Upload folder using huggingface_hub
Browse files
results_code/mistral/hmt_r8/train_metrics.json
ADDED
|
@@ -0,0 +1,429 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"model": "mistral",
|
| 3 |
+
"method": "hmt",
|
| 4 |
+
"rank": 8,
|
| 5 |
+
"dataset": "m-a-p/Code-Feedback",
|
| 6 |
+
"n_samples": 66383,
|
| 7 |
+
"epochs": 1,
|
| 8 |
+
"total_params": 7262703616,
|
| 9 |
+
"trainable_params": 20971520,
|
| 10 |
+
"trainable_pct": 0.2888,
|
| 11 |
+
"init_time_sec": 2.71,
|
| 12 |
+
"train_time_sec": 31995.0,
|
| 13 |
+
"peak_gpu_mem_gb": 18.9,
|
| 14 |
+
"final_loss": 0.5973,
|
| 15 |
+
"loss_curve": [
|
| 16 |
+
{
|
| 17 |
+
"step": 10,
|
| 18 |
+
"loss": 0.804
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"step": 20,
|
| 22 |
+
"loss": 0.7247
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"step": 30,
|
| 26 |
+
"loss": 0.7052
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"step": 40,
|
| 30 |
+
"loss": 0.6809
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"step": 50,
|
| 34 |
+
"loss": 0.6919
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"step": 60,
|
| 38 |
+
"loss": 0.6749
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"step": 70,
|
| 42 |
+
"loss": 0.6878
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"step": 80,
|
| 46 |
+
"loss": 0.6504
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"step": 90,
|
| 50 |
+
"loss": 0.6478
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"step": 100,
|
| 54 |
+
"loss": 0.6385
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"step": 110,
|
| 58 |
+
"loss": 0.6567
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"step": 120,
|
| 62 |
+
"loss": 0.6112
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"step": 130,
|
| 66 |
+
"loss": 0.6813
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"step": 140,
|
| 70 |
+
"loss": 0.6141
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"step": 150,
|
| 74 |
+
"loss": 0.6401
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"step": 160,
|
| 78 |
+
"loss": 0.6255
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"step": 170,
|
| 82 |
+
"loss": 0.6449
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"step": 180,
|
| 86 |
+
"loss": 0.6113
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"step": 190,
|
| 90 |
+
"loss": 0.6071
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"step": 200,
|
| 94 |
+
"loss": 0.607
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"step": 210,
|
| 98 |
+
"loss": 0.617
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"step": 220,
|
| 102 |
+
"loss": 0.6177
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"step": 230,
|
| 106 |
+
"loss": 0.6408
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"step": 240,
|
| 110 |
+
"loss": 0.5988
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"step": 250,
|
| 114 |
+
"loss": 0.618
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"step": 260,
|
| 118 |
+
"loss": 0.603
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"step": 270,
|
| 122 |
+
"loss": 0.6081
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"step": 280,
|
| 126 |
+
"loss": 0.6052
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"step": 290,
|
| 130 |
+
"loss": 0.5816
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"step": 300,
|
| 134 |
+
"loss": 0.6342
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"step": 310,
|
| 138 |
+
"loss": 0.6167
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"step": 320,
|
| 142 |
+
"loss": 0.624
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"step": 330,
|
| 146 |
+
"loss": 0.6123
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"step": 340,
|
| 150 |
+
"loss": 0.6009
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"step": 350,
|
| 154 |
+
"loss": 0.6214
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"step": 360,
|
| 158 |
+
"loss": 0.6045
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"step": 370,
|
| 162 |
+
"loss": 0.5993
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"step": 380,
|
| 166 |
+
"loss": 0.6155
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"step": 390,
|
| 170 |
+
"loss": 0.5883
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"step": 400,
|
| 174 |
+
"loss": 0.6287
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"step": 410,
|
| 178 |
+
"loss": 0.6063
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"step": 420,
|
| 182 |
+
"loss": 0.6057
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"step": 430,
|
| 186 |
+
"loss": 0.5961
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"step": 440,
|
| 190 |
+
"loss": 0.6267
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"step": 450,
|
| 194 |
+
"loss": 0.5952
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"step": 460,
|
| 198 |
+
"loss": 0.5969
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"step": 470,
|
| 202 |
+
"loss": 0.6283
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"step": 480,
|
| 206 |
+
"loss": 0.6008
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"step": 490,
|
| 210 |
+
"loss": 0.6077
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"step": 500,
|
| 214 |
+
"loss": 0.6015
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"step": 510,
|
| 218 |
+
"loss": 0.6063
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"step": 520,
|
| 222 |
+
"loss": 0.5804
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"step": 530,
|
| 226 |
+
"loss": 0.591
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"step": 540,
|
| 230 |
+
"loss": 0.6167
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"step": 550,
|
| 234 |
+
"loss": 0.5906
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"step": 560,
|
| 238 |
+
"loss": 0.5951
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"step": 570,
|
| 242 |
+
"loss": 0.6183
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"step": 580,
|
| 246 |
+
"loss": 0.6026
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"step": 590,
|
| 250 |
+
"loss": 0.567
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"step": 600,
|
| 254 |
+
"loss": 0.5989
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"step": 610,
|
| 258 |
+
"loss": 0.584
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"step": 620,
|
| 262 |
+
"loss": 0.5931
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"step": 630,
|
| 266 |
+
"loss": 0.634
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"step": 640,
|
| 270 |
+
"loss": 0.598
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"step": 650,
|
| 274 |
+
"loss": 0.6089
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"step": 660,
|
| 278 |
+
"loss": 0.5837
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"step": 670,
|
| 282 |
+
"loss": 0.5906
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"step": 680,
|
| 286 |
+
"loss": 0.5846
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"step": 690,
|
| 290 |
+
"loss": 0.589
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"step": 700,
|
| 294 |
+
"loss": 0.5828
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"step": 710,
|
| 298 |
+
"loss": 0.6046
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"step": 720,
|
| 302 |
+
"loss": 0.603
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"step": 730,
|
| 306 |
+
"loss": 0.5849
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"step": 740,
|
| 310 |
+
"loss": 0.5995
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"step": 750,
|
| 314 |
+
"loss": 0.5845
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"step": 760,
|
| 318 |
+
"loss": 0.5896
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"step": 770,
|
| 322 |
+
"loss": 0.5847
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"step": 780,
|
| 326 |
+
"loss": 0.6012
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"step": 790,
|
| 330 |
+
"loss": 0.5907
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"step": 800,
|
| 334 |
+
"loss": 0.5799
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"step": 810,
|
| 338 |
+
"loss": 0.5861
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"step": 820,
|
| 342 |
+
"loss": 0.5731
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"step": 830,
|
| 346 |
+
"loss": 0.5857
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"step": 840,
|
| 350 |
+
"loss": 0.5988
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"step": 850,
|
| 354 |
+
"loss": 0.5981
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"step": 860,
|
| 358 |
+
"loss": 0.5939
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"step": 870,
|
| 362 |
+
"loss": 0.5795
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"step": 880,
|
| 366 |
+
"loss": 0.5945
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"step": 890,
|
| 370 |
+
"loss": 0.5862
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"step": 900,
|
| 374 |
+
"loss": 0.5938
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"step": 910,
|
| 378 |
+
"loss": 0.5813
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"step": 920,
|
| 382 |
+
"loss": 0.5762
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"step": 930,
|
| 386 |
+
"loss": 0.5802
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"step": 940,
|
| 390 |
+
"loss": 0.5892
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"step": 950,
|
| 394 |
+
"loss": 0.5943
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"step": 960,
|
| 398 |
+
"loss": 0.5665
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"step": 970,
|
| 402 |
+
"loss": 0.5604
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"step": 980,
|
| 406 |
+
"loss": 0.5764
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"step": 990,
|
| 410 |
+
"loss": 0.5889
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"step": 1000,
|
| 414 |
+
"loss": 0.5704
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"step": 1010,
|
| 418 |
+
"loss": 0.591
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"step": 1020,
|
| 422 |
+
"loss": 0.5659
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"step": 1030,
|
| 426 |
+
"loss": 0.5973
|
| 427 |
+
}
|
| 428 |
+
]
|
| 429 |
+
}
|