.5038466930389405,
|
| 115 |
-
"step": 280
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.04679259507182988,
|
| 119 |
-
"grad_norm": 3.7527730464935303,
|
| 120 |
-
"learning_rate": 4.997854107683623e-05,
|
| 121 |
-
"loss": 0.42429962158203127,
|
| 122 |
-
"step": 300
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.04991210140995188,
|
| 126 |
-
"grad_norm": 3.1628715991973877,
|
| 127 |
-
"learning_rate": 4.997554374315774e-05,
|
| 128 |
-
"loss": 0.43418636322021487,
|
| 129 |
-
"step": 320
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.05303160774807387,
|
| 133 |
-
"grad_norm": 6.594268798828125,
|
| 134 |
-
"learning_rate": 4.997235063715855e-05,
|
| 135 |
-
"loss": 0.3593674898147583,
|
| 136 |
-
"step": 340
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.05615111408619586,
|
| 140 |
-
"grad_norm": 3.662322521209717,
|
| 141 |
-
"learning_rate": 4.996896178386803e-05,
|
| 142 |
-
"loss": 0.3210789680480957,
|
| 143 |
-
"step": 360
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.059270620424317855,
|
| 147 |
-
"grad_norm": 3.9782955646514893,
|
| 148 |
-
"learning_rate": 4.9965377209849904e-05,
|
| 149 |
-
"loss": 0.30758669376373293,
|
| 150 |
-
"step": 380
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.062390126762439845,
|
| 154 |
-
"grad_norm": 5.329382419586182,
|
| 155 |
-
"learning_rate": 4.996159694320207e-05,
|
| 156 |
-
"loss": 0.2638844966888428,
|
| 157 |
-
"step": 400
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.06550963310056183,
|
| 161 |
-
"grad_norm": 3.7222423553466797,
|
| 162 |
-
"learning_rate": 4.995762101355639e-05,
|
| 163 |
-
"loss": 0.21310510635375976,
|
| 164 |
-
"step": 420
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.06862913943868383,
|
| 168 |
-
"grad_norm": 3.9508121013641357,
|
| 169 |
-
"learning_rate": 4.9953449452078414e-05,
|
| 170 |
-
"loss": 0.21608681678771974,
|
| 171 |
-
"step": 440
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.07174864577680581,
|
| 175 |
-
"grad_norm": 2.9517951011657715,
|
| 176 |
-
"learning_rate": 4.99490822914672e-05,
|
| 177 |
-
"loss": 0.20024800300598145,
|
| 178 |
-
"step": 460
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.07486815211492781,
|
| 182 |
-
"grad_norm": 1.2055240869522095,
|
| 183 |
-
"learning_rate": 4.994451956595499e-05,
|
| 184 |
-
"loss": 0.1932399868965149,
|
| 185 |
-
"step": 480
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.07798765845304981,
|
| 189 |
-
"grad_norm": 1.6800686120986938,
|
| 190 |
-
"learning_rate": 4.993976131130699e-05,
|
| 191 |
-
"loss": 0.1796867847442627,
|
| 192 |
-
"step": 500
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.08110716479117179,
|
| 196 |
-
"grad_norm": 3.093266725540161,
|
| 197 |
-
"learning_rate": 4.9934807564821065e-05,
|
| 198 |
-
"loss": 0.14106053113937378,
|
| 199 |
-
"step": 520
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.08422667112929379,
|
| 203 |
-
"grad_norm": 1.3461378812789917,
|
| 204 |
-
"learning_rate": 4.992965836532746e-05,
|
| 205 |
-
"loss": 0.15992594957351686,
|
| 206 |
-
"step": 540
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.08734617746741578,
|
| 210 |
-
"grad_norm": 1.4411957263946533,
|
| 211 |
-
"learning_rate": 4.992431375318849e-05,
|
| 212 |
-
"loss": 0.13545843362808227,
|
| 213 |
-
"step": 560
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.09046568380553778,
|
| 217 |
-
"grad_norm": 2.334609031677246,
|
| 218 |
-
"learning_rate": 4.991877377029823e-05,
|
| 219 |
-
"loss": 0.12385048866271972,
|
| 220 |
-
"step": 580
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.09358519014365976,
|
| 224 |
-
"grad_norm": 1.9647951126098633,
|
| 225 |
-
"learning_rate": 4.991303846008216e-05,
|
| 226 |
-
"loss": 0.15953426361083983,
|
| 227 |
-
"step": 600
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.09670469648178176,
|
| 231 |
-
"grad_norm": 0.9261314868927002,
|
| 232 |
-
"learning_rate": 4.990710786749686e-05,
|
| 233 |
-
"loss": 0.13601958751678467,
|
| 234 |
-
"step": 620
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.09982420281990376,
|
| 238 |
-
"grad_norm": 2.8501081466674805,
|
| 239 |
-
"learning_rate": 4.990098203902965e-05,
|
| 240 |
-
"loss": 0.12980480194091798,
|
| 241 |
-
"step": 640
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.10294370915802574,
|
| 245 |
-
"grad_norm": 2.030174493789673,
|
| 246 |
-
"learning_rate": 4.989466102269821e-05,
|
| 247 |
-
"loss": 0.11676372289657592,
|
| 248 |
-
"step": 660
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.10606321549614774,
|
| 252 |
-
"grad_norm": 2.160588502883911,
|
| 253 |
-
"learning_rate": 4.9888144868050194e-05,
|
| 254 |
-
"loss": 0.11160085201263428,
|
| 255 |
-
"step": 680
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.10918272183426973,
|
| 259 |
-
"grad_norm": 1.138016939163208,
|
| 260 |
-
"learning_rate": 4.9881433626162896e-05,
|
| 261 |
-
"loss": 0.09880762696266174,
|
| 262 |
-
"step": 700
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.11230222817239172,
|
| 266 |
-
"grad_norm": 2.918583869934082,
|
| 267 |
-
"learning_rate": 4.98745273496428e-05,
|
| 268 |
-
"loss": 0.08724074363708496,
|
| 269 |
-
"step": 720
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.11542173451051371,
|
| 273 |
-
"grad_norm": 2.165208578109741,
|
| 274 |
-
"learning_rate": 4.986742609262515e-05,
|
| 275 |
-
"loss": 0.11841287612915039,
|
| 276 |
-
"step": 740
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.11854124084863571,
|
| 280 |
-
"grad_norm": 2.163217306137085,
|
| 281 |
-
"learning_rate": 4.986012991077359e-05,
|
| 282 |
-
"loss": 0.10830552577972412,
|
| 283 |
-
"step": 760
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.12166074718675769,
|
| 287 |
-
"grad_norm": 1.270262598991394,
|
| 288 |
-
"learning_rate": 4.98526388612797e-05,
|
| 289 |
-
"loss": 0.06507933735847474,
|
| 290 |
-
"step": 780
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.12478025352487969,
|
| 294 |
-
"grad_norm": 1.2272371053695679,
|
| 295 |
-
"learning_rate": 4.984495300286252e-05,
|
| 296 |
-
"loss": 0.07868309020996093,
|
| 297 |
-
"step": 800
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.12789975986300167,
|
| 301 |
-
"grad_norm": 1.2555146217346191,
|
| 302 |
-
"learning_rate": 4.98370723957681e-05,
|
| 303 |
-
"loss": 0.06831848621368408,
|
| 304 |
-
"step": 820
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.13101926620112367,
|
| 308 |
-
"grad_norm": 0.2989479601383209,
|
| 309 |
-
"learning_rate": 4.982899710176909e-05,
|
| 310 |
-
"loss": 0.10411404371261597,
|
| 311 |
-
"step": 840
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.13413877253924567,
|
| 315 |
-
"grad_norm": 1.6604547500610352,
|
| 316 |
-
"learning_rate": 4.982072718416416e-05,
|
| 317 |
-
"loss": 0.07505239248275757,
|
| 318 |
-
"step": 860
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.13725827887736766,
|
| 322 |
-
"grad_norm": 6.027090549468994,
|
| 323 |
-
"learning_rate": 4.981226270777756e-05,
|
| 324 |
-
"loss": 0.08949573040008545,
|
| 325 |
-
"step": 880
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.14037778521548966,
|
| 329 |
-
"grad_norm": 5.2599077224731445,
|
| 330 |
-
"learning_rate": 4.9803603738958604e-05,
|
| 331 |
-
"loss": 0.04615107178688049,
|
| 332 |
-
"step": 900
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.14349729155361163,
|
| 336 |
-
"grad_norm": 2.0441677570343018,
|
| 337 |
-
"learning_rate": 4.979475034558115e-05,
|
| 338 |
-
"loss": 0.08460248112678528,
|
| 339 |
-
"step": 920
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.14661679789173362,
|
| 343 |
-
"grad_norm": 1.892174482345581,
|
| 344 |
-
"learning_rate": 4.9785702597043064e-05,
|
| 345 |
-
"loss": 0.06970399022102355,
|
| 346 |
-
"step": 940
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.14973630422985562,
|
| 350 |
-
"grad_norm": 1.8726390600204468,
|
| 351 |
-
"learning_rate": 4.977646056426567e-05,
|
| 352 |
-
"loss": 0.07341498732566834,
|
| 353 |
-
"step": 960
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.15285581056797762,
|
| 357 |
-
"grad_norm": 3.0138962268829346,
|
| 358 |
-
"learning_rate": 4.97670243196932e-05,
|
| 359 |
-
"loss": 0.06328916549682617,
|
| 360 |
-
"step": 980
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.15597531690609961,
|
| 364 |
-
"grad_norm": 1.26374089717865,
|
| 365 |
-
"learning_rate": 4.975739393729224e-05,
|
| 366 |
-
"loss": 0.06133637428283691,
|
| 367 |
-
"step": 1000
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 0.1590948232442216,
|
| 371 |
-
"grad_norm": 2.468918561935425,
|
| 372 |
-
"learning_rate": 4.974756949255112e-05,
|
| 373 |
-
"loss": 0.040934398770332336,
|
| 374 |
-
"step": 1020
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 0.16221432958234358,
|
| 378 |
-
"grad_norm": 1.0652263164520264,
|
| 379 |
-
"learning_rate": 4.973755106247936e-05,
|
| 380 |
-
"loss": 0.053039383888244626,
|
| 381 |
-
"step": 1040
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 0.16533383592046558,
|
| 385 |
-
"grad_norm": 8.056346893310547,
|
| 386 |
-
"learning_rate": 4.972733872560702e-05,
|
| 387 |
-
"loss": 0.05064538717269897,
|
| 388 |
-
"step": 1060
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 0.16845334225858757,
|
| 392 |
-
"grad_norm": 0.27906128764152527,
|
| 393 |
-
"learning_rate": 4.971693256198414e-05,
|
| 394 |
-
"loss": 0.038862642645835874,
|
| 395 |
-
"step": 1080
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 0.17157284859670957,
|
| 399 |
-
"grad_norm": 0.9866332411766052,
|
| 400 |
-
"learning_rate": 4.9706332653180055e-05,
|
| 401 |
-
"loss": 0.05685908198356628,
|
| 402 |
-
"step": 1100
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 0.17469235493483157,
|
| 406 |
-
"grad_norm": 0.1504131406545639,
|
| 407 |
-
"learning_rate": 4.9695539082282805e-05,
|
| 408 |
-
"loss": 0.054278552532196045,
|
| 409 |
-
"step": 1120
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 0.17781186127295356,
|
| 413 |
-
"grad_norm": 0.6855164170265198,
|
| 414 |
-
"learning_rate": 4.968455193389845e-05,
|
| 415 |
-
"loss": 0.038184747099876404,
|
| 416 |
-
"step": 1140
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 0.18093136761107556,
|
| 420 |
-
"grad_norm": 1.5174620151519775,
|
| 421 |
-
"learning_rate": 4.967337129415041e-05,
|
| 422 |
-
"loss": 0.032207447290420535,
|
| 423 |
-
"step": 1160
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 0.18405087394919753,
|
| 427 |
-
"grad_norm": 2.1656675338745117,
|
| 428 |
-
"learning_rate": 4.966199725067883e-05,
|
| 429 |
-
"loss": 0.04032683968544006,
|
| 430 |
-
"step": 1180
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 0.18717038028731953,
|
| 434 |
-
"grad_norm": 0.16235563158988953,
|
| 435 |
-
"learning_rate": 4.9650429892639824e-05,
|
| 436 |
-
"loss": 0.03260311484336853,
|
| 437 |
-
"step": 1200
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 0.19028988662544152,
|
| 441 |
-
"grad_norm": 0.7270789742469788,
|
| 442 |
-
"learning_rate": 4.9638669310704856e-05,
|
| 443 |
-
"loss": 0.039445298910140994,
|
| 444 |
-
"step": 1220
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 0.19340939296356352,
|
| 448 |
-
"grad_norm": 1.4053725004196167,
|
| 449 |
-
"learning_rate": 4.9626715597059936e-05,
|
| 450 |
-
"loss": 0.05901872515678406,
|
| 451 |
-
"step": 1240
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 0.19652889930168552,
|
| 455 |
-
"grad_norm": 0.149334117770195,
|
| 456 |
-
"learning_rate": 4.9614568845404994e-05,
|
| 457 |
-
"loss": 0.05250124335289001,
|
| 458 |
-
"step": 1260
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 0.1996484056398075,
|
| 462 |
-
"grad_norm": 1.7250016927719116,
|
| 463 |
-
"learning_rate": 4.9602229150953086e-05,
|
| 464 |
-
"loss": 0.03500291109085083,
|
| 465 |
-
"step": 1280
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 0.20276791197792948,
|
| 469 |
-
"grad_norm": 0.3273993730545044,
|
| 470 |
-
"learning_rate": 4.9589696610429635e-05,
|
| 471 |
-
"loss": 0.04656350910663605,
|
| 472 |
-
"step": 1300
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 0.20588741831605148,
|
| 476 |
-
"grad_norm": 2.355738878250122,
|
| 477 |
-
"learning_rate": 4.957697132207175e-05,
|
| 478 |
-
"loss": 0.0341707855463028,
|
| 479 |
-
"step": 1320
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 0.20900692465417348,
|
| 483 |
-
"grad_norm": 0.39558863639831543,
|
| 484 |
-
"learning_rate": 4.9564053385627375e-05,
|
| 485 |
-
"loss": 0.04657787382602692,
|
| 486 |
-
"step": 1340
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 0.21212643099229547,
|
| 490 |
-
"grad_norm": 0.7477391362190247,
|
| 491 |
-
"learning_rate": 4.955094290235452e-05,
|
| 492 |
-
"loss": 0.032107046246528624,
|
| 493 |
-
"step": 1360
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 0.21524593733041747,
|
| 497 |
-
"grad_norm": 8.392775535583496,
|
| 498 |
-
"learning_rate": 4.9537639975020525e-05,
|
| 499 |
-
"loss": 0.034580078721046445,
|
| 500 |
-
"step": 1380
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 0.21836544366853947,
|
| 504 |
-
"grad_norm": 1.3500334024429321,
|
| 505 |
-
"learning_rate": 4.95241447079012e-05,
|
| 506 |
-
"loss": 0.031853339076042174,
|
| 507 |
-
"step": 1400
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 0.22148495000666144,
|
| 511 |
-
"grad_norm": 0.7828538417816162,
|
| 512 |
-
"learning_rate": 4.9510457206779996e-05,
|
| 513 |
-
"loss": 0.041383099555969236,
|
| 514 |
-
"step": 1420
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 0.22460445634478343,
|
| 518 |
-
"grad_norm": 0.8445055484771729,
|
| 519 |
-
"learning_rate": 4.949657757894724e-05,
|
| 520 |
-
"loss": 0.037648487091064456,
|
| 521 |
-
"step": 1440
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 0.22772396268290543,
|
| 525 |
-
"grad_norm": 0.20844776928424835,
|
| 526 |
-
"learning_rate": 4.948250593319925e-05,
|
| 527 |
-
"loss": 0.03933250904083252,
|
| 528 |
-
"step": 1460
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 0.23084346902102743,
|
| 532 |
-
"grad_norm": 2.855363368988037,
|
| 533 |
-
"learning_rate": 4.946824237983747e-05,
|
| 534 |
-
"loss": 0.03083827197551727,
|
| 535 |
-
"step": 1480
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 0.23396297535914942,
|
| 539 |
-
"grad_norm": 0.9697449803352356,
|
| 540 |
-
"learning_rate": 4.9453787030667616e-05,
|
| 541 |
-
"loss": 0.027672284841537477,
|
| 542 |
-
"step": 1500
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 0.23708248169727142,
|
| 546 |
-
"grad_norm": 0.08931494504213333,
|
| 547 |
-
"learning_rate": 4.943913999899884e-05,
|
| 548 |
-
"loss": 0.016587555408477783,
|
| 549 |
-
"step": 1520
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 0.2402019880353934,
|
| 553 |
-
"grad_norm": 0.7060425281524658,
|
| 554 |
-
"learning_rate": 4.942430139964278e-05,
|
| 555 |
-
"loss": 0.02475900799036026,
|
| 556 |
-
"step": 1540
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 0.24332149437351538,
|
| 560 |
-
"grad_norm": 0.17042799293994904,
|
| 561 |
-
"learning_rate": 4.9409271348912714e-05,
|
| 562 |
-
"loss": 0.023083586990833283,
|
| 563 |
-
"step": 1560
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 0.24644100071163738,
|
| 567 |
-
"grad_norm": 0.21033896505832672,
|
| 568 |
-
"learning_rate": 4.9394049964622604e-05,
|
| 569 |
-
"loss": 0.023305554687976838,
|
| 570 |
-
"step": 1580
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 0.24956050704975938,
|
| 574 |
-
"grad_norm": 1.6797473430633545,
|
| 575 |
-
"learning_rate": 4.93786373660862e-05,
|
| 576 |
-
"loss": 0.027982226014137267,
|
| 577 |
-
"step": 1600
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 0.2526800133878814,
|
| 581 |
-
"grad_norm": 0.19112282991409302,
|
| 582 |
-
"learning_rate": 4.936303367411609e-05,
|
| 583 |
-
"loss": 0.041162309050559995,
|
| 584 |
-
"step": 1620
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 0.25579951972600334,
|
| 588 |
-
"grad_norm": 1.997085452079773,
|
| 589 |
-
"learning_rate": 4.934723901102276e-05,
|
| 590 |
-
"loss": 0.030818694829940797,
|
| 591 |
-
"step": 1640
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 0.25891902606412537,
|
| 595 |
-
"grad_norm": 0.10342924296855927,
|
| 596 |
-
"learning_rate": 4.933125350061365e-05,
|
| 597 |
-
"loss": 0.011925507336854935,
|
| 598 |
-
"step": 1660
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 0.26203853240224734,
|
| 602 |
-
"grad_norm": 0.10754784196615219,
|
| 603 |
-
"learning_rate": 4.931507726819215e-05,
|
| 604 |
-
"loss": 0.038025134801864625,
|
| 605 |
-
"step": 1680
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 0.26515803874036936,
|
| 609 |
-
"grad_norm": 0.7322559356689453,
|
| 610 |
-
"learning_rate": 4.929871044055664e-05,
|
| 611 |
-
"loss": 0.01450435072183609,
|
| 612 |
-
"step": 1700
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 0.26827754507849133,
|
| 616 |
-
"grad_norm": 0.11053226888179779,
|
| 617 |
-
"learning_rate": 4.92821531459995e-05,
|
| 618 |
-
"loss": 0.03104289174079895,
|
| 619 |
-
"step": 1720
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 0.2713970514166133,
|
| 623 |
-
"grad_norm": 0.2635640799999237,
|
| 624 |
-
"learning_rate": 4.92654055143061e-05,
|
| 625 |
-
"loss": 0.020085930824279785,
|
| 626 |
-
"step": 1740
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 0.2745165577547353,
|
| 630 |
-
"grad_norm": 1.1885840892791748,
|
| 631 |
-
"learning_rate": 4.924846767675375e-05,
|
| 632 |
-
"loss": 0.03624064326286316,
|
| 633 |
-
"step": 1760
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 0.2776360640928573,
|
| 637 |
-
"grad_norm": 0.12489036470651627,
|
| 638 |
-
"learning_rate": 4.923133976611074e-05,
|
| 639 |
-
"loss": 0.037028759717941284,
|
| 640 |
-
"step": 1780
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.2807555704309793,
|
| 644 |
-
"grad_norm": 0.13448038697242737,
|
| 645 |
-
"learning_rate": 4.9214021916635234e-05,
|
| 646 |
-
"loss": 0.016252121329307555,
|
| 647 |
-
"step": 1800
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.2838750767691013,
|
| 651 |
-
"grad_norm": 3.755352020263672,
|
| 652 |
-
"learning_rate": 4.9196514264074245e-05,
|
| 653 |
-
"loss": 0.027153733372688293,
|
| 654 |
-
"step": 1820
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.28699458310722326,
|
| 658 |
-
"grad_norm": 0.13555799424648285,
|
| 659 |
-
"learning_rate": 4.917881694566259e-05,
|
| 660 |
-
"loss": 0.029864606261253358,
|
| 661 |
-
"step": 1840
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.2901140894453453,
|
| 665 |
-
"grad_norm": 0.13309070467948914,
|
| 666 |
-
"learning_rate": 4.916093010012175e-05,
|
| 667 |
-
"loss": 0.013042651116847992,
|
| 668 |
-
"step": 1860
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.29323359578346725,
|
| 672 |
-
"grad_norm": 2.035360097885132,
|
| 673 |
-
"learning_rate": 4.9142853867658875e-05,
|
| 674 |
-
"loss": 0.015907545387744904,
|
| 675 |
-
"step": 1880
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.2963531021215893,
|
| 679 |
-
"grad_norm": 0.9238037467002869,
|
| 680 |
-
"learning_rate": 4.912458838996561e-05,
|
| 681 |
-
"loss": 0.023724390566349028,
|
| 682 |
-
"step": 1900
|
| 683 |
-
},
|
| 684 |
-
{
|
| 685 |
-
"epoch": 0.29947260845971124,
|
| 686 |
-
"grad_norm": 0.22935695946216583,
|
| 687 |
-
"learning_rate": 4.910613381021702e-05,
|
| 688 |
-
"loss": 0.014851535856723785,
|
| 689 |
-
"step": 1920
|
| 690 |
-
},
|
| 691 |
-
{
|
| 692 |
-
"epoch": 0.30259211479783327,
|
| 693 |
-
"grad_norm": 0.3207225799560547,
|
| 694 |
-
"learning_rate": 4.908749027307043e-05,
|
| 695 |
-
"loss": 0.013988447189331055,
|
| 696 |
-
"step": 1940
|
| 697 |
-
},
|
| 698 |
-
{
|
| 699 |
-
"epoch": 0.30571162113595524,
|
| 700 |
-
"grad_norm": 0.19095349311828613,
|
| 701 |
-
"learning_rate": 4.9068657924664366e-05,
|
| 702 |
-
"loss": 0.03171408772468567,
|
| 703 |
-
"step": 1960
|
| 704 |
-
},
|
| 705 |
-
{
|
| 706 |
-
"epoch": 0.3088311274740772,
|
| 707 |
-
"grad_norm": 3.989168643951416,
|
| 708 |
-
"learning_rate": 4.9049636912617316e-05,
|
| 709 |
-
"loss": 0.034608349204063416,
|
| 710 |
-
"step": 1980
|
| 711 |
-
},
|
| 712 |
-
{
|
| 713 |
-
"epoch": 0.31195063381219923,
|
| 714 |
-
"grad_norm": 0.1929292231798172,
|
| 715 |
-
"learning_rate": 4.9030427386026664e-05,
|
| 716 |
-
"loss": 0.02513568103313446,
|
| 717 |
-
"step": 2000
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 0.3150701401503212,
|
| 721 |
-
"grad_norm": 1.7010163068771362,
|
| 722 |
-
"learning_rate": 4.9011029495467445e-05,
|
| 723 |
-
"loss": 0.017113180458545686,
|
| 724 |
-
"step": 2020
|
| 725 |
-
},
|
| 726 |
-
{
|
| 727 |
-
"epoch": 0.3181896464884432,
|
| 728 |
-
"grad_norm": 0.07719412446022034,
|
| 729 |
-
"learning_rate": 4.899144339299119e-05,
|
| 730 |
-
"loss": 0.021139779686927797,
|
| 731 |
-
"step": 2040
|
| 732 |
-
},
|
| 733 |
-
{
|
| 734 |
-
"epoch": 0.3213091528265652,
|
| 735 |
-
"grad_norm": 3.5760796070098877,
|
| 736 |
-
"learning_rate": 4.8971669232124784e-05,
|
| 737 |
-
"loss": 0.014710186421871186,
|
| 738 |
-
"step": 2060
|
| 739 |
-
},
|
| 740 |
-
{
|
| 741 |
-
"epoch": 0.32442865916468716,
|
| 742 |
-
"grad_norm": 1.4730585813522339,
|
| 743 |
-
"learning_rate": 4.8951707167869164e-05,
|
| 744 |
-
"loss": 0.017758625745773315,
|
| 745 |
-
"step": 2080
|
| 746 |
-
},
|
| 747 |
-
{
|
| 748 |
-
"epoch": 0.3275481655028092,
|
| 749 |
-
"grad_norm": 0.4192917048931122,
|
| 750 |
-
"learning_rate": 4.893155735669821e-05,
|
| 751 |
-
"loss": 0.02882348895072937,
|
| 752 |
-
"step": 2100
|
| 753 |
-
},
|
| 754 |
-
{
|
| 755 |
-
"epoch": 0.33066767184093115,
|
| 756 |
-
"grad_norm": 0.6030257344245911,
|
| 757 |
-
"learning_rate": 4.891121995655743e-05,
|
| 758 |
-
"loss": 0.010555204749107362,
|
| 759 |
-
"step": 2120
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 0.3337871781790532,
|
| 763 |
-
"grad_norm": 0.498300164937973,
|
| 764 |
-
"learning_rate": 4.8890695126862776e-05,
|
| 765 |
-
"loss": 0.02531702518463135,
|
| 766 |
-
"step": 2140
|
| 767 |
-
},
|
| 768 |
-
{
|
| 769 |
-
"epoch": 0.33690668451717515,
|
| 770 |
-
"grad_norm": 0.2421785444021225,
|
| 771 |
-
"learning_rate": 4.886998302849938e-05,
|
| 772 |
-
"loss": 0.012451396137475968,
|
| 773 |
-
"step": 2160
|
| 774 |
-
},
|
| 775 |
-
{
|
| 776 |
-
"epoch": 0.3400261908552972,
|
| 777 |
-
"grad_norm": 0.4294167160987854,
|
| 778 |
-
"learning_rate": 4.884908382382029e-05,
|
| 779 |
-
"loss": 0.013656674325466156,
|
| 780 |
-
"step": 2180
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 0.34314569719341914,
|
| 784 |
-
"grad_norm": 0.10389716178178787,
|
| 785 |
-
"learning_rate": 4.882799767664517e-05,
|
| 786 |
-
"loss": 0.008202654123306275,
|
| 787 |
-
"step": 2200
|
| 788 |
-
},
|
| 789 |
-
{
|
| 790 |
-
"epoch": 0.3462652035315411,
|
| 791 |
-
"grad_norm": 0.10655222833156586,
|
| 792 |
-
"learning_rate": 4.8806724752259095e-05,
|
| 793 |
-
"loss": 0.015621550381183624,
|
| 794 |
-
"step": 2220
|
| 795 |
-
}
|
| 796 |
-
],
|
| 797 |
-
"logging_steps": 20,
|
| 798 |
-
"max_steps": 22442,
|
| 799 |
-
"num_input_tokens_seen": 0,
|
| 800 |
-
"num_train_epochs": 4,
|
| 801 |
-
"save_steps": 1000000000,
|
| 802 |
-
"stateful_callbacks": {
|
| 803 |
-
"TrainerControl": {
|
| 804 |
-
"args": {
|
| 805 |
-
"should_epoch_stop": false,
|
| 806 |
-
"should_evaluate": false,
|
| 807 |
-
"should_log": false,
|
| 808 |
-
"should_save": true,
|
| 809 |
-
"should_training_stop": false
|
| 810 |
-
},
|
| 811 |
-
"attributes": {}
|
| 812 |
-
}
|
| 813 |
-
},
|
| 814 |
-
"total_flos": 8552741277450240.0,
|
| 815 |
-
"train_batch_size": 1,
|
| 816 |
-
"trial_name": null,
|
| 817 |
-
"trial_params": null
|
| 818 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|