Commit
·
33f8089
1
Parent(s):
f14fd41
Refined (v5.1) update
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- checkpoints/checkpoint-46000/trainer_state.json +0 -678
- checkpoints/checkpoint-46500/trainer_state.json +0 -685
- checkpoints/checkpoint-46670/trainer_state.json +0 -685
- checkpoints/{checkpoint-46000 → checkpoint-52500}/config.json +0 -0
- checkpoints/{checkpoint-46000 → checkpoint-52500}/generation_config.json +0 -0
- checkpoints/{checkpoint-46000 → checkpoint-52500}/model.safetensors +1 -1
- checkpoints/{checkpoint-46500 → checkpoint-52500}/optimizer.pt +1 -1
- checkpoints/{checkpoint-46670 → checkpoint-52500}/rng_state.pth +1 -1
- checkpoints/{checkpoint-46670 → checkpoint-52500}/scaler.pt +1 -1
- checkpoints/{checkpoint-46000 → checkpoint-52500}/scheduler.pt +1 -1
- checkpoints/{checkpoint-46000 → checkpoint-52500}/special_tokens_map.json +0 -0
- checkpoints/{checkpoint-46500 → checkpoint-52500}/spiece.model +2 -2
- checkpoints/{checkpoint-46000 → checkpoint-52500}/tokenizer.json +0 -0
- checkpoints/{checkpoint-46000 → checkpoint-52500}/tokenizer_config.json +0 -0
- checkpoints/checkpoint-52500/trainer_state.json +769 -0
- checkpoints/{checkpoint-46000 → checkpoint-52500}/training_args.bin +0 -0
- checkpoints/{checkpoint-46500 → checkpoint-53000}/config.json +0 -0
- checkpoints/{checkpoint-46500 → checkpoint-53000}/generation_config.json +0 -0
- checkpoints/{checkpoint-46500 → checkpoint-53000}/model.safetensors +1 -1
- checkpoints/{checkpoint-46000 → checkpoint-53000}/optimizer.pt +1 -1
- checkpoints/{checkpoint-46500 → checkpoint-53000}/rng_state.pth +1 -1
- checkpoints/{checkpoint-46500 → checkpoint-53000}/scaler.pt +1 -1
- checkpoints/{checkpoint-46670 → checkpoint-53000}/scheduler.pt +1 -1
- checkpoints/{checkpoint-46500 → checkpoint-53000}/special_tokens_map.json +0 -0
- checkpoints/{checkpoint-46000 → checkpoint-53000}/spiece.model +2 -2
- checkpoints/{checkpoint-46500 → checkpoint-53000}/tokenizer.json +0 -0
- checkpoints/{checkpoint-46500 → checkpoint-53000}/tokenizer_config.json +0 -0
- checkpoints/checkpoint-53000/trainer_state.json +776 -0
- checkpoints/{checkpoint-46500 → checkpoint-53000}/training_args.bin +0 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/config.json +0 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/generation_config.json +0 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/model.safetensors +1 -1
- checkpoints/{checkpoint-46670 → checkpoint-53280}/optimizer.pt +1 -1
- checkpoints/{checkpoint-46000 → checkpoint-53280}/rng_state.pth +1 -1
- checkpoints/{checkpoint-46000 → checkpoint-53280}/scaler.pt +1 -1
- checkpoints/{checkpoint-46500 → checkpoint-53280}/scheduler.pt +1 -1
- checkpoints/{checkpoint-46670 → checkpoint-53280}/special_tokens_map.json +0 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/spiece.model +2 -2
- checkpoints/{checkpoint-46670 → checkpoint-53280}/tokenizer.json +0 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/tokenizer_config.json +0 -0
- checkpoints/checkpoint-53280/trainer_state.json +776 -0
- checkpoints/{checkpoint-46670 → checkpoint-53280}/training_args.bin +0 -0
- model.safetensors +1 -1
- spiece.model +2 -2
- src/data/generate_cyr_lat_pairs.py +1 -1
- src/tokeniser/dalat5_sp.model +2 -2
- src/tokeniser/dalat5_sp.vocab +0 -0
- src/tokeniser/spiece.model +2 -2
- src/train_t5.py +1 -0
- src/train_tokeniser.py +1 -1
checkpoints/checkpoint-46000/trainer_state.json
DELETED
|
@@ -1,678 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 1.9712241006192284,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 46000,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.02142658181740267,
|
| 14 |
-
"grad_norm": 0.4460393786430359,
|
| 15 |
-
"learning_rate": 4.946539532890508e-05,
|
| 16 |
-
"loss": 2.2597,
|
| 17 |
-
"step": 500
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.04285316363480534,
|
| 21 |
-
"grad_norm": 0.4649476110935211,
|
| 22 |
-
"learning_rate": 4.8929719305763877e-05,
|
| 23 |
-
"loss": 2.0158,
|
| 24 |
-
"step": 1000
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.064279745452208,
|
| 28 |
-
"grad_norm": 0.5113596320152283,
|
| 29 |
-
"learning_rate": 4.839404328262267e-05,
|
| 30 |
-
"loss": 1.9311,
|
| 31 |
-
"step": 1500
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.08570632726961068,
|
| 35 |
-
"grad_norm": 0.49711284041404724,
|
| 36 |
-
"learning_rate": 4.785836725948147e-05,
|
| 37 |
-
"loss": 1.8587,
|
| 38 |
-
"step": 2000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.10713290908701335,
|
| 42 |
-
"grad_norm": 0.6695660352706909,
|
| 43 |
-
"learning_rate": 4.732269123634027e-05,
|
| 44 |
-
"loss": 1.8219,
|
| 45 |
-
"step": 2500
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.128559490904416,
|
| 49 |
-
"grad_norm": 0.5876463651657104,
|
| 50 |
-
"learning_rate": 4.678701521319906e-05,
|
| 51 |
-
"loss": 1.7645,
|
| 52 |
-
"step": 3000
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.1499860727218187,
|
| 56 |
-
"grad_norm": 0.554977297782898,
|
| 57 |
-
"learning_rate": 4.625133919005786e-05,
|
| 58 |
-
"loss": 1.7136,
|
| 59 |
-
"step": 3500
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.17141265453922136,
|
| 63 |
-
"grad_norm": 0.5401861667633057,
|
| 64 |
-
"learning_rate": 4.571566316691665e-05,
|
| 65 |
-
"loss": 1.696,
|
| 66 |
-
"step": 4000
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.19283923635662403,
|
| 70 |
-
"grad_norm": 0.5951809287071228,
|
| 71 |
-
"learning_rate": 4.517998714377545e-05,
|
| 72 |
-
"loss": 1.6519,
|
| 73 |
-
"step": 4500
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.2142658181740267,
|
| 77 |
-
"grad_norm": 0.6037197113037109,
|
| 78 |
-
"learning_rate": 4.464431112063424e-05,
|
| 79 |
-
"loss": 1.6279,
|
| 80 |
-
"step": 5000
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.23569239999142938,
|
| 84 |
-
"grad_norm": 0.6542425751686096,
|
| 85 |
-
"learning_rate": 4.410863509749304e-05,
|
| 86 |
-
"loss": 1.6104,
|
| 87 |
-
"step": 5500
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.257118981808832,
|
| 91 |
-
"grad_norm": 0.5781915783882141,
|
| 92 |
-
"learning_rate": 4.357295907435183e-05,
|
| 93 |
-
"loss": 1.5976,
|
| 94 |
-
"step": 6000
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.2785455636262347,
|
| 98 |
-
"grad_norm": 0.575715184211731,
|
| 99 |
-
"learning_rate": 4.303728305121063e-05,
|
| 100 |
-
"loss": 1.5373,
|
| 101 |
-
"step": 6500
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.2999721454436374,
|
| 105 |
-
"grad_norm": 0.6789396405220032,
|
| 106 |
-
"learning_rate": 4.2501607028069424e-05,
|
| 107 |
-
"loss": 1.5435,
|
| 108 |
-
"step": 7000
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.3213987272610401,
|
| 112 |
-
"grad_norm": 0.6032198667526245,
|
| 113 |
-
"learning_rate": 4.196593100492822e-05,
|
| 114 |
-
"loss": 1.5081,
|
| 115 |
-
"step": 7500
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.3428253090784427,
|
| 119 |
-
"grad_norm": 0.6690296530723572,
|
| 120 |
-
"learning_rate": 4.1430254981787015e-05,
|
| 121 |
-
"loss": 1.5059,
|
| 122 |
-
"step": 8000
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.36425189089584536,
|
| 126 |
-
"grad_norm": 0.6951475739479065,
|
| 127 |
-
"learning_rate": 4.0894578958645814e-05,
|
| 128 |
-
"loss": 1.4716,
|
| 129 |
-
"step": 8500
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.38567847271324807,
|
| 133 |
-
"grad_norm": 0.7095356583595276,
|
| 134 |
-
"learning_rate": 4.035890293550461e-05,
|
| 135 |
-
"loss": 1.455,
|
| 136 |
-
"step": 9000
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.4071050545306507,
|
| 140 |
-
"grad_norm": 0.8375953435897827,
|
| 141 |
-
"learning_rate": 3.9823226912363405e-05,
|
| 142 |
-
"loss": 1.4527,
|
| 143 |
-
"step": 9500
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.4285316363480534,
|
| 147 |
-
"grad_norm": 0.8582206964492798,
|
| 148 |
-
"learning_rate": 3.9287550889222204e-05,
|
| 149 |
-
"loss": 1.4345,
|
| 150 |
-
"step": 10000
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.44995821816545606,
|
| 154 |
-
"grad_norm": 0.8591541051864624,
|
| 155 |
-
"learning_rate": 3.8751874866081e-05,
|
| 156 |
-
"loss": 1.4268,
|
| 157 |
-
"step": 10500
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.47138479998285876,
|
| 161 |
-
"grad_norm": 0.835361897945404,
|
| 162 |
-
"learning_rate": 3.8216198842939796e-05,
|
| 163 |
-
"loss": 1.4045,
|
| 164 |
-
"step": 11000
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.4928113818002614,
|
| 168 |
-
"grad_norm": 0.9994288086891174,
|
| 169 |
-
"learning_rate": 3.768052281979859e-05,
|
| 170 |
-
"loss": 1.3965,
|
| 171 |
-
"step": 11500
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.514237963617664,
|
| 175 |
-
"grad_norm": 0.8310408592224121,
|
| 176 |
-
"learning_rate": 3.714484679665739e-05,
|
| 177 |
-
"loss": 1.3577,
|
| 178 |
-
"step": 12000
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.5356645454350667,
|
| 182 |
-
"grad_norm": 1.061630368232727,
|
| 183 |
-
"learning_rate": 3.660917077351618e-05,
|
| 184 |
-
"loss": 1.3311,
|
| 185 |
-
"step": 12500
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.5570911272524695,
|
| 189 |
-
"grad_norm": 1.0928676128387451,
|
| 190 |
-
"learning_rate": 3.607349475037498e-05,
|
| 191 |
-
"loss": 1.3203,
|
| 192 |
-
"step": 13000
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.5785177090698721,
|
| 196 |
-
"grad_norm": 1.4662648439407349,
|
| 197 |
-
"learning_rate": 3.553781872723377e-05,
|
| 198 |
-
"loss": 1.2778,
|
| 199 |
-
"step": 13500
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.5999442908872747,
|
| 203 |
-
"grad_norm": 1.2254618406295776,
|
| 204 |
-
"learning_rate": 3.500214270409257e-05,
|
| 205 |
-
"loss": 1.2572,
|
| 206 |
-
"step": 14000
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.6213708727046774,
|
| 210 |
-
"grad_norm": 1.275029182434082,
|
| 211 |
-
"learning_rate": 3.446860938504393e-05,
|
| 212 |
-
"loss": 1.2707,
|
| 213 |
-
"step": 14500
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.6427974545220801,
|
| 217 |
-
"grad_norm": 1.3197990655899048,
|
| 218 |
-
"learning_rate": 3.3932933361902726e-05,
|
| 219 |
-
"loss": 1.2308,
|
| 220 |
-
"step": 15000
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.6642240363394828,
|
| 224 |
-
"grad_norm": 1.934342861175537,
|
| 225 |
-
"learning_rate": 3.339725733876152e-05,
|
| 226 |
-
"loss": 1.2268,
|
| 227 |
-
"step": 15500
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.6856506181568854,
|
| 231 |
-
"grad_norm": 1.445302128791809,
|
| 232 |
-
"learning_rate": 3.286158131562032e-05,
|
| 233 |
-
"loss": 1.2007,
|
| 234 |
-
"step": 16000
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.7070771999742881,
|
| 238 |
-
"grad_norm": 1.2702606916427612,
|
| 239 |
-
"learning_rate": 3.232697664452539e-05,
|
| 240 |
-
"loss": 1.1785,
|
| 241 |
-
"step": 16500
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.7285037817916907,
|
| 245 |
-
"grad_norm": 1.254937767982483,
|
| 246 |
-
"learning_rate": 3.179130062138419e-05,
|
| 247 |
-
"loss": 1.1616,
|
| 248 |
-
"step": 17000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.7499303636090935,
|
| 252 |
-
"grad_norm": 1.3628320693969727,
|
| 253 |
-
"learning_rate": 3.1255624598242984e-05,
|
| 254 |
-
"loss": 1.1352,
|
| 255 |
-
"step": 17500
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.7713569454264961,
|
| 259 |
-
"grad_norm": 3.0748207569122314,
|
| 260 |
-
"learning_rate": 3.071994857510178e-05,
|
| 261 |
-
"loss": 1.1243,
|
| 262 |
-
"step": 18000
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.7927835272438988,
|
| 266 |
-
"grad_norm": 1.4262775182724,
|
| 267 |
-
"learning_rate": 3.0185343904006858e-05,
|
| 268 |
-
"loss": 1.1011,
|
| 269 |
-
"step": 18500
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.8142101090613014,
|
| 273 |
-
"grad_norm": 2.2326557636260986,
|
| 274 |
-
"learning_rate": 2.9650739232911933e-05,
|
| 275 |
-
"loss": 1.0974,
|
| 276 |
-
"step": 19000
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.8356366908787041,
|
| 280 |
-
"grad_norm": 2.791055679321289,
|
| 281 |
-
"learning_rate": 2.9115063209770732e-05,
|
| 282 |
-
"loss": 1.0762,
|
| 283 |
-
"step": 19500
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.8570632726961068,
|
| 287 |
-
"grad_norm": 3.0886242389678955,
|
| 288 |
-
"learning_rate": 2.857938718662953e-05,
|
| 289 |
-
"loss": 1.0698,
|
| 290 |
-
"step": 20000
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.8784898545135095,
|
| 294 |
-
"grad_norm": 1.4234964847564697,
|
| 295 |
-
"learning_rate": 2.8043711163488323e-05,
|
| 296 |
-
"loss": 1.0541,
|
| 297 |
-
"step": 20500
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.8999164363309121,
|
| 301 |
-
"grad_norm": 3.275305986404419,
|
| 302 |
-
"learning_rate": 2.75091064923934e-05,
|
| 303 |
-
"loss": 1.0396,
|
| 304 |
-
"step": 21000
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.9213430181483148,
|
| 308 |
-
"grad_norm": 1.6365532875061035,
|
| 309 |
-
"learning_rate": 2.6973430469252198e-05,
|
| 310 |
-
"loss": 1.0146,
|
| 311 |
-
"step": 21500
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.9427695999657175,
|
| 315 |
-
"grad_norm": 3.1808159351348877,
|
| 316 |
-
"learning_rate": 2.6437754446110997e-05,
|
| 317 |
-
"loss": 0.9997,
|
| 318 |
-
"step": 22000
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.9641961817831202,
|
| 322 |
-
"grad_norm": 1.8644700050354004,
|
| 323 |
-
"learning_rate": 2.590207842296979e-05,
|
| 324 |
-
"loss": 0.9905,
|
| 325 |
-
"step": 22500
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.9856227636005228,
|
| 329 |
-
"grad_norm": 1.5963947772979736,
|
| 330 |
-
"learning_rate": 2.5367473751874864e-05,
|
| 331 |
-
"loss": 0.9624,
|
| 332 |
-
"step": 23000
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 1.007027918836108,
|
| 336 |
-
"grad_norm": 1.5474953651428223,
|
| 337 |
-
"learning_rate": 2.4831797728733663e-05,
|
| 338 |
-
"loss": 0.958,
|
| 339 |
-
"step": 23500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.0284545006535109,
|
| 343 |
-
"grad_norm": 3.8564932346343994,
|
| 344 |
-
"learning_rate": 2.429612170559246e-05,
|
| 345 |
-
"loss": 0.9388,
|
| 346 |
-
"step": 24000
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.0498810824709135,
|
| 350 |
-
"grad_norm": 2.154879331588745,
|
| 351 |
-
"learning_rate": 2.3760445682451254e-05,
|
| 352 |
-
"loss": 0.9132,
|
| 353 |
-
"step": 24500
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 1.0713076642883161,
|
| 357 |
-
"grad_norm": 4.118302822113037,
|
| 358 |
-
"learning_rate": 2.3225841011356333e-05,
|
| 359 |
-
"loss": 0.9132,
|
| 360 |
-
"step": 25000
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 1.0927342461057188,
|
| 364 |
-
"grad_norm": 2.0971686840057373,
|
| 365 |
-
"learning_rate": 2.2690164988215128e-05,
|
| 366 |
-
"loss": 0.8947,
|
| 367 |
-
"step": 25500
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 1.1141608279231214,
|
| 371 |
-
"grad_norm": 3.480602741241455,
|
| 372 |
-
"learning_rate": 2.2154488965073924e-05,
|
| 373 |
-
"loss": 0.8922,
|
| 374 |
-
"step": 26000
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 1.135587409740524,
|
| 378 |
-
"grad_norm": 4.0732879638671875,
|
| 379 |
-
"learning_rate": 2.161881294193272e-05,
|
| 380 |
-
"loss": 0.87,
|
| 381 |
-
"step": 26500
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.1570139915579267,
|
| 385 |
-
"grad_norm": 2.2193567752838135,
|
| 386 |
-
"learning_rate": 2.1084208270837798e-05,
|
| 387 |
-
"loss": 0.8476,
|
| 388 |
-
"step": 27000
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 1.1784405733753294,
|
| 392 |
-
"grad_norm": 2.1723575592041016,
|
| 393 |
-
"learning_rate": 2.0548532247696594e-05,
|
| 394 |
-
"loss": 0.8322,
|
| 395 |
-
"step": 27500
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 1.199867155192732,
|
| 399 |
-
"grad_norm": 3.3503777980804443,
|
| 400 |
-
"learning_rate": 2.001285622455539e-05,
|
| 401 |
-
"loss": 0.8171,
|
| 402 |
-
"step": 28000
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 1.2212937370101349,
|
| 406 |
-
"grad_norm": 3.6080329418182373,
|
| 407 |
-
"learning_rate": 1.9477180201414185e-05,
|
| 408 |
-
"loss": 0.8105,
|
| 409 |
-
"step": 28500
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 1.2427203188275375,
|
| 413 |
-
"grad_norm": 2.708294630050659,
|
| 414 |
-
"learning_rate": 1.894150417827298e-05,
|
| 415 |
-
"loss": 0.7808,
|
| 416 |
-
"step": 29000
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 1.2641469006449402,
|
| 420 |
-
"grad_norm": 3.300328493118286,
|
| 421 |
-
"learning_rate": 1.8405828155131776e-05,
|
| 422 |
-
"loss": 0.7774,
|
| 423 |
-
"step": 29500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 1.2855734824623428,
|
| 427 |
-
"grad_norm": 2.1943624019622803,
|
| 428 |
-
"learning_rate": 1.7870152131990572e-05,
|
| 429 |
-
"loss": 0.7637,
|
| 430 |
-
"step": 30000
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 1.3070000642797455,
|
| 434 |
-
"grad_norm": 2.7251927852630615,
|
| 435 |
-
"learning_rate": 1.733554746089565e-05,
|
| 436 |
-
"loss": 0.7481,
|
| 437 |
-
"step": 30500
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 1.3284266460971481,
|
| 441 |
-
"grad_norm": 4.672070026397705,
|
| 442 |
-
"learning_rate": 1.6799871437754446e-05,
|
| 443 |
-
"loss": 0.7255,
|
| 444 |
-
"step": 31000
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.3498532279145508,
|
| 448 |
-
"grad_norm": 2.4984779357910156,
|
| 449 |
-
"learning_rate": 1.626419541461324e-05,
|
| 450 |
-
"loss": 0.7231,
|
| 451 |
-
"step": 31500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 1.3712798097319534,
|
| 455 |
-
"grad_norm": 1.9306334257125854,
|
| 456 |
-
"learning_rate": 1.5728519391472037e-05,
|
| 457 |
-
"loss": 0.7032,
|
| 458 |
-
"step": 32000
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 1.392706391549356,
|
| 462 |
-
"grad_norm": 3.015226125717163,
|
| 463 |
-
"learning_rate": 1.5192843368330834e-05,
|
| 464 |
-
"loss": 0.6923,
|
| 465 |
-
"step": 32500
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.4141329733667587,
|
| 469 |
-
"grad_norm": 3.646979331970215,
|
| 470 |
-
"learning_rate": 1.4658238697235913e-05,
|
| 471 |
-
"loss": 0.6893,
|
| 472 |
-
"step": 33000
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 1.4355595551841613,
|
| 476 |
-
"grad_norm": 1.8951635360717773,
|
| 477 |
-
"learning_rate": 1.4122562674094708e-05,
|
| 478 |
-
"loss": 0.6693,
|
| 479 |
-
"step": 33500
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 1.4569861370015642,
|
| 483 |
-
"grad_norm": 2.3496530055999756,
|
| 484 |
-
"learning_rate": 1.3586886650953504e-05,
|
| 485 |
-
"loss": 0.6586,
|
| 486 |
-
"step": 34000
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 1.4784127188189669,
|
| 490 |
-
"grad_norm": 2.899231195449829,
|
| 491 |
-
"learning_rate": 1.30512106278123e-05,
|
| 492 |
-
"loss": 0.6338,
|
| 493 |
-
"step": 34500
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.4998393006363695,
|
| 497 |
-
"grad_norm": 2.7465250492095947,
|
| 498 |
-
"learning_rate": 1.2515534604671095e-05,
|
| 499 |
-
"loss": 0.6291,
|
| 500 |
-
"step": 35000
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.5212658824537721,
|
| 504 |
-
"grad_norm": 2.3695950508117676,
|
| 505 |
-
"learning_rate": 1.1979858581529891e-05,
|
| 506 |
-
"loss": 0.6192,
|
| 507 |
-
"step": 35500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.5426924642711748,
|
| 511 |
-
"grad_norm": 3.0361921787261963,
|
| 512 |
-
"learning_rate": 1.1444182558388687e-05,
|
| 513 |
-
"loss": 0.6147,
|
| 514 |
-
"step": 36000
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 1.5641190460885774,
|
| 518 |
-
"grad_norm": 3.5043649673461914,
|
| 519 |
-
"learning_rate": 1.0908506535247482e-05,
|
| 520 |
-
"loss": 0.6024,
|
| 521 |
-
"step": 36500
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 1.5855456279059803,
|
| 525 |
-
"grad_norm": 2.1166698932647705,
|
| 526 |
-
"learning_rate": 1.037390186415256e-05,
|
| 527 |
-
"loss": 0.6033,
|
| 528 |
-
"step": 37000
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 1.606972209723383,
|
| 532 |
-
"grad_norm": 2.8012290000915527,
|
| 533 |
-
"learning_rate": 9.838225841011358e-06,
|
| 534 |
-
"loss": 0.5935,
|
| 535 |
-
"step": 37500
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 1.6283987915407856,
|
| 539 |
-
"grad_norm": 2.49540638923645,
|
| 540 |
-
"learning_rate": 9.302549817870154e-06,
|
| 541 |
-
"loss": 0.5819,
|
| 542 |
-
"step": 38000
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 1.6498253733581882,
|
| 546 |
-
"grad_norm": 2.233430862426758,
|
| 547 |
-
"learning_rate": 8.766873794728948e-06,
|
| 548 |
-
"loss": 0.5674,
|
| 549 |
-
"step": 38500
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.6712519551755909,
|
| 553 |
-
"grad_norm": 2.823101282119751,
|
| 554 |
-
"learning_rate": 8.232269123634026e-06,
|
| 555 |
-
"loss": 0.552,
|
| 556 |
-
"step": 39000
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 1.6926785369929935,
|
| 560 |
-
"grad_norm": 2.389265775680542,
|
| 561 |
-
"learning_rate": 7.696593100492823e-06,
|
| 562 |
-
"loss": 0.5564,
|
| 563 |
-
"step": 39500
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 1.7141051188103962,
|
| 567 |
-
"grad_norm": 2.1136868000030518,
|
| 568 |
-
"learning_rate": 7.161988429397901e-06,
|
| 569 |
-
"loss": 0.5477,
|
| 570 |
-
"step": 40000
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 1.7355317006277988,
|
| 574 |
-
"grad_norm": 2.6307172775268555,
|
| 575 |
-
"learning_rate": 6.626312406256696e-06,
|
| 576 |
-
"loss": 0.545,
|
| 577 |
-
"step": 40500
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 1.7569582824452015,
|
| 581 |
-
"grad_norm": 3.4348325729370117,
|
| 582 |
-
"learning_rate": 6.090636383115492e-06,
|
| 583 |
-
"loss": 0.5359,
|
| 584 |
-
"step": 41000
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 1.7783848642626041,
|
| 588 |
-
"grad_norm": 2.1959691047668457,
|
| 589 |
-
"learning_rate": 5.554960359974288e-06,
|
| 590 |
-
"loss": 0.5267,
|
| 591 |
-
"step": 41500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 1.7998114460800068,
|
| 595 |
-
"grad_norm": 2.747068405151367,
|
| 596 |
-
"learning_rate": 5.0192843368330835e-06,
|
| 597 |
-
"loss": 0.5316,
|
| 598 |
-
"step": 42000
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 1.8212380278974094,
|
| 602 |
-
"grad_norm": 2.965233325958252,
|
| 603 |
-
"learning_rate": 4.48360831369188e-06,
|
| 604 |
-
"loss": 0.5129,
|
| 605 |
-
"step": 42500
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 1.842664609714812,
|
| 609 |
-
"grad_norm": 2.3425612449645996,
|
| 610 |
-
"learning_rate": 3.9479322905506756e-06,
|
| 611 |
-
"loss": 0.5058,
|
| 612 |
-
"step": 43000
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 1.864091191532215,
|
| 616 |
-
"grad_norm": 1.5650664567947388,
|
| 617 |
-
"learning_rate": 3.4122562674094708e-06,
|
| 618 |
-
"loss": 0.5155,
|
| 619 |
-
"step": 43500
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 1.8855177733496176,
|
| 623 |
-
"grad_norm": 2.501758098602295,
|
| 624 |
-
"learning_rate": 2.8776515963145492e-06,
|
| 625 |
-
"loss": 0.5076,
|
| 626 |
-
"step": 44000
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 1.9069443551670202,
|
| 630 |
-
"grad_norm": 4.431907653808594,
|
| 631 |
-
"learning_rate": 2.341975573173345e-06,
|
| 632 |
-
"loss": 0.5119,
|
| 633 |
-
"step": 44500
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 1.9283709369844229,
|
| 637 |
-
"grad_norm": 3.3445732593536377,
|
| 638 |
-
"learning_rate": 1.8062995500321405e-06,
|
| 639 |
-
"loss": 0.5021,
|
| 640 |
-
"step": 45000
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 1.9497975188018255,
|
| 644 |
-
"grad_norm": 2.6224589347839355,
|
| 645 |
-
"learning_rate": 1.2706235268909363e-06,
|
| 646 |
-
"loss": 0.5019,
|
| 647 |
-
"step": 45500
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 1.9712241006192284,
|
| 651 |
-
"grad_norm": 2.954880952835083,
|
| 652 |
-
"learning_rate": 7.360188557960147e-07,
|
| 653 |
-
"loss": 0.5055,
|
| 654 |
-
"step": 46000
|
| 655 |
-
}
|
| 656 |
-
],
|
| 657 |
-
"logging_steps": 500,
|
| 658 |
-
"max_steps": 46670,
|
| 659 |
-
"num_input_tokens_seen": 0,
|
| 660 |
-
"num_train_epochs": 2,
|
| 661 |
-
"save_steps": 500,
|
| 662 |
-
"stateful_callbacks": {
|
| 663 |
-
"TrainerControl": {
|
| 664 |
-
"args": {
|
| 665 |
-
"should_epoch_stop": false,
|
| 666 |
-
"should_evaluate": false,
|
| 667 |
-
"should_log": false,
|
| 668 |
-
"should_save": true,
|
| 669 |
-
"should_training_stop": false
|
| 670 |
-
},
|
| 671 |
-
"attributes": {}
|
| 672 |
-
}
|
| 673 |
-
},
|
| 674 |
-
"total_flos": 9.961028013627802e+16,
|
| 675 |
-
"train_batch_size": 32,
|
| 676 |
-
"trial_name": null,
|
| 677 |
-
"trial_params": null
|
| 678 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints/checkpoint-46500/trainer_state.json
DELETED
|
@@ -1,685 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 1.992650682436631,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 46500,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.02142658181740267,
|
| 14 |
-
"grad_norm": 0.4460393786430359,
|
| 15 |
-
"learning_rate": 4.946539532890508e-05,
|
| 16 |
-
"loss": 2.2597,
|
| 17 |
-
"step": 500
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.04285316363480534,
|
| 21 |
-
"grad_norm": 0.4649476110935211,
|
| 22 |
-
"learning_rate": 4.8929719305763877e-05,
|
| 23 |
-
"loss": 2.0158,
|
| 24 |
-
"step": 1000
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.064279745452208,
|
| 28 |
-
"grad_norm": 0.5113596320152283,
|
| 29 |
-
"learning_rate": 4.839404328262267e-05,
|
| 30 |
-
"loss": 1.9311,
|
| 31 |
-
"step": 1500
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.08570632726961068,
|
| 35 |
-
"grad_norm": 0.49711284041404724,
|
| 36 |
-
"learning_rate": 4.785836725948147e-05,
|
| 37 |
-
"loss": 1.8587,
|
| 38 |
-
"step": 2000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.10713290908701335,
|
| 42 |
-
"grad_norm": 0.6695660352706909,
|
| 43 |
-
"learning_rate": 4.732269123634027e-05,
|
| 44 |
-
"loss": 1.8219,
|
| 45 |
-
"step": 2500
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.128559490904416,
|
| 49 |
-
"grad_norm": 0.5876463651657104,
|
| 50 |
-
"learning_rate": 4.678701521319906e-05,
|
| 51 |
-
"loss": 1.7645,
|
| 52 |
-
"step": 3000
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.1499860727218187,
|
| 56 |
-
"grad_norm": 0.554977297782898,
|
| 57 |
-
"learning_rate": 4.625133919005786e-05,
|
| 58 |
-
"loss": 1.7136,
|
| 59 |
-
"step": 3500
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.17141265453922136,
|
| 63 |
-
"grad_norm": 0.5401861667633057,
|
| 64 |
-
"learning_rate": 4.571566316691665e-05,
|
| 65 |
-
"loss": 1.696,
|
| 66 |
-
"step": 4000
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.19283923635662403,
|
| 70 |
-
"grad_norm": 0.5951809287071228,
|
| 71 |
-
"learning_rate": 4.517998714377545e-05,
|
| 72 |
-
"loss": 1.6519,
|
| 73 |
-
"step": 4500
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.2142658181740267,
|
| 77 |
-
"grad_norm": 0.6037197113037109,
|
| 78 |
-
"learning_rate": 4.464431112063424e-05,
|
| 79 |
-
"loss": 1.6279,
|
| 80 |
-
"step": 5000
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.23569239999142938,
|
| 84 |
-
"grad_norm": 0.6542425751686096,
|
| 85 |
-
"learning_rate": 4.410863509749304e-05,
|
| 86 |
-
"loss": 1.6104,
|
| 87 |
-
"step": 5500
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.257118981808832,
|
| 91 |
-
"grad_norm": 0.5781915783882141,
|
| 92 |
-
"learning_rate": 4.357295907435183e-05,
|
| 93 |
-
"loss": 1.5976,
|
| 94 |
-
"step": 6000
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.2785455636262347,
|
| 98 |
-
"grad_norm": 0.575715184211731,
|
| 99 |
-
"learning_rate": 4.303728305121063e-05,
|
| 100 |
-
"loss": 1.5373,
|
| 101 |
-
"step": 6500
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.2999721454436374,
|
| 105 |
-
"grad_norm": 0.6789396405220032,
|
| 106 |
-
"learning_rate": 4.2501607028069424e-05,
|
| 107 |
-
"loss": 1.5435,
|
| 108 |
-
"step": 7000
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.3213987272610401,
|
| 112 |
-
"grad_norm": 0.6032198667526245,
|
| 113 |
-
"learning_rate": 4.196593100492822e-05,
|
| 114 |
-
"loss": 1.5081,
|
| 115 |
-
"step": 7500
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.3428253090784427,
|
| 119 |
-
"grad_norm": 0.6690296530723572,
|
| 120 |
-
"learning_rate": 4.1430254981787015e-05,
|
| 121 |
-
"loss": 1.5059,
|
| 122 |
-
"step": 8000
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.36425189089584536,
|
| 126 |
-
"grad_norm": 0.6951475739479065,
|
| 127 |
-
"learning_rate": 4.0894578958645814e-05,
|
| 128 |
-
"loss": 1.4716,
|
| 129 |
-
"step": 8500
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.38567847271324807,
|
| 133 |
-
"grad_norm": 0.7095356583595276,
|
| 134 |
-
"learning_rate": 4.035890293550461e-05,
|
| 135 |
-
"loss": 1.455,
|
| 136 |
-
"step": 9000
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.4071050545306507,
|
| 140 |
-
"grad_norm": 0.8375953435897827,
|
| 141 |
-
"learning_rate": 3.9823226912363405e-05,
|
| 142 |
-
"loss": 1.4527,
|
| 143 |
-
"step": 9500
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.4285316363480534,
|
| 147 |
-
"grad_norm": 0.8582206964492798,
|
| 148 |
-
"learning_rate": 3.9287550889222204e-05,
|
| 149 |
-
"loss": 1.4345,
|
| 150 |
-
"step": 10000
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.44995821816545606,
|
| 154 |
-
"grad_norm": 0.8591541051864624,
|
| 155 |
-
"learning_rate": 3.8751874866081e-05,
|
| 156 |
-
"loss": 1.4268,
|
| 157 |
-
"step": 10500
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.47138479998285876,
|
| 161 |
-
"grad_norm": 0.835361897945404,
|
| 162 |
-
"learning_rate": 3.8216198842939796e-05,
|
| 163 |
-
"loss": 1.4045,
|
| 164 |
-
"step": 11000
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.4928113818002614,
|
| 168 |
-
"grad_norm": 0.9994288086891174,
|
| 169 |
-
"learning_rate": 3.768052281979859e-05,
|
| 170 |
-
"loss": 1.3965,
|
| 171 |
-
"step": 11500
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.514237963617664,
|
| 175 |
-
"grad_norm": 0.8310408592224121,
|
| 176 |
-
"learning_rate": 3.714484679665739e-05,
|
| 177 |
-
"loss": 1.3577,
|
| 178 |
-
"step": 12000
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.5356645454350667,
|
| 182 |
-
"grad_norm": 1.061630368232727,
|
| 183 |
-
"learning_rate": 3.660917077351618e-05,
|
| 184 |
-
"loss": 1.3311,
|
| 185 |
-
"step": 12500
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.5570911272524695,
|
| 189 |
-
"grad_norm": 1.0928676128387451,
|
| 190 |
-
"learning_rate": 3.607349475037498e-05,
|
| 191 |
-
"loss": 1.3203,
|
| 192 |
-
"step": 13000
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.5785177090698721,
|
| 196 |
-
"grad_norm": 1.4662648439407349,
|
| 197 |
-
"learning_rate": 3.553781872723377e-05,
|
| 198 |
-
"loss": 1.2778,
|
| 199 |
-
"step": 13500
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.5999442908872747,
|
| 203 |
-
"grad_norm": 1.2254618406295776,
|
| 204 |
-
"learning_rate": 3.500214270409257e-05,
|
| 205 |
-
"loss": 1.2572,
|
| 206 |
-
"step": 14000
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.6213708727046774,
|
| 210 |
-
"grad_norm": 1.275029182434082,
|
| 211 |
-
"learning_rate": 3.446860938504393e-05,
|
| 212 |
-
"loss": 1.2707,
|
| 213 |
-
"step": 14500
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.6427974545220801,
|
| 217 |
-
"grad_norm": 1.3197990655899048,
|
| 218 |
-
"learning_rate": 3.3932933361902726e-05,
|
| 219 |
-
"loss": 1.2308,
|
| 220 |
-
"step": 15000
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.6642240363394828,
|
| 224 |
-
"grad_norm": 1.934342861175537,
|
| 225 |
-
"learning_rate": 3.339725733876152e-05,
|
| 226 |
-
"loss": 1.2268,
|
| 227 |
-
"step": 15500
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.6856506181568854,
|
| 231 |
-
"grad_norm": 1.445302128791809,
|
| 232 |
-
"learning_rate": 3.286158131562032e-05,
|
| 233 |
-
"loss": 1.2007,
|
| 234 |
-
"step": 16000
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.7070771999742881,
|
| 238 |
-
"grad_norm": 1.2702606916427612,
|
| 239 |
-
"learning_rate": 3.232697664452539e-05,
|
| 240 |
-
"loss": 1.1785,
|
| 241 |
-
"step": 16500
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.7285037817916907,
|
| 245 |
-
"grad_norm": 1.254937767982483,
|
| 246 |
-
"learning_rate": 3.179130062138419e-05,
|
| 247 |
-
"loss": 1.1616,
|
| 248 |
-
"step": 17000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.7499303636090935,
|
| 252 |
-
"grad_norm": 1.3628320693969727,
|
| 253 |
-
"learning_rate": 3.1255624598242984e-05,
|
| 254 |
-
"loss": 1.1352,
|
| 255 |
-
"step": 17500
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.7713569454264961,
|
| 259 |
-
"grad_norm": 3.0748207569122314,
|
| 260 |
-
"learning_rate": 3.071994857510178e-05,
|
| 261 |
-
"loss": 1.1243,
|
| 262 |
-
"step": 18000
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.7927835272438988,
|
| 266 |
-
"grad_norm": 1.4262775182724,
|
| 267 |
-
"learning_rate": 3.0185343904006858e-05,
|
| 268 |
-
"loss": 1.1011,
|
| 269 |
-
"step": 18500
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.8142101090613014,
|
| 273 |
-
"grad_norm": 2.2326557636260986,
|
| 274 |
-
"learning_rate": 2.9650739232911933e-05,
|
| 275 |
-
"loss": 1.0974,
|
| 276 |
-
"step": 19000
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.8356366908787041,
|
| 280 |
-
"grad_norm": 2.791055679321289,
|
| 281 |
-
"learning_rate": 2.9115063209770732e-05,
|
| 282 |
-
"loss": 1.0762,
|
| 283 |
-
"step": 19500
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.8570632726961068,
|
| 287 |
-
"grad_norm": 3.0886242389678955,
|
| 288 |
-
"learning_rate": 2.857938718662953e-05,
|
| 289 |
-
"loss": 1.0698,
|
| 290 |
-
"step": 20000
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.8784898545135095,
|
| 294 |
-
"grad_norm": 1.4234964847564697,
|
| 295 |
-
"learning_rate": 2.8043711163488323e-05,
|
| 296 |
-
"loss": 1.0541,
|
| 297 |
-
"step": 20500
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.8999164363309121,
|
| 301 |
-
"grad_norm": 3.275305986404419,
|
| 302 |
-
"learning_rate": 2.75091064923934e-05,
|
| 303 |
-
"loss": 1.0396,
|
| 304 |
-
"step": 21000
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.9213430181483148,
|
| 308 |
-
"grad_norm": 1.6365532875061035,
|
| 309 |
-
"learning_rate": 2.6973430469252198e-05,
|
| 310 |
-
"loss": 1.0146,
|
| 311 |
-
"step": 21500
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.9427695999657175,
|
| 315 |
-
"grad_norm": 3.1808159351348877,
|
| 316 |
-
"learning_rate": 2.6437754446110997e-05,
|
| 317 |
-
"loss": 0.9997,
|
| 318 |
-
"step": 22000
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.9641961817831202,
|
| 322 |
-
"grad_norm": 1.8644700050354004,
|
| 323 |
-
"learning_rate": 2.590207842296979e-05,
|
| 324 |
-
"loss": 0.9905,
|
| 325 |
-
"step": 22500
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.9856227636005228,
|
| 329 |
-
"grad_norm": 1.5963947772979736,
|
| 330 |
-
"learning_rate": 2.5367473751874864e-05,
|
| 331 |
-
"loss": 0.9624,
|
| 332 |
-
"step": 23000
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 1.007027918836108,
|
| 336 |
-
"grad_norm": 1.5474953651428223,
|
| 337 |
-
"learning_rate": 2.4831797728733663e-05,
|
| 338 |
-
"loss": 0.958,
|
| 339 |
-
"step": 23500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.0284545006535109,
|
| 343 |
-
"grad_norm": 3.8564932346343994,
|
| 344 |
-
"learning_rate": 2.429612170559246e-05,
|
| 345 |
-
"loss": 0.9388,
|
| 346 |
-
"step": 24000
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.0498810824709135,
|
| 350 |
-
"grad_norm": 2.154879331588745,
|
| 351 |
-
"learning_rate": 2.3760445682451254e-05,
|
| 352 |
-
"loss": 0.9132,
|
| 353 |
-
"step": 24500
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 1.0713076642883161,
|
| 357 |
-
"grad_norm": 4.118302822113037,
|
| 358 |
-
"learning_rate": 2.3225841011356333e-05,
|
| 359 |
-
"loss": 0.9132,
|
| 360 |
-
"step": 25000
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 1.0927342461057188,
|
| 364 |
-
"grad_norm": 2.0971686840057373,
|
| 365 |
-
"learning_rate": 2.2690164988215128e-05,
|
| 366 |
-
"loss": 0.8947,
|
| 367 |
-
"step": 25500
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 1.1141608279231214,
|
| 371 |
-
"grad_norm": 3.480602741241455,
|
| 372 |
-
"learning_rate": 2.2154488965073924e-05,
|
| 373 |
-
"loss": 0.8922,
|
| 374 |
-
"step": 26000
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 1.135587409740524,
|
| 378 |
-
"grad_norm": 4.0732879638671875,
|
| 379 |
-
"learning_rate": 2.161881294193272e-05,
|
| 380 |
-
"loss": 0.87,
|
| 381 |
-
"step": 26500
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.1570139915579267,
|
| 385 |
-
"grad_norm": 2.2193567752838135,
|
| 386 |
-
"learning_rate": 2.1084208270837798e-05,
|
| 387 |
-
"loss": 0.8476,
|
| 388 |
-
"step": 27000
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 1.1784405733753294,
|
| 392 |
-
"grad_norm": 2.1723575592041016,
|
| 393 |
-
"learning_rate": 2.0548532247696594e-05,
|
| 394 |
-
"loss": 0.8322,
|
| 395 |
-
"step": 27500
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 1.199867155192732,
|
| 399 |
-
"grad_norm": 3.3503777980804443,
|
| 400 |
-
"learning_rate": 2.001285622455539e-05,
|
| 401 |
-
"loss": 0.8171,
|
| 402 |
-
"step": 28000
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 1.2212937370101349,
|
| 406 |
-
"grad_norm": 3.6080329418182373,
|
| 407 |
-
"learning_rate": 1.9477180201414185e-05,
|
| 408 |
-
"loss": 0.8105,
|
| 409 |
-
"step": 28500
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 1.2427203188275375,
|
| 413 |
-
"grad_norm": 2.708294630050659,
|
| 414 |
-
"learning_rate": 1.894150417827298e-05,
|
| 415 |
-
"loss": 0.7808,
|
| 416 |
-
"step": 29000
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 1.2641469006449402,
|
| 420 |
-
"grad_norm": 3.300328493118286,
|
| 421 |
-
"learning_rate": 1.8405828155131776e-05,
|
| 422 |
-
"loss": 0.7774,
|
| 423 |
-
"step": 29500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 1.2855734824623428,
|
| 427 |
-
"grad_norm": 2.1943624019622803,
|
| 428 |
-
"learning_rate": 1.7870152131990572e-05,
|
| 429 |
-
"loss": 0.7637,
|
| 430 |
-
"step": 30000
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 1.3070000642797455,
|
| 434 |
-
"grad_norm": 2.7251927852630615,
|
| 435 |
-
"learning_rate": 1.733554746089565e-05,
|
| 436 |
-
"loss": 0.7481,
|
| 437 |
-
"step": 30500
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 1.3284266460971481,
|
| 441 |
-
"grad_norm": 4.672070026397705,
|
| 442 |
-
"learning_rate": 1.6799871437754446e-05,
|
| 443 |
-
"loss": 0.7255,
|
| 444 |
-
"step": 31000
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.3498532279145508,
|
| 448 |
-
"grad_norm": 2.4984779357910156,
|
| 449 |
-
"learning_rate": 1.626419541461324e-05,
|
| 450 |
-
"loss": 0.7231,
|
| 451 |
-
"step": 31500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 1.3712798097319534,
|
| 455 |
-
"grad_norm": 1.9306334257125854,
|
| 456 |
-
"learning_rate": 1.5728519391472037e-05,
|
| 457 |
-
"loss": 0.7032,
|
| 458 |
-
"step": 32000
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 1.392706391549356,
|
| 462 |
-
"grad_norm": 3.015226125717163,
|
| 463 |
-
"learning_rate": 1.5192843368330834e-05,
|
| 464 |
-
"loss": 0.6923,
|
| 465 |
-
"step": 32500
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.4141329733667587,
|
| 469 |
-
"grad_norm": 3.646979331970215,
|
| 470 |
-
"learning_rate": 1.4658238697235913e-05,
|
| 471 |
-
"loss": 0.6893,
|
| 472 |
-
"step": 33000
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 1.4355595551841613,
|
| 476 |
-
"grad_norm": 1.8951635360717773,
|
| 477 |
-
"learning_rate": 1.4122562674094708e-05,
|
| 478 |
-
"loss": 0.6693,
|
| 479 |
-
"step": 33500
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 1.4569861370015642,
|
| 483 |
-
"grad_norm": 2.3496530055999756,
|
| 484 |
-
"learning_rate": 1.3586886650953504e-05,
|
| 485 |
-
"loss": 0.6586,
|
| 486 |
-
"step": 34000
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 1.4784127188189669,
|
| 490 |
-
"grad_norm": 2.899231195449829,
|
| 491 |
-
"learning_rate": 1.30512106278123e-05,
|
| 492 |
-
"loss": 0.6338,
|
| 493 |
-
"step": 34500
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.4998393006363695,
|
| 497 |
-
"grad_norm": 2.7465250492095947,
|
| 498 |
-
"learning_rate": 1.2515534604671095e-05,
|
| 499 |
-
"loss": 0.6291,
|
| 500 |
-
"step": 35000
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.5212658824537721,
|
| 504 |
-
"grad_norm": 2.3695950508117676,
|
| 505 |
-
"learning_rate": 1.1979858581529891e-05,
|
| 506 |
-
"loss": 0.6192,
|
| 507 |
-
"step": 35500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.5426924642711748,
|
| 511 |
-
"grad_norm": 3.0361921787261963,
|
| 512 |
-
"learning_rate": 1.1444182558388687e-05,
|
| 513 |
-
"loss": 0.6147,
|
| 514 |
-
"step": 36000
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 1.5641190460885774,
|
| 518 |
-
"grad_norm": 3.5043649673461914,
|
| 519 |
-
"learning_rate": 1.0908506535247482e-05,
|
| 520 |
-
"loss": 0.6024,
|
| 521 |
-
"step": 36500
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 1.5855456279059803,
|
| 525 |
-
"grad_norm": 2.1166698932647705,
|
| 526 |
-
"learning_rate": 1.037390186415256e-05,
|
| 527 |
-
"loss": 0.6033,
|
| 528 |
-
"step": 37000
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 1.606972209723383,
|
| 532 |
-
"grad_norm": 2.8012290000915527,
|
| 533 |
-
"learning_rate": 9.838225841011358e-06,
|
| 534 |
-
"loss": 0.5935,
|
| 535 |
-
"step": 37500
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 1.6283987915407856,
|
| 539 |
-
"grad_norm": 2.49540638923645,
|
| 540 |
-
"learning_rate": 9.302549817870154e-06,
|
| 541 |
-
"loss": 0.5819,
|
| 542 |
-
"step": 38000
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 1.6498253733581882,
|
| 546 |
-
"grad_norm": 2.233430862426758,
|
| 547 |
-
"learning_rate": 8.766873794728948e-06,
|
| 548 |
-
"loss": 0.5674,
|
| 549 |
-
"step": 38500
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.6712519551755909,
|
| 553 |
-
"grad_norm": 2.823101282119751,
|
| 554 |
-
"learning_rate": 8.232269123634026e-06,
|
| 555 |
-
"loss": 0.552,
|
| 556 |
-
"step": 39000
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 1.6926785369929935,
|
| 560 |
-
"grad_norm": 2.389265775680542,
|
| 561 |
-
"learning_rate": 7.696593100492823e-06,
|
| 562 |
-
"loss": 0.5564,
|
| 563 |
-
"step": 39500
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 1.7141051188103962,
|
| 567 |
-
"grad_norm": 2.1136868000030518,
|
| 568 |
-
"learning_rate": 7.161988429397901e-06,
|
| 569 |
-
"loss": 0.5477,
|
| 570 |
-
"step": 40000
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 1.7355317006277988,
|
| 574 |
-
"grad_norm": 2.6307172775268555,
|
| 575 |
-
"learning_rate": 6.626312406256696e-06,
|
| 576 |
-
"loss": 0.545,
|
| 577 |
-
"step": 40500
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 1.7569582824452015,
|
| 581 |
-
"grad_norm": 3.4348325729370117,
|
| 582 |
-
"learning_rate": 6.090636383115492e-06,
|
| 583 |
-
"loss": 0.5359,
|
| 584 |
-
"step": 41000
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 1.7783848642626041,
|
| 588 |
-
"grad_norm": 2.1959691047668457,
|
| 589 |
-
"learning_rate": 5.554960359974288e-06,
|
| 590 |
-
"loss": 0.5267,
|
| 591 |
-
"step": 41500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 1.7998114460800068,
|
| 595 |
-
"grad_norm": 2.747068405151367,
|
| 596 |
-
"learning_rate": 5.0192843368330835e-06,
|
| 597 |
-
"loss": 0.5316,
|
| 598 |
-
"step": 42000
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 1.8212380278974094,
|
| 602 |
-
"grad_norm": 2.965233325958252,
|
| 603 |
-
"learning_rate": 4.48360831369188e-06,
|
| 604 |
-
"loss": 0.5129,
|
| 605 |
-
"step": 42500
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 1.842664609714812,
|
| 609 |
-
"grad_norm": 2.3425612449645996,
|
| 610 |
-
"learning_rate": 3.9479322905506756e-06,
|
| 611 |
-
"loss": 0.5058,
|
| 612 |
-
"step": 43000
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 1.864091191532215,
|
| 616 |
-
"grad_norm": 1.5650664567947388,
|
| 617 |
-
"learning_rate": 3.4122562674094708e-06,
|
| 618 |
-
"loss": 0.5155,
|
| 619 |
-
"step": 43500
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 1.8855177733496176,
|
| 623 |
-
"grad_norm": 2.501758098602295,
|
| 624 |
-
"learning_rate": 2.8776515963145492e-06,
|
| 625 |
-
"loss": 0.5076,
|
| 626 |
-
"step": 44000
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 1.9069443551670202,
|
| 630 |
-
"grad_norm": 4.431907653808594,
|
| 631 |
-
"learning_rate": 2.341975573173345e-06,
|
| 632 |
-
"loss": 0.5119,
|
| 633 |
-
"step": 44500
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 1.9283709369844229,
|
| 637 |
-
"grad_norm": 3.3445732593536377,
|
| 638 |
-
"learning_rate": 1.8062995500321405e-06,
|
| 639 |
-
"loss": 0.5021,
|
| 640 |
-
"step": 45000
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 1.9497975188018255,
|
| 644 |
-
"grad_norm": 2.6224589347839355,
|
| 645 |
-
"learning_rate": 1.2706235268909363e-06,
|
| 646 |
-
"loss": 0.5019,
|
| 647 |
-
"step": 45500
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 1.9712241006192284,
|
| 651 |
-
"grad_norm": 2.954880952835083,
|
| 652 |
-
"learning_rate": 7.360188557960147e-07,
|
| 653 |
-
"loss": 0.5055,
|
| 654 |
-
"step": 46000
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 1.992650682436631,
|
| 658 |
-
"grad_norm": 2.903310775756836,
|
| 659 |
-
"learning_rate": 2.003428326548104e-07,
|
| 660 |
-
"loss": 0.5043,
|
| 661 |
-
"step": 46500
|
| 662 |
-
}
|
| 663 |
-
],
|
| 664 |
-
"logging_steps": 500,
|
| 665 |
-
"max_steps": 46670,
|
| 666 |
-
"num_input_tokens_seen": 0,
|
| 667 |
-
"num_train_epochs": 2,
|
| 668 |
-
"save_steps": 500,
|
| 669 |
-
"stateful_callbacks": {
|
| 670 |
-
"TrainerControl": {
|
| 671 |
-
"args": {
|
| 672 |
-
"should_epoch_stop": false,
|
| 673 |
-
"should_evaluate": false,
|
| 674 |
-
"should_log": false,
|
| 675 |
-
"should_save": true,
|
| 676 |
-
"should_training_stop": false
|
| 677 |
-
},
|
| 678 |
-
"attributes": {}
|
| 679 |
-
}
|
| 680 |
-
},
|
| 681 |
-
"total_flos": 1.0069301454805402e+17,
|
| 682 |
-
"train_batch_size": 32,
|
| 683 |
-
"trial_name": null,
|
| 684 |
-
"trial_params": null
|
| 685 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints/checkpoint-46670/trainer_state.json
DELETED
|
@@ -1,685 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 1.9999357202545478,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 46670,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.02142658181740267,
|
| 14 |
-
"grad_norm": 0.4460393786430359,
|
| 15 |
-
"learning_rate": 4.946539532890508e-05,
|
| 16 |
-
"loss": 2.2597,
|
| 17 |
-
"step": 500
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.04285316363480534,
|
| 21 |
-
"grad_norm": 0.4649476110935211,
|
| 22 |
-
"learning_rate": 4.8929719305763877e-05,
|
| 23 |
-
"loss": 2.0158,
|
| 24 |
-
"step": 1000
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.064279745452208,
|
| 28 |
-
"grad_norm": 0.5113596320152283,
|
| 29 |
-
"learning_rate": 4.839404328262267e-05,
|
| 30 |
-
"loss": 1.9311,
|
| 31 |
-
"step": 1500
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.08570632726961068,
|
| 35 |
-
"grad_norm": 0.49711284041404724,
|
| 36 |
-
"learning_rate": 4.785836725948147e-05,
|
| 37 |
-
"loss": 1.8587,
|
| 38 |
-
"step": 2000
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.10713290908701335,
|
| 42 |
-
"grad_norm": 0.6695660352706909,
|
| 43 |
-
"learning_rate": 4.732269123634027e-05,
|
| 44 |
-
"loss": 1.8219,
|
| 45 |
-
"step": 2500
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.128559490904416,
|
| 49 |
-
"grad_norm": 0.5876463651657104,
|
| 50 |
-
"learning_rate": 4.678701521319906e-05,
|
| 51 |
-
"loss": 1.7645,
|
| 52 |
-
"step": 3000
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.1499860727218187,
|
| 56 |
-
"grad_norm": 0.554977297782898,
|
| 57 |
-
"learning_rate": 4.625133919005786e-05,
|
| 58 |
-
"loss": 1.7136,
|
| 59 |
-
"step": 3500
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.17141265453922136,
|
| 63 |
-
"grad_norm": 0.5401861667633057,
|
| 64 |
-
"learning_rate": 4.571566316691665e-05,
|
| 65 |
-
"loss": 1.696,
|
| 66 |
-
"step": 4000
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.19283923635662403,
|
| 70 |
-
"grad_norm": 0.5951809287071228,
|
| 71 |
-
"learning_rate": 4.517998714377545e-05,
|
| 72 |
-
"loss": 1.6519,
|
| 73 |
-
"step": 4500
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.2142658181740267,
|
| 77 |
-
"grad_norm": 0.6037197113037109,
|
| 78 |
-
"learning_rate": 4.464431112063424e-05,
|
| 79 |
-
"loss": 1.6279,
|
| 80 |
-
"step": 5000
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.23569239999142938,
|
| 84 |
-
"grad_norm": 0.6542425751686096,
|
| 85 |
-
"learning_rate": 4.410863509749304e-05,
|
| 86 |
-
"loss": 1.6104,
|
| 87 |
-
"step": 5500
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.257118981808832,
|
| 91 |
-
"grad_norm": 0.5781915783882141,
|
| 92 |
-
"learning_rate": 4.357295907435183e-05,
|
| 93 |
-
"loss": 1.5976,
|
| 94 |
-
"step": 6000
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.2785455636262347,
|
| 98 |
-
"grad_norm": 0.575715184211731,
|
| 99 |
-
"learning_rate": 4.303728305121063e-05,
|
| 100 |
-
"loss": 1.5373,
|
| 101 |
-
"step": 6500
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.2999721454436374,
|
| 105 |
-
"grad_norm": 0.6789396405220032,
|
| 106 |
-
"learning_rate": 4.2501607028069424e-05,
|
| 107 |
-
"loss": 1.5435,
|
| 108 |
-
"step": 7000
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.3213987272610401,
|
| 112 |
-
"grad_norm": 0.6032198667526245,
|
| 113 |
-
"learning_rate": 4.196593100492822e-05,
|
| 114 |
-
"loss": 1.5081,
|
| 115 |
-
"step": 7500
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.3428253090784427,
|
| 119 |
-
"grad_norm": 0.6690296530723572,
|
| 120 |
-
"learning_rate": 4.1430254981787015e-05,
|
| 121 |
-
"loss": 1.5059,
|
| 122 |
-
"step": 8000
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.36425189089584536,
|
| 126 |
-
"grad_norm": 0.6951475739479065,
|
| 127 |
-
"learning_rate": 4.0894578958645814e-05,
|
| 128 |
-
"loss": 1.4716,
|
| 129 |
-
"step": 8500
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.38567847271324807,
|
| 133 |
-
"grad_norm": 0.7095356583595276,
|
| 134 |
-
"learning_rate": 4.035890293550461e-05,
|
| 135 |
-
"loss": 1.455,
|
| 136 |
-
"step": 9000
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.4071050545306507,
|
| 140 |
-
"grad_norm": 0.8375953435897827,
|
| 141 |
-
"learning_rate": 3.9823226912363405e-05,
|
| 142 |
-
"loss": 1.4527,
|
| 143 |
-
"step": 9500
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.4285316363480534,
|
| 147 |
-
"grad_norm": 0.8582206964492798,
|
| 148 |
-
"learning_rate": 3.9287550889222204e-05,
|
| 149 |
-
"loss": 1.4345,
|
| 150 |
-
"step": 10000
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.44995821816545606,
|
| 154 |
-
"grad_norm": 0.8591541051864624,
|
| 155 |
-
"learning_rate": 3.8751874866081e-05,
|
| 156 |
-
"loss": 1.4268,
|
| 157 |
-
"step": 10500
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.47138479998285876,
|
| 161 |
-
"grad_norm": 0.835361897945404,
|
| 162 |
-
"learning_rate": 3.8216198842939796e-05,
|
| 163 |
-
"loss": 1.4045,
|
| 164 |
-
"step": 11000
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.4928113818002614,
|
| 168 |
-
"grad_norm": 0.9994288086891174,
|
| 169 |
-
"learning_rate": 3.768052281979859e-05,
|
| 170 |
-
"loss": 1.3965,
|
| 171 |
-
"step": 11500
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.514237963617664,
|
| 175 |
-
"grad_norm": 0.8310408592224121,
|
| 176 |
-
"learning_rate": 3.714484679665739e-05,
|
| 177 |
-
"loss": 1.3577,
|
| 178 |
-
"step": 12000
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.5356645454350667,
|
| 182 |
-
"grad_norm": 1.061630368232727,
|
| 183 |
-
"learning_rate": 3.660917077351618e-05,
|
| 184 |
-
"loss": 1.3311,
|
| 185 |
-
"step": 12500
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.5570911272524695,
|
| 189 |
-
"grad_norm": 1.0928676128387451,
|
| 190 |
-
"learning_rate": 3.607349475037498e-05,
|
| 191 |
-
"loss": 1.3203,
|
| 192 |
-
"step": 13000
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.5785177090698721,
|
| 196 |
-
"grad_norm": 1.4662648439407349,
|
| 197 |
-
"learning_rate": 3.553781872723377e-05,
|
| 198 |
-
"loss": 1.2778,
|
| 199 |
-
"step": 13500
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.5999442908872747,
|
| 203 |
-
"grad_norm": 1.2254618406295776,
|
| 204 |
-
"learning_rate": 3.500214270409257e-05,
|
| 205 |
-
"loss": 1.2572,
|
| 206 |
-
"step": 14000
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.6213708727046774,
|
| 210 |
-
"grad_norm": 1.275029182434082,
|
| 211 |
-
"learning_rate": 3.446860938504393e-05,
|
| 212 |
-
"loss": 1.2707,
|
| 213 |
-
"step": 14500
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.6427974545220801,
|
| 217 |
-
"grad_norm": 1.3197990655899048,
|
| 218 |
-
"learning_rate": 3.3932933361902726e-05,
|
| 219 |
-
"loss": 1.2308,
|
| 220 |
-
"step": 15000
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.6642240363394828,
|
| 224 |
-
"grad_norm": 1.934342861175537,
|
| 225 |
-
"learning_rate": 3.339725733876152e-05,
|
| 226 |
-
"loss": 1.2268,
|
| 227 |
-
"step": 15500
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.6856506181568854,
|
| 231 |
-
"grad_norm": 1.445302128791809,
|
| 232 |
-
"learning_rate": 3.286158131562032e-05,
|
| 233 |
-
"loss": 1.2007,
|
| 234 |
-
"step": 16000
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.7070771999742881,
|
| 238 |
-
"grad_norm": 1.2702606916427612,
|
| 239 |
-
"learning_rate": 3.232697664452539e-05,
|
| 240 |
-
"loss": 1.1785,
|
| 241 |
-
"step": 16500
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.7285037817916907,
|
| 245 |
-
"grad_norm": 1.254937767982483,
|
| 246 |
-
"learning_rate": 3.179130062138419e-05,
|
| 247 |
-
"loss": 1.1616,
|
| 248 |
-
"step": 17000
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.7499303636090935,
|
| 252 |
-
"grad_norm": 1.3628320693969727,
|
| 253 |
-
"learning_rate": 3.1255624598242984e-05,
|
| 254 |
-
"loss": 1.1352,
|
| 255 |
-
"step": 17500
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.7713569454264961,
|
| 259 |
-
"grad_norm": 3.0748207569122314,
|
| 260 |
-
"learning_rate": 3.071994857510178e-05,
|
| 261 |
-
"loss": 1.1243,
|
| 262 |
-
"step": 18000
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.7927835272438988,
|
| 266 |
-
"grad_norm": 1.4262775182724,
|
| 267 |
-
"learning_rate": 3.0185343904006858e-05,
|
| 268 |
-
"loss": 1.1011,
|
| 269 |
-
"step": 18500
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.8142101090613014,
|
| 273 |
-
"grad_norm": 2.2326557636260986,
|
| 274 |
-
"learning_rate": 2.9650739232911933e-05,
|
| 275 |
-
"loss": 1.0974,
|
| 276 |
-
"step": 19000
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.8356366908787041,
|
| 280 |
-
"grad_norm": 2.791055679321289,
|
| 281 |
-
"learning_rate": 2.9115063209770732e-05,
|
| 282 |
-
"loss": 1.0762,
|
| 283 |
-
"step": 19500
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.8570632726961068,
|
| 287 |
-
"grad_norm": 3.0886242389678955,
|
| 288 |
-
"learning_rate": 2.857938718662953e-05,
|
| 289 |
-
"loss": 1.0698,
|
| 290 |
-
"step": 20000
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.8784898545135095,
|
| 294 |
-
"grad_norm": 1.4234964847564697,
|
| 295 |
-
"learning_rate": 2.8043711163488323e-05,
|
| 296 |
-
"loss": 1.0541,
|
| 297 |
-
"step": 20500
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.8999164363309121,
|
| 301 |
-
"grad_norm": 3.275305986404419,
|
| 302 |
-
"learning_rate": 2.75091064923934e-05,
|
| 303 |
-
"loss": 1.0396,
|
| 304 |
-
"step": 21000
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.9213430181483148,
|
| 308 |
-
"grad_norm": 1.6365532875061035,
|
| 309 |
-
"learning_rate": 2.6973430469252198e-05,
|
| 310 |
-
"loss": 1.0146,
|
| 311 |
-
"step": 21500
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.9427695999657175,
|
| 315 |
-
"grad_norm": 3.1808159351348877,
|
| 316 |
-
"learning_rate": 2.6437754446110997e-05,
|
| 317 |
-
"loss": 0.9997,
|
| 318 |
-
"step": 22000
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.9641961817831202,
|
| 322 |
-
"grad_norm": 1.8644700050354004,
|
| 323 |
-
"learning_rate": 2.590207842296979e-05,
|
| 324 |
-
"loss": 0.9905,
|
| 325 |
-
"step": 22500
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.9856227636005228,
|
| 329 |
-
"grad_norm": 1.5963947772979736,
|
| 330 |
-
"learning_rate": 2.5367473751874864e-05,
|
| 331 |
-
"loss": 0.9624,
|
| 332 |
-
"step": 23000
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 1.007027918836108,
|
| 336 |
-
"grad_norm": 1.5474953651428223,
|
| 337 |
-
"learning_rate": 2.4831797728733663e-05,
|
| 338 |
-
"loss": 0.958,
|
| 339 |
-
"step": 23500
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.0284545006535109,
|
| 343 |
-
"grad_norm": 3.8564932346343994,
|
| 344 |
-
"learning_rate": 2.429612170559246e-05,
|
| 345 |
-
"loss": 0.9388,
|
| 346 |
-
"step": 24000
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.0498810824709135,
|
| 350 |
-
"grad_norm": 2.154879331588745,
|
| 351 |
-
"learning_rate": 2.3760445682451254e-05,
|
| 352 |
-
"loss": 0.9132,
|
| 353 |
-
"step": 24500
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 1.0713076642883161,
|
| 357 |
-
"grad_norm": 4.118302822113037,
|
| 358 |
-
"learning_rate": 2.3225841011356333e-05,
|
| 359 |
-
"loss": 0.9132,
|
| 360 |
-
"step": 25000
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 1.0927342461057188,
|
| 364 |
-
"grad_norm": 2.0971686840057373,
|
| 365 |
-
"learning_rate": 2.2690164988215128e-05,
|
| 366 |
-
"loss": 0.8947,
|
| 367 |
-
"step": 25500
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 1.1141608279231214,
|
| 371 |
-
"grad_norm": 3.480602741241455,
|
| 372 |
-
"learning_rate": 2.2154488965073924e-05,
|
| 373 |
-
"loss": 0.8922,
|
| 374 |
-
"step": 26000
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 1.135587409740524,
|
| 378 |
-
"grad_norm": 4.0732879638671875,
|
| 379 |
-
"learning_rate": 2.161881294193272e-05,
|
| 380 |
-
"loss": 0.87,
|
| 381 |
-
"step": 26500
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.1570139915579267,
|
| 385 |
-
"grad_norm": 2.2193567752838135,
|
| 386 |
-
"learning_rate": 2.1084208270837798e-05,
|
| 387 |
-
"loss": 0.8476,
|
| 388 |
-
"step": 27000
|
| 389 |
-
},
|
| 390 |
-
{
|
| 391 |
-
"epoch": 1.1784405733753294,
|
| 392 |
-
"grad_norm": 2.1723575592041016,
|
| 393 |
-
"learning_rate": 2.0548532247696594e-05,
|
| 394 |
-
"loss": 0.8322,
|
| 395 |
-
"step": 27500
|
| 396 |
-
},
|
| 397 |
-
{
|
| 398 |
-
"epoch": 1.199867155192732,
|
| 399 |
-
"grad_norm": 3.3503777980804443,
|
| 400 |
-
"learning_rate": 2.001285622455539e-05,
|
| 401 |
-
"loss": 0.8171,
|
| 402 |
-
"step": 28000
|
| 403 |
-
},
|
| 404 |
-
{
|
| 405 |
-
"epoch": 1.2212937370101349,
|
| 406 |
-
"grad_norm": 3.6080329418182373,
|
| 407 |
-
"learning_rate": 1.9477180201414185e-05,
|
| 408 |
-
"loss": 0.8105,
|
| 409 |
-
"step": 28500
|
| 410 |
-
},
|
| 411 |
-
{
|
| 412 |
-
"epoch": 1.2427203188275375,
|
| 413 |
-
"grad_norm": 2.708294630050659,
|
| 414 |
-
"learning_rate": 1.894150417827298e-05,
|
| 415 |
-
"loss": 0.7808,
|
| 416 |
-
"step": 29000
|
| 417 |
-
},
|
| 418 |
-
{
|
| 419 |
-
"epoch": 1.2641469006449402,
|
| 420 |
-
"grad_norm": 3.300328493118286,
|
| 421 |
-
"learning_rate": 1.8405828155131776e-05,
|
| 422 |
-
"loss": 0.7774,
|
| 423 |
-
"step": 29500
|
| 424 |
-
},
|
| 425 |
-
{
|
| 426 |
-
"epoch": 1.2855734824623428,
|
| 427 |
-
"grad_norm": 2.1943624019622803,
|
| 428 |
-
"learning_rate": 1.7870152131990572e-05,
|
| 429 |
-
"loss": 0.7637,
|
| 430 |
-
"step": 30000
|
| 431 |
-
},
|
| 432 |
-
{
|
| 433 |
-
"epoch": 1.3070000642797455,
|
| 434 |
-
"grad_norm": 2.7251927852630615,
|
| 435 |
-
"learning_rate": 1.733554746089565e-05,
|
| 436 |
-
"loss": 0.7481,
|
| 437 |
-
"step": 30500
|
| 438 |
-
},
|
| 439 |
-
{
|
| 440 |
-
"epoch": 1.3284266460971481,
|
| 441 |
-
"grad_norm": 4.672070026397705,
|
| 442 |
-
"learning_rate": 1.6799871437754446e-05,
|
| 443 |
-
"loss": 0.7255,
|
| 444 |
-
"step": 31000
|
| 445 |
-
},
|
| 446 |
-
{
|
| 447 |
-
"epoch": 1.3498532279145508,
|
| 448 |
-
"grad_norm": 2.4984779357910156,
|
| 449 |
-
"learning_rate": 1.626419541461324e-05,
|
| 450 |
-
"loss": 0.7231,
|
| 451 |
-
"step": 31500
|
| 452 |
-
},
|
| 453 |
-
{
|
| 454 |
-
"epoch": 1.3712798097319534,
|
| 455 |
-
"grad_norm": 1.9306334257125854,
|
| 456 |
-
"learning_rate": 1.5728519391472037e-05,
|
| 457 |
-
"loss": 0.7032,
|
| 458 |
-
"step": 32000
|
| 459 |
-
},
|
| 460 |
-
{
|
| 461 |
-
"epoch": 1.392706391549356,
|
| 462 |
-
"grad_norm": 3.015226125717163,
|
| 463 |
-
"learning_rate": 1.5192843368330834e-05,
|
| 464 |
-
"loss": 0.6923,
|
| 465 |
-
"step": 32500
|
| 466 |
-
},
|
| 467 |
-
{
|
| 468 |
-
"epoch": 1.4141329733667587,
|
| 469 |
-
"grad_norm": 3.646979331970215,
|
| 470 |
-
"learning_rate": 1.4658238697235913e-05,
|
| 471 |
-
"loss": 0.6893,
|
| 472 |
-
"step": 33000
|
| 473 |
-
},
|
| 474 |
-
{
|
| 475 |
-
"epoch": 1.4355595551841613,
|
| 476 |
-
"grad_norm": 1.8951635360717773,
|
| 477 |
-
"learning_rate": 1.4122562674094708e-05,
|
| 478 |
-
"loss": 0.6693,
|
| 479 |
-
"step": 33500
|
| 480 |
-
},
|
| 481 |
-
{
|
| 482 |
-
"epoch": 1.4569861370015642,
|
| 483 |
-
"grad_norm": 2.3496530055999756,
|
| 484 |
-
"learning_rate": 1.3586886650953504e-05,
|
| 485 |
-
"loss": 0.6586,
|
| 486 |
-
"step": 34000
|
| 487 |
-
},
|
| 488 |
-
{
|
| 489 |
-
"epoch": 1.4784127188189669,
|
| 490 |
-
"grad_norm": 2.899231195449829,
|
| 491 |
-
"learning_rate": 1.30512106278123e-05,
|
| 492 |
-
"loss": 0.6338,
|
| 493 |
-
"step": 34500
|
| 494 |
-
},
|
| 495 |
-
{
|
| 496 |
-
"epoch": 1.4998393006363695,
|
| 497 |
-
"grad_norm": 2.7465250492095947,
|
| 498 |
-
"learning_rate": 1.2515534604671095e-05,
|
| 499 |
-
"loss": 0.6291,
|
| 500 |
-
"step": 35000
|
| 501 |
-
},
|
| 502 |
-
{
|
| 503 |
-
"epoch": 1.5212658824537721,
|
| 504 |
-
"grad_norm": 2.3695950508117676,
|
| 505 |
-
"learning_rate": 1.1979858581529891e-05,
|
| 506 |
-
"loss": 0.6192,
|
| 507 |
-
"step": 35500
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 1.5426924642711748,
|
| 511 |
-
"grad_norm": 3.0361921787261963,
|
| 512 |
-
"learning_rate": 1.1444182558388687e-05,
|
| 513 |
-
"loss": 0.6147,
|
| 514 |
-
"step": 36000
|
| 515 |
-
},
|
| 516 |
-
{
|
| 517 |
-
"epoch": 1.5641190460885774,
|
| 518 |
-
"grad_norm": 3.5043649673461914,
|
| 519 |
-
"learning_rate": 1.0908506535247482e-05,
|
| 520 |
-
"loss": 0.6024,
|
| 521 |
-
"step": 36500
|
| 522 |
-
},
|
| 523 |
-
{
|
| 524 |
-
"epoch": 1.5855456279059803,
|
| 525 |
-
"grad_norm": 2.1166698932647705,
|
| 526 |
-
"learning_rate": 1.037390186415256e-05,
|
| 527 |
-
"loss": 0.6033,
|
| 528 |
-
"step": 37000
|
| 529 |
-
},
|
| 530 |
-
{
|
| 531 |
-
"epoch": 1.606972209723383,
|
| 532 |
-
"grad_norm": 2.8012290000915527,
|
| 533 |
-
"learning_rate": 9.838225841011358e-06,
|
| 534 |
-
"loss": 0.5935,
|
| 535 |
-
"step": 37500
|
| 536 |
-
},
|
| 537 |
-
{
|
| 538 |
-
"epoch": 1.6283987915407856,
|
| 539 |
-
"grad_norm": 2.49540638923645,
|
| 540 |
-
"learning_rate": 9.302549817870154e-06,
|
| 541 |
-
"loss": 0.5819,
|
| 542 |
-
"step": 38000
|
| 543 |
-
},
|
| 544 |
-
{
|
| 545 |
-
"epoch": 1.6498253733581882,
|
| 546 |
-
"grad_norm": 2.233430862426758,
|
| 547 |
-
"learning_rate": 8.766873794728948e-06,
|
| 548 |
-
"loss": 0.5674,
|
| 549 |
-
"step": 38500
|
| 550 |
-
},
|
| 551 |
-
{
|
| 552 |
-
"epoch": 1.6712519551755909,
|
| 553 |
-
"grad_norm": 2.823101282119751,
|
| 554 |
-
"learning_rate": 8.232269123634026e-06,
|
| 555 |
-
"loss": 0.552,
|
| 556 |
-
"step": 39000
|
| 557 |
-
},
|
| 558 |
-
{
|
| 559 |
-
"epoch": 1.6926785369929935,
|
| 560 |
-
"grad_norm": 2.389265775680542,
|
| 561 |
-
"learning_rate": 7.696593100492823e-06,
|
| 562 |
-
"loss": 0.5564,
|
| 563 |
-
"step": 39500
|
| 564 |
-
},
|
| 565 |
-
{
|
| 566 |
-
"epoch": 1.7141051188103962,
|
| 567 |
-
"grad_norm": 2.1136868000030518,
|
| 568 |
-
"learning_rate": 7.161988429397901e-06,
|
| 569 |
-
"loss": 0.5477,
|
| 570 |
-
"step": 40000
|
| 571 |
-
},
|
| 572 |
-
{
|
| 573 |
-
"epoch": 1.7355317006277988,
|
| 574 |
-
"grad_norm": 2.6307172775268555,
|
| 575 |
-
"learning_rate": 6.626312406256696e-06,
|
| 576 |
-
"loss": 0.545,
|
| 577 |
-
"step": 40500
|
| 578 |
-
},
|
| 579 |
-
{
|
| 580 |
-
"epoch": 1.7569582824452015,
|
| 581 |
-
"grad_norm": 3.4348325729370117,
|
| 582 |
-
"learning_rate": 6.090636383115492e-06,
|
| 583 |
-
"loss": 0.5359,
|
| 584 |
-
"step": 41000
|
| 585 |
-
},
|
| 586 |
-
{
|
| 587 |
-
"epoch": 1.7783848642626041,
|
| 588 |
-
"grad_norm": 2.1959691047668457,
|
| 589 |
-
"learning_rate": 5.554960359974288e-06,
|
| 590 |
-
"loss": 0.5267,
|
| 591 |
-
"step": 41500
|
| 592 |
-
},
|
| 593 |
-
{
|
| 594 |
-
"epoch": 1.7998114460800068,
|
| 595 |
-
"grad_norm": 2.747068405151367,
|
| 596 |
-
"learning_rate": 5.0192843368330835e-06,
|
| 597 |
-
"loss": 0.5316,
|
| 598 |
-
"step": 42000
|
| 599 |
-
},
|
| 600 |
-
{
|
| 601 |
-
"epoch": 1.8212380278974094,
|
| 602 |
-
"grad_norm": 2.965233325958252,
|
| 603 |
-
"learning_rate": 4.48360831369188e-06,
|
| 604 |
-
"loss": 0.5129,
|
| 605 |
-
"step": 42500
|
| 606 |
-
},
|
| 607 |
-
{
|
| 608 |
-
"epoch": 1.842664609714812,
|
| 609 |
-
"grad_norm": 2.3425612449645996,
|
| 610 |
-
"learning_rate": 3.9479322905506756e-06,
|
| 611 |
-
"loss": 0.5058,
|
| 612 |
-
"step": 43000
|
| 613 |
-
},
|
| 614 |
-
{
|
| 615 |
-
"epoch": 1.864091191532215,
|
| 616 |
-
"grad_norm": 1.5650664567947388,
|
| 617 |
-
"learning_rate": 3.4122562674094708e-06,
|
| 618 |
-
"loss": 0.5155,
|
| 619 |
-
"step": 43500
|
| 620 |
-
},
|
| 621 |
-
{
|
| 622 |
-
"epoch": 1.8855177733496176,
|
| 623 |
-
"grad_norm": 2.501758098602295,
|
| 624 |
-
"learning_rate": 2.8776515963145492e-06,
|
| 625 |
-
"loss": 0.5076,
|
| 626 |
-
"step": 44000
|
| 627 |
-
},
|
| 628 |
-
{
|
| 629 |
-
"epoch": 1.9069443551670202,
|
| 630 |
-
"grad_norm": 4.431907653808594,
|
| 631 |
-
"learning_rate": 2.341975573173345e-06,
|
| 632 |
-
"loss": 0.5119,
|
| 633 |
-
"step": 44500
|
| 634 |
-
},
|
| 635 |
-
{
|
| 636 |
-
"epoch": 1.9283709369844229,
|
| 637 |
-
"grad_norm": 3.3445732593536377,
|
| 638 |
-
"learning_rate": 1.8062995500321405e-06,
|
| 639 |
-
"loss": 0.5021,
|
| 640 |
-
"step": 45000
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 1.9497975188018255,
|
| 644 |
-
"grad_norm": 2.6224589347839355,
|
| 645 |
-
"learning_rate": 1.2706235268909363e-06,
|
| 646 |
-
"loss": 0.5019,
|
| 647 |
-
"step": 45500
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 1.9712241006192284,
|
| 651 |
-
"grad_norm": 2.954880952835083,
|
| 652 |
-
"learning_rate": 7.360188557960147e-07,
|
| 653 |
-
"loss": 0.5055,
|
| 654 |
-
"step": 46000
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 1.992650682436631,
|
| 658 |
-
"grad_norm": 2.903310775756836,
|
| 659 |
-
"learning_rate": 2.003428326548104e-07,
|
| 660 |
-
"loss": 0.5043,
|
| 661 |
-
"step": 46500
|
| 662 |
-
}
|
| 663 |
-
],
|
| 664 |
-
"logging_steps": 500,
|
| 665 |
-
"max_steps": 46670,
|
| 666 |
-
"num_input_tokens_seen": 0,
|
| 667 |
-
"num_train_epochs": 2,
|
| 668 |
-
"save_steps": 500,
|
| 669 |
-
"stateful_callbacks": {
|
| 670 |
-
"TrainerControl": {
|
| 671 |
-
"args": {
|
| 672 |
-
"should_epoch_stop": false,
|
| 673 |
-
"should_evaluate": false,
|
| 674 |
-
"should_log": false,
|
| 675 |
-
"should_save": true,
|
| 676 |
-
"should_training_stop": true
|
| 677 |
-
},
|
| 678 |
-
"attributes": {}
|
| 679 |
-
}
|
| 680 |
-
},
|
| 681 |
-
"total_flos": 1.0106114424805786e+17,
|
| 682 |
-
"train_batch_size": 32,
|
| 683 |
-
"trial_name": null,
|
| 684 |
-
"trial_params": null
|
| 685 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/generation_config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/model.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 258368552
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b030c0ed6a72d8033b32aaa2546414c98985ca56999d893f5195834f54d7a11
|
| 3 |
size 258368552
|
checkpoints/{checkpoint-46500 → checkpoint-52500}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 516816826
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8bb31247b55ae7908b4293b7181b29e83feb85d9e6a0daf0dc5782ca9ed8993
|
| 3 |
size 516816826
|
checkpoints/{checkpoint-46670 → checkpoint-52500}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac57e18a18d8d46c3f4b8eca8bfce15d75ce4c754bcfc35c83828ce60edeb2bb
|
| 3 |
size 14244
|
checkpoints/{checkpoint-46670 → checkpoint-52500}/scaler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5c5cf0b53ea935f765905ba2841a47a2a2388f189b78f81fc0c5f7d2004819f
|
| 3 |
size 988
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a8ba48807d9ae4635c45ea3f73ce208ef40ee042a6465963b590cfa72b46f89
|
| 3 |
size 1064
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/special_tokens_map.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46500 → checkpoint-52500}/spiece.model
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/tokenizer.json
RENAMED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/tokenizer_config.json
RENAMED
|
File without changes
|
checkpoints/checkpoint-52500/trainer_state.json
ADDED
|
@@ -0,0 +1,769 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9707207207207207,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 52500,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01876876876876877,
|
| 14 |
+
"grad_norm": 0.431496798992157,
|
| 15 |
+
"learning_rate": 4.953171921921922e-05,
|
| 16 |
+
"loss": 2.2319,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03753753753753754,
|
| 21 |
+
"grad_norm": 0.5443540811538696,
|
| 22 |
+
"learning_rate": 4.90625e-05,
|
| 23 |
+
"loss": 1.9685,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05630630630630631,
|
| 28 |
+
"grad_norm": 0.4891342520713806,
|
| 29 |
+
"learning_rate": 4.859328078078078e-05,
|
| 30 |
+
"loss": 1.9115,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07507507507507508,
|
| 35 |
+
"grad_norm": 0.5062255263328552,
|
| 36 |
+
"learning_rate": 4.812406156156156e-05,
|
| 37 |
+
"loss": 1.835,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09384384384384384,
|
| 42 |
+
"grad_norm": 0.5304695963859558,
|
| 43 |
+
"learning_rate": 4.765484234234234e-05,
|
| 44 |
+
"loss": 1.7728,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11261261261261261,
|
| 49 |
+
"grad_norm": 0.5526185631752014,
|
| 50 |
+
"learning_rate": 4.7185623123123124e-05,
|
| 51 |
+
"loss": 1.7251,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.1313813813813814,
|
| 56 |
+
"grad_norm": 0.5777105689048767,
|
| 57 |
+
"learning_rate": 4.6716403903903905e-05,
|
| 58 |
+
"loss": 1.6773,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.15015015015015015,
|
| 63 |
+
"grad_norm": 0.49299120903015137,
|
| 64 |
+
"learning_rate": 4.6247184684684686e-05,
|
| 65 |
+
"loss": 1.653,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16891891891891891,
|
| 70 |
+
"grad_norm": 0.5166372060775757,
|
| 71 |
+
"learning_rate": 4.577796546546547e-05,
|
| 72 |
+
"loss": 1.6323,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.18768768768768768,
|
| 77 |
+
"grad_norm": 0.6083309054374695,
|
| 78 |
+
"learning_rate": 4.530874624624625e-05,
|
| 79 |
+
"loss": 1.5978,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.20645645645645647,
|
| 84 |
+
"grad_norm": 0.6018310189247131,
|
| 85 |
+
"learning_rate": 4.483952702702703e-05,
|
| 86 |
+
"loss": 1.5637,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22522522522522523,
|
| 91 |
+
"grad_norm": 0.6258436441421509,
|
| 92 |
+
"learning_rate": 4.437030780780781e-05,
|
| 93 |
+
"loss": 1.5415,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.243993993993994,
|
| 98 |
+
"grad_norm": 0.589625895023346,
|
| 99 |
+
"learning_rate": 4.390108858858859e-05,
|
| 100 |
+
"loss": 1.5393,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2627627627627628,
|
| 105 |
+
"grad_norm": 0.714053213596344,
|
| 106 |
+
"learning_rate": 4.343186936936937e-05,
|
| 107 |
+
"loss": 1.5081,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.28153153153153154,
|
| 112 |
+
"grad_norm": 0.6066734194755554,
|
| 113 |
+
"learning_rate": 4.2962650150150154e-05,
|
| 114 |
+
"loss": 1.5088,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3003003003003003,
|
| 119 |
+
"grad_norm": 0.6267812848091125,
|
| 120 |
+
"learning_rate": 4.2493430930930935e-05,
|
| 121 |
+
"loss": 1.4817,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.31906906906906907,
|
| 126 |
+
"grad_norm": 0.78885817527771,
|
| 127 |
+
"learning_rate": 4.2024211711711716e-05,
|
| 128 |
+
"loss": 1.4647,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33783783783783783,
|
| 133 |
+
"grad_norm": 0.7244852185249329,
|
| 134 |
+
"learning_rate": 4.15549924924925e-05,
|
| 135 |
+
"loss": 1.4486,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3566066066066066,
|
| 140 |
+
"grad_norm": 0.8008115291595459,
|
| 141 |
+
"learning_rate": 4.108577327327328e-05,
|
| 142 |
+
"loss": 1.4265,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.37537537537537535,
|
| 147 |
+
"grad_norm": 0.8214899301528931,
|
| 148 |
+
"learning_rate": 4.061655405405406e-05,
|
| 149 |
+
"loss": 1.3988,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.39414414414414417,
|
| 154 |
+
"grad_norm": 0.824074387550354,
|
| 155 |
+
"learning_rate": 4.014733483483484e-05,
|
| 156 |
+
"loss": 1.3838,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.41291291291291293,
|
| 161 |
+
"grad_norm": 0.957482099533081,
|
| 162 |
+
"learning_rate": 3.967811561561562e-05,
|
| 163 |
+
"loss": 1.3667,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4316816816816817,
|
| 168 |
+
"grad_norm": 0.9371333718299866,
|
| 169 |
+
"learning_rate": 3.92088963963964e-05,
|
| 170 |
+
"loss": 1.3309,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.45045045045045046,
|
| 175 |
+
"grad_norm": 1.1298383474349976,
|
| 176 |
+
"learning_rate": 3.873967717717718e-05,
|
| 177 |
+
"loss": 1.3116,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4692192192192192,
|
| 182 |
+
"grad_norm": 0.9902416467666626,
|
| 183 |
+
"learning_rate": 3.8270457957957964e-05,
|
| 184 |
+
"loss": 1.2831,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.487987987987988,
|
| 189 |
+
"grad_norm": 1.0585743188858032,
|
| 190 |
+
"learning_rate": 3.7801238738738745e-05,
|
| 191 |
+
"loss": 1.2783,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5067567567567568,
|
| 196 |
+
"grad_norm": 1.0553358793258667,
|
| 197 |
+
"learning_rate": 3.733295795795796e-05,
|
| 198 |
+
"loss": 1.2451,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5255255255255256,
|
| 203 |
+
"grad_norm": 1.130255103111267,
|
| 204 |
+
"learning_rate": 3.686373873873874e-05,
|
| 205 |
+
"loss": 1.2277,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5442942942942943,
|
| 210 |
+
"grad_norm": 1.5941433906555176,
|
| 211 |
+
"learning_rate": 3.6394519519519524e-05,
|
| 212 |
+
"loss": 1.1952,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5630630630630631,
|
| 217 |
+
"grad_norm": 1.3499375581741333,
|
| 218 |
+
"learning_rate": 3.592623873873874e-05,
|
| 219 |
+
"loss": 1.2071,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5818318318318318,
|
| 224 |
+
"grad_norm": 1.6456067562103271,
|
| 225 |
+
"learning_rate": 3.545701951951952e-05,
|
| 226 |
+
"loss": 1.175,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6006006006006006,
|
| 231 |
+
"grad_norm": 1.5855329036712646,
|
| 232 |
+
"learning_rate": 3.49878003003003e-05,
|
| 233 |
+
"loss": 1.1348,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6193693693693694,
|
| 238 |
+
"grad_norm": 1.4513354301452637,
|
| 239 |
+
"learning_rate": 3.4518581081081084e-05,
|
| 240 |
+
"loss": 1.1411,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6381381381381381,
|
| 245 |
+
"grad_norm": 1.5460500717163086,
|
| 246 |
+
"learning_rate": 3.4049361861861865e-05,
|
| 247 |
+
"loss": 1.118,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6569069069069069,
|
| 252 |
+
"grad_norm": 1.18765389919281,
|
| 253 |
+
"learning_rate": 3.358108108108108e-05,
|
| 254 |
+
"loss": 1.0947,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6756756756756757,
|
| 259 |
+
"grad_norm": 2.74556040763855,
|
| 260 |
+
"learning_rate": 3.311186186186186e-05,
|
| 261 |
+
"loss": 1.0946,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6944444444444444,
|
| 266 |
+
"grad_norm": 1.7159161567687988,
|
| 267 |
+
"learning_rate": 3.264264264264264e-05,
|
| 268 |
+
"loss": 1.0557,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7132132132132132,
|
| 273 |
+
"grad_norm": 1.7812073230743408,
|
| 274 |
+
"learning_rate": 3.217436186186186e-05,
|
| 275 |
+
"loss": 1.0353,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7319819819819819,
|
| 280 |
+
"grad_norm": 1.616114854812622,
|
| 281 |
+
"learning_rate": 3.170514264264264e-05,
|
| 282 |
+
"loss": 1.0303,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7507507507507507,
|
| 287 |
+
"grad_norm": 1.718697190284729,
|
| 288 |
+
"learning_rate": 3.123592342342342e-05,
|
| 289 |
+
"loss": 1.0243,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7695195195195195,
|
| 294 |
+
"grad_norm": 1.679237961769104,
|
| 295 |
+
"learning_rate": 3.07667042042042e-05,
|
| 296 |
+
"loss": 0.9882,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7882882882882883,
|
| 301 |
+
"grad_norm": 2.228395938873291,
|
| 302 |
+
"learning_rate": 3.0298423423423423e-05,
|
| 303 |
+
"loss": 0.9913,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8070570570570571,
|
| 308 |
+
"grad_norm": 4.01596736907959,
|
| 309 |
+
"learning_rate": 2.9829204204204204e-05,
|
| 310 |
+
"loss": 0.9756,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8258258258258259,
|
| 315 |
+
"grad_norm": 2.9847664833068848,
|
| 316 |
+
"learning_rate": 2.9359984984984985e-05,
|
| 317 |
+
"loss": 0.9362,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8445945945945946,
|
| 322 |
+
"grad_norm": 2.6638920307159424,
|
| 323 |
+
"learning_rate": 2.8890765765765766e-05,
|
| 324 |
+
"loss": 0.9295,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8633633633633634,
|
| 329 |
+
"grad_norm": 2.0596117973327637,
|
| 330 |
+
"learning_rate": 2.8423423423423425e-05,
|
| 331 |
+
"loss": 0.9021,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8821321321321322,
|
| 336 |
+
"grad_norm": 2.9540657997131348,
|
| 337 |
+
"learning_rate": 2.7954204204204206e-05,
|
| 338 |
+
"loss": 0.8909,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9009009009009009,
|
| 343 |
+
"grad_norm": 3.7612063884735107,
|
| 344 |
+
"learning_rate": 2.7484984984984983e-05,
|
| 345 |
+
"loss": 0.8637,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9196696696696697,
|
| 350 |
+
"grad_norm": 4.234411716461182,
|
| 351 |
+
"learning_rate": 2.7015765765765764e-05,
|
| 352 |
+
"loss": 0.8465,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9384384384384384,
|
| 357 |
+
"grad_norm": 2.0504462718963623,
|
| 358 |
+
"learning_rate": 2.6547484984984988e-05,
|
| 359 |
+
"loss": 0.82,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9572072072072072,
|
| 364 |
+
"grad_norm": 4.168994426727295,
|
| 365 |
+
"learning_rate": 2.607826576576577e-05,
|
| 366 |
+
"loss": 0.8003,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.975975975975976,
|
| 371 |
+
"grad_norm": 1.9281890392303467,
|
| 372 |
+
"learning_rate": 2.560904654654655e-05,
|
| 373 |
+
"loss": 0.7895,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9947447447447447,
|
| 378 |
+
"grad_norm": 2.6740379333496094,
|
| 379 |
+
"learning_rate": 2.513982732732733e-05,
|
| 380 |
+
"loss": 0.772,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0135135135135136,
|
| 385 |
+
"grad_norm": 3.448998212814331,
|
| 386 |
+
"learning_rate": 2.4671546546546547e-05,
|
| 387 |
+
"loss": 0.7612,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0322822822822824,
|
| 392 |
+
"grad_norm": 1.8563182353973389,
|
| 393 |
+
"learning_rate": 2.420232732732733e-05,
|
| 394 |
+
"loss": 0.7195,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0510510510510511,
|
| 399 |
+
"grad_norm": 2.4365921020507812,
|
| 400 |
+
"learning_rate": 2.373310810810811e-05,
|
| 401 |
+
"loss": 0.6996,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0698198198198199,
|
| 406 |
+
"grad_norm": 2.343167781829834,
|
| 407 |
+
"learning_rate": 2.326388888888889e-05,
|
| 408 |
+
"loss": 0.6947,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0885885885885886,
|
| 413 |
+
"grad_norm": 3.7776143550872803,
|
| 414 |
+
"learning_rate": 2.279560810810811e-05,
|
| 415 |
+
"loss": 0.6685,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1073573573573574,
|
| 420 |
+
"grad_norm": 6.413450717926025,
|
| 421 |
+
"learning_rate": 2.232638888888889e-05,
|
| 422 |
+
"loss": 0.6401,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1261261261261262,
|
| 427 |
+
"grad_norm": 2.64066743850708,
|
| 428 |
+
"learning_rate": 2.1857169669669672e-05,
|
| 429 |
+
"loss": 0.6276,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.144894894894895,
|
| 434 |
+
"grad_norm": 1.6746728420257568,
|
| 435 |
+
"learning_rate": 2.138795045045045e-05,
|
| 436 |
+
"loss": 0.6072,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1636636636636637,
|
| 441 |
+
"grad_norm": 5.016355514526367,
|
| 442 |
+
"learning_rate": 2.091873123123123e-05,
|
| 443 |
+
"loss": 0.5864,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1824324324324325,
|
| 448 |
+
"grad_norm": 2.3659191131591797,
|
| 449 |
+
"learning_rate": 2.0449512012012012e-05,
|
| 450 |
+
"loss": 0.57,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2012012012012012,
|
| 455 |
+
"grad_norm": 2.68090558052063,
|
| 456 |
+
"learning_rate": 1.9981231231231232e-05,
|
| 457 |
+
"loss": 0.5612,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.21996996996997,
|
| 462 |
+
"grad_norm": 3.082831621170044,
|
| 463 |
+
"learning_rate": 1.9512012012012013e-05,
|
| 464 |
+
"loss": 0.5476,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2387387387387387,
|
| 469 |
+
"grad_norm": 2.6346185207366943,
|
| 470 |
+
"learning_rate": 1.9042792792792794e-05,
|
| 471 |
+
"loss": 0.5342,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2575075075075075,
|
| 476 |
+
"grad_norm": 2.0800909996032715,
|
| 477 |
+
"learning_rate": 1.8573573573573575e-05,
|
| 478 |
+
"loss": 0.5268,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2762762762762763,
|
| 483 |
+
"grad_norm": 2.693878650665283,
|
| 484 |
+
"learning_rate": 1.8105292792792795e-05,
|
| 485 |
+
"loss": 0.5074,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.295045045045045,
|
| 490 |
+
"grad_norm": 2.1206555366516113,
|
| 491 |
+
"learning_rate": 1.7636073573573576e-05,
|
| 492 |
+
"loss": 0.5003,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3138138138138138,
|
| 497 |
+
"grad_norm": 1.7897528409957886,
|
| 498 |
+
"learning_rate": 1.7166854354354357e-05,
|
| 499 |
+
"loss": 0.4749,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3325825825825826,
|
| 504 |
+
"grad_norm": 1.4757843017578125,
|
| 505 |
+
"learning_rate": 1.6697635135135138e-05,
|
| 506 |
+
"loss": 0.4688,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3513513513513513,
|
| 511 |
+
"grad_norm": 2.02535343170166,
|
| 512 |
+
"learning_rate": 1.6229354354354355e-05,
|
| 513 |
+
"loss": 0.4554,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.37012012012012,
|
| 518 |
+
"grad_norm": 1.7602412700653076,
|
| 519 |
+
"learning_rate": 1.5760135135135136e-05,
|
| 520 |
+
"loss": 0.4421,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3888888888888888,
|
| 525 |
+
"grad_norm": 1.8384273052215576,
|
| 526 |
+
"learning_rate": 1.5290915915915917e-05,
|
| 527 |
+
"loss": 0.4288,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4076576576576576,
|
| 532 |
+
"grad_norm": 2.2813003063201904,
|
| 533 |
+
"learning_rate": 1.4821696696696696e-05,
|
| 534 |
+
"loss": 0.4223,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.4264264264264264,
|
| 539 |
+
"grad_norm": 1.6081308126449585,
|
| 540 |
+
"learning_rate": 1.4352477477477477e-05,
|
| 541 |
+
"loss": 0.4116,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4451951951951951,
|
| 546 |
+
"grad_norm": 1.9083577394485474,
|
| 547 |
+
"learning_rate": 1.3884196696696697e-05,
|
| 548 |
+
"loss": 0.3985,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4639639639639639,
|
| 553 |
+
"grad_norm": 2.3188250064849854,
|
| 554 |
+
"learning_rate": 1.3414977477477478e-05,
|
| 555 |
+
"loss": 0.3967,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4827327327327327,
|
| 560 |
+
"grad_norm": 2.214327096939087,
|
| 561 |
+
"learning_rate": 1.294575825825826e-05,
|
| 562 |
+
"loss": 0.3906,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5015015015015014,
|
| 567 |
+
"grad_norm": 1.9394819736480713,
|
| 568 |
+
"learning_rate": 1.247653903903904e-05,
|
| 569 |
+
"loss": 0.3749,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5202702702702702,
|
| 574 |
+
"grad_norm": 2.485877275466919,
|
| 575 |
+
"learning_rate": 1.2008258258258259e-05,
|
| 576 |
+
"loss": 0.3724,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.539039039039039,
|
| 581 |
+
"grad_norm": 1.7980527877807617,
|
| 582 |
+
"learning_rate": 1.153903903903904e-05,
|
| 583 |
+
"loss": 0.3657,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5578078078078077,
|
| 588 |
+
"grad_norm": 2.025686740875244,
|
| 589 |
+
"learning_rate": 1.1069819819819819e-05,
|
| 590 |
+
"loss": 0.3545,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5765765765765765,
|
| 595 |
+
"grad_norm": 1.5588300228118896,
|
| 596 |
+
"learning_rate": 1.06006006006006e-05,
|
| 597 |
+
"loss": 0.3497,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5953453453453452,
|
| 602 |
+
"grad_norm": 2.663644790649414,
|
| 603 |
+
"learning_rate": 1.013231981981982e-05,
|
| 604 |
+
"loss": 0.35,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.614114114114114,
|
| 609 |
+
"grad_norm": 2.828707695007324,
|
| 610 |
+
"learning_rate": 9.663100600600601e-06,
|
| 611 |
+
"loss": 0.3412,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6328828828828827,
|
| 616 |
+
"grad_norm": 1.8409276008605957,
|
| 617 |
+
"learning_rate": 9.193881381381382e-06,
|
| 618 |
+
"loss": 0.3326,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6516516516516515,
|
| 623 |
+
"grad_norm": 1.8295179605484009,
|
| 624 |
+
"learning_rate": 8.724662162162163e-06,
|
| 625 |
+
"loss": 0.3315,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6704204204204203,
|
| 630 |
+
"grad_norm": 1.9252082109451294,
|
| 631 |
+
"learning_rate": 8.255442942942944e-06,
|
| 632 |
+
"loss": 0.3264,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.689189189189189,
|
| 637 |
+
"grad_norm": 2.1823854446411133,
|
| 638 |
+
"learning_rate": 7.786223723723723e-06,
|
| 639 |
+
"loss": 0.3239,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7079579579579578,
|
| 644 |
+
"grad_norm": 1.524289846420288,
|
| 645 |
+
"learning_rate": 7.3170045045045045e-06,
|
| 646 |
+
"loss": 0.3128,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7267267267267268,
|
| 651 |
+
"grad_norm": 1.5701310634613037,
|
| 652 |
+
"learning_rate": 6.8477852852852856e-06,
|
| 653 |
+
"loss": 0.3098,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7454954954954955,
|
| 658 |
+
"grad_norm": 2.198347330093384,
|
| 659 |
+
"learning_rate": 6.378566066066067e-06,
|
| 660 |
+
"loss": 0.313,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7642642642642643,
|
| 665 |
+
"grad_norm": 1.6907720565795898,
|
| 666 |
+
"learning_rate": 5.911223723723724e-06,
|
| 667 |
+
"loss": 0.3051,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.783033033033033,
|
| 672 |
+
"grad_norm": 2.1231484413146973,
|
| 673 |
+
"learning_rate": 5.442004504504505e-06,
|
| 674 |
+
"loss": 0.3069,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.8018018018018018,
|
| 679 |
+
"grad_norm": 2.2364184856414795,
|
| 680 |
+
"learning_rate": 4.972785285285286e-06,
|
| 681 |
+
"loss": 0.298,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8205705705705706,
|
| 686 |
+
"grad_norm": 2.3122572898864746,
|
| 687 |
+
"learning_rate": 4.503566066066067e-06,
|
| 688 |
+
"loss": 0.2982,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8393393393393394,
|
| 693 |
+
"grad_norm": 1.4891096353530884,
|
| 694 |
+
"learning_rate": 4.034346846846847e-06,
|
| 695 |
+
"loss": 0.2924,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8581081081081081,
|
| 700 |
+
"grad_norm": 1.6942968368530273,
|
| 701 |
+
"learning_rate": 3.5651276276276276e-06,
|
| 702 |
+
"loss": 0.2903,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8768768768768769,
|
| 707 |
+
"grad_norm": 1.5200239419937134,
|
| 708 |
+
"learning_rate": 3.096846846846847e-06,
|
| 709 |
+
"loss": 0.2914,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8956456456456456,
|
| 714 |
+
"grad_norm": 1.5315965414047241,
|
| 715 |
+
"learning_rate": 2.6276276276276277e-06,
|
| 716 |
+
"loss": 0.286,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9144144144144144,
|
| 721 |
+
"grad_norm": 1.921713948249817,
|
| 722 |
+
"learning_rate": 2.1584084084084083e-06,
|
| 723 |
+
"loss": 0.2841,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9331831831831832,
|
| 728 |
+
"grad_norm": 1.4052722454071045,
|
| 729 |
+
"learning_rate": 1.6891891891891894e-06,
|
| 730 |
+
"loss": 0.2916,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.951951951951952,
|
| 735 |
+
"grad_norm": 1.8074671030044556,
|
| 736 |
+
"learning_rate": 1.2209084084084084e-06,
|
| 737 |
+
"loss": 0.2806,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9707207207207207,
|
| 742 |
+
"grad_norm": 1.321871042251587,
|
| 743 |
+
"learning_rate": 7.516891891891893e-07,
|
| 744 |
+
"loss": 0.2801,
|
| 745 |
+
"step": 52500
|
| 746 |
+
}
|
| 747 |
+
],
|
| 748 |
+
"logging_steps": 500,
|
| 749 |
+
"max_steps": 53280,
|
| 750 |
+
"num_input_tokens_seen": 0,
|
| 751 |
+
"num_train_epochs": 2,
|
| 752 |
+
"save_steps": 500,
|
| 753 |
+
"stateful_callbacks": {
|
| 754 |
+
"TrainerControl": {
|
| 755 |
+
"args": {
|
| 756 |
+
"should_epoch_stop": false,
|
| 757 |
+
"should_evaluate": false,
|
| 758 |
+
"should_log": false,
|
| 759 |
+
"should_save": true,
|
| 760 |
+
"should_training_stop": false
|
| 761 |
+
},
|
| 762 |
+
"attributes": {}
|
| 763 |
+
}
|
| 764 |
+
},
|
| 765 |
+
"total_flos": 1.1368640269202227e+17,
|
| 766 |
+
"train_batch_size": 32,
|
| 767 |
+
"trial_name": null,
|
| 768 |
+
"trial_params": null
|
| 769 |
+
}
|
checkpoints/{checkpoint-46000 → checkpoint-52500}/training_args.bin
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/generation_config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/model.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 258368552
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa79af008d93bbef586f127e31322218dca89bc7d5ce02d5fb46f36c3371e2d1
|
| 3 |
size 258368552
|
checkpoints/{checkpoint-46000 → checkpoint-53000}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 516816826
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2c07699081386fc1f316496592ec95a86ad4a9a99b29a75fcd3253918bb431d3
|
| 3 |
size 516816826
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:240df16f6ae9eafc62c9e7b4c1be0d60f43684a3916b221610f7555a5c40ce0c
|
| 3 |
size 14244
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/scaler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43115ce28d85adc3bdae70505e0fe690c1f0e11c01abd6864f113bad760bdd12
|
| 3 |
size 988
|
checkpoints/{checkpoint-46670 → checkpoint-53000}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7f390b883ecc2cc4189fe182e8d7ef8863346ed35e8d1114e10348f6d66c390
|
| 3 |
size 1064
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/special_tokens_map.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46000 → checkpoint-53000}/spiece.model
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/tokenizer.json
RENAMED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/tokenizer_config.json
RENAMED
|
File without changes
|
checkpoints/checkpoint-53000/trainer_state.json
ADDED
|
@@ -0,0 +1,776 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9894894894894894,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 53000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01876876876876877,
|
| 14 |
+
"grad_norm": 0.431496798992157,
|
| 15 |
+
"learning_rate": 4.953171921921922e-05,
|
| 16 |
+
"loss": 2.2319,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03753753753753754,
|
| 21 |
+
"grad_norm": 0.5443540811538696,
|
| 22 |
+
"learning_rate": 4.90625e-05,
|
| 23 |
+
"loss": 1.9685,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05630630630630631,
|
| 28 |
+
"grad_norm": 0.4891342520713806,
|
| 29 |
+
"learning_rate": 4.859328078078078e-05,
|
| 30 |
+
"loss": 1.9115,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07507507507507508,
|
| 35 |
+
"grad_norm": 0.5062255263328552,
|
| 36 |
+
"learning_rate": 4.812406156156156e-05,
|
| 37 |
+
"loss": 1.835,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09384384384384384,
|
| 42 |
+
"grad_norm": 0.5304695963859558,
|
| 43 |
+
"learning_rate": 4.765484234234234e-05,
|
| 44 |
+
"loss": 1.7728,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11261261261261261,
|
| 49 |
+
"grad_norm": 0.5526185631752014,
|
| 50 |
+
"learning_rate": 4.7185623123123124e-05,
|
| 51 |
+
"loss": 1.7251,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.1313813813813814,
|
| 56 |
+
"grad_norm": 0.5777105689048767,
|
| 57 |
+
"learning_rate": 4.6716403903903905e-05,
|
| 58 |
+
"loss": 1.6773,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.15015015015015015,
|
| 63 |
+
"grad_norm": 0.49299120903015137,
|
| 64 |
+
"learning_rate": 4.6247184684684686e-05,
|
| 65 |
+
"loss": 1.653,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16891891891891891,
|
| 70 |
+
"grad_norm": 0.5166372060775757,
|
| 71 |
+
"learning_rate": 4.577796546546547e-05,
|
| 72 |
+
"loss": 1.6323,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.18768768768768768,
|
| 77 |
+
"grad_norm": 0.6083309054374695,
|
| 78 |
+
"learning_rate": 4.530874624624625e-05,
|
| 79 |
+
"loss": 1.5978,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.20645645645645647,
|
| 84 |
+
"grad_norm": 0.6018310189247131,
|
| 85 |
+
"learning_rate": 4.483952702702703e-05,
|
| 86 |
+
"loss": 1.5637,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22522522522522523,
|
| 91 |
+
"grad_norm": 0.6258436441421509,
|
| 92 |
+
"learning_rate": 4.437030780780781e-05,
|
| 93 |
+
"loss": 1.5415,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.243993993993994,
|
| 98 |
+
"grad_norm": 0.589625895023346,
|
| 99 |
+
"learning_rate": 4.390108858858859e-05,
|
| 100 |
+
"loss": 1.5393,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2627627627627628,
|
| 105 |
+
"grad_norm": 0.714053213596344,
|
| 106 |
+
"learning_rate": 4.343186936936937e-05,
|
| 107 |
+
"loss": 1.5081,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.28153153153153154,
|
| 112 |
+
"grad_norm": 0.6066734194755554,
|
| 113 |
+
"learning_rate": 4.2962650150150154e-05,
|
| 114 |
+
"loss": 1.5088,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3003003003003003,
|
| 119 |
+
"grad_norm": 0.6267812848091125,
|
| 120 |
+
"learning_rate": 4.2493430930930935e-05,
|
| 121 |
+
"loss": 1.4817,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.31906906906906907,
|
| 126 |
+
"grad_norm": 0.78885817527771,
|
| 127 |
+
"learning_rate": 4.2024211711711716e-05,
|
| 128 |
+
"loss": 1.4647,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33783783783783783,
|
| 133 |
+
"grad_norm": 0.7244852185249329,
|
| 134 |
+
"learning_rate": 4.15549924924925e-05,
|
| 135 |
+
"loss": 1.4486,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3566066066066066,
|
| 140 |
+
"grad_norm": 0.8008115291595459,
|
| 141 |
+
"learning_rate": 4.108577327327328e-05,
|
| 142 |
+
"loss": 1.4265,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.37537537537537535,
|
| 147 |
+
"grad_norm": 0.8214899301528931,
|
| 148 |
+
"learning_rate": 4.061655405405406e-05,
|
| 149 |
+
"loss": 1.3988,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.39414414414414417,
|
| 154 |
+
"grad_norm": 0.824074387550354,
|
| 155 |
+
"learning_rate": 4.014733483483484e-05,
|
| 156 |
+
"loss": 1.3838,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.41291291291291293,
|
| 161 |
+
"grad_norm": 0.957482099533081,
|
| 162 |
+
"learning_rate": 3.967811561561562e-05,
|
| 163 |
+
"loss": 1.3667,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4316816816816817,
|
| 168 |
+
"grad_norm": 0.9371333718299866,
|
| 169 |
+
"learning_rate": 3.92088963963964e-05,
|
| 170 |
+
"loss": 1.3309,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.45045045045045046,
|
| 175 |
+
"grad_norm": 1.1298383474349976,
|
| 176 |
+
"learning_rate": 3.873967717717718e-05,
|
| 177 |
+
"loss": 1.3116,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4692192192192192,
|
| 182 |
+
"grad_norm": 0.9902416467666626,
|
| 183 |
+
"learning_rate": 3.8270457957957964e-05,
|
| 184 |
+
"loss": 1.2831,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.487987987987988,
|
| 189 |
+
"grad_norm": 1.0585743188858032,
|
| 190 |
+
"learning_rate": 3.7801238738738745e-05,
|
| 191 |
+
"loss": 1.2783,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5067567567567568,
|
| 196 |
+
"grad_norm": 1.0553358793258667,
|
| 197 |
+
"learning_rate": 3.733295795795796e-05,
|
| 198 |
+
"loss": 1.2451,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5255255255255256,
|
| 203 |
+
"grad_norm": 1.130255103111267,
|
| 204 |
+
"learning_rate": 3.686373873873874e-05,
|
| 205 |
+
"loss": 1.2277,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5442942942942943,
|
| 210 |
+
"grad_norm": 1.5941433906555176,
|
| 211 |
+
"learning_rate": 3.6394519519519524e-05,
|
| 212 |
+
"loss": 1.1952,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5630630630630631,
|
| 217 |
+
"grad_norm": 1.3499375581741333,
|
| 218 |
+
"learning_rate": 3.592623873873874e-05,
|
| 219 |
+
"loss": 1.2071,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5818318318318318,
|
| 224 |
+
"grad_norm": 1.6456067562103271,
|
| 225 |
+
"learning_rate": 3.545701951951952e-05,
|
| 226 |
+
"loss": 1.175,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6006006006006006,
|
| 231 |
+
"grad_norm": 1.5855329036712646,
|
| 232 |
+
"learning_rate": 3.49878003003003e-05,
|
| 233 |
+
"loss": 1.1348,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6193693693693694,
|
| 238 |
+
"grad_norm": 1.4513354301452637,
|
| 239 |
+
"learning_rate": 3.4518581081081084e-05,
|
| 240 |
+
"loss": 1.1411,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6381381381381381,
|
| 245 |
+
"grad_norm": 1.5460500717163086,
|
| 246 |
+
"learning_rate": 3.4049361861861865e-05,
|
| 247 |
+
"loss": 1.118,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6569069069069069,
|
| 252 |
+
"grad_norm": 1.18765389919281,
|
| 253 |
+
"learning_rate": 3.358108108108108e-05,
|
| 254 |
+
"loss": 1.0947,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6756756756756757,
|
| 259 |
+
"grad_norm": 2.74556040763855,
|
| 260 |
+
"learning_rate": 3.311186186186186e-05,
|
| 261 |
+
"loss": 1.0946,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6944444444444444,
|
| 266 |
+
"grad_norm": 1.7159161567687988,
|
| 267 |
+
"learning_rate": 3.264264264264264e-05,
|
| 268 |
+
"loss": 1.0557,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7132132132132132,
|
| 273 |
+
"grad_norm": 1.7812073230743408,
|
| 274 |
+
"learning_rate": 3.217436186186186e-05,
|
| 275 |
+
"loss": 1.0353,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7319819819819819,
|
| 280 |
+
"grad_norm": 1.616114854812622,
|
| 281 |
+
"learning_rate": 3.170514264264264e-05,
|
| 282 |
+
"loss": 1.0303,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7507507507507507,
|
| 287 |
+
"grad_norm": 1.718697190284729,
|
| 288 |
+
"learning_rate": 3.123592342342342e-05,
|
| 289 |
+
"loss": 1.0243,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7695195195195195,
|
| 294 |
+
"grad_norm": 1.679237961769104,
|
| 295 |
+
"learning_rate": 3.07667042042042e-05,
|
| 296 |
+
"loss": 0.9882,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7882882882882883,
|
| 301 |
+
"grad_norm": 2.228395938873291,
|
| 302 |
+
"learning_rate": 3.0298423423423423e-05,
|
| 303 |
+
"loss": 0.9913,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8070570570570571,
|
| 308 |
+
"grad_norm": 4.01596736907959,
|
| 309 |
+
"learning_rate": 2.9829204204204204e-05,
|
| 310 |
+
"loss": 0.9756,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8258258258258259,
|
| 315 |
+
"grad_norm": 2.9847664833068848,
|
| 316 |
+
"learning_rate": 2.9359984984984985e-05,
|
| 317 |
+
"loss": 0.9362,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8445945945945946,
|
| 322 |
+
"grad_norm": 2.6638920307159424,
|
| 323 |
+
"learning_rate": 2.8890765765765766e-05,
|
| 324 |
+
"loss": 0.9295,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8633633633633634,
|
| 329 |
+
"grad_norm": 2.0596117973327637,
|
| 330 |
+
"learning_rate": 2.8423423423423425e-05,
|
| 331 |
+
"loss": 0.9021,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8821321321321322,
|
| 336 |
+
"grad_norm": 2.9540657997131348,
|
| 337 |
+
"learning_rate": 2.7954204204204206e-05,
|
| 338 |
+
"loss": 0.8909,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9009009009009009,
|
| 343 |
+
"grad_norm": 3.7612063884735107,
|
| 344 |
+
"learning_rate": 2.7484984984984983e-05,
|
| 345 |
+
"loss": 0.8637,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9196696696696697,
|
| 350 |
+
"grad_norm": 4.234411716461182,
|
| 351 |
+
"learning_rate": 2.7015765765765764e-05,
|
| 352 |
+
"loss": 0.8465,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9384384384384384,
|
| 357 |
+
"grad_norm": 2.0504462718963623,
|
| 358 |
+
"learning_rate": 2.6547484984984988e-05,
|
| 359 |
+
"loss": 0.82,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9572072072072072,
|
| 364 |
+
"grad_norm": 4.168994426727295,
|
| 365 |
+
"learning_rate": 2.607826576576577e-05,
|
| 366 |
+
"loss": 0.8003,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.975975975975976,
|
| 371 |
+
"grad_norm": 1.9281890392303467,
|
| 372 |
+
"learning_rate": 2.560904654654655e-05,
|
| 373 |
+
"loss": 0.7895,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9947447447447447,
|
| 378 |
+
"grad_norm": 2.6740379333496094,
|
| 379 |
+
"learning_rate": 2.513982732732733e-05,
|
| 380 |
+
"loss": 0.772,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0135135135135136,
|
| 385 |
+
"grad_norm": 3.448998212814331,
|
| 386 |
+
"learning_rate": 2.4671546546546547e-05,
|
| 387 |
+
"loss": 0.7612,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0322822822822824,
|
| 392 |
+
"grad_norm": 1.8563182353973389,
|
| 393 |
+
"learning_rate": 2.420232732732733e-05,
|
| 394 |
+
"loss": 0.7195,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0510510510510511,
|
| 399 |
+
"grad_norm": 2.4365921020507812,
|
| 400 |
+
"learning_rate": 2.373310810810811e-05,
|
| 401 |
+
"loss": 0.6996,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0698198198198199,
|
| 406 |
+
"grad_norm": 2.343167781829834,
|
| 407 |
+
"learning_rate": 2.326388888888889e-05,
|
| 408 |
+
"loss": 0.6947,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0885885885885886,
|
| 413 |
+
"grad_norm": 3.7776143550872803,
|
| 414 |
+
"learning_rate": 2.279560810810811e-05,
|
| 415 |
+
"loss": 0.6685,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1073573573573574,
|
| 420 |
+
"grad_norm": 6.413450717926025,
|
| 421 |
+
"learning_rate": 2.232638888888889e-05,
|
| 422 |
+
"loss": 0.6401,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1261261261261262,
|
| 427 |
+
"grad_norm": 2.64066743850708,
|
| 428 |
+
"learning_rate": 2.1857169669669672e-05,
|
| 429 |
+
"loss": 0.6276,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.144894894894895,
|
| 434 |
+
"grad_norm": 1.6746728420257568,
|
| 435 |
+
"learning_rate": 2.138795045045045e-05,
|
| 436 |
+
"loss": 0.6072,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1636636636636637,
|
| 441 |
+
"grad_norm": 5.016355514526367,
|
| 442 |
+
"learning_rate": 2.091873123123123e-05,
|
| 443 |
+
"loss": 0.5864,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1824324324324325,
|
| 448 |
+
"grad_norm": 2.3659191131591797,
|
| 449 |
+
"learning_rate": 2.0449512012012012e-05,
|
| 450 |
+
"loss": 0.57,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2012012012012012,
|
| 455 |
+
"grad_norm": 2.68090558052063,
|
| 456 |
+
"learning_rate": 1.9981231231231232e-05,
|
| 457 |
+
"loss": 0.5612,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.21996996996997,
|
| 462 |
+
"grad_norm": 3.082831621170044,
|
| 463 |
+
"learning_rate": 1.9512012012012013e-05,
|
| 464 |
+
"loss": 0.5476,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2387387387387387,
|
| 469 |
+
"grad_norm": 2.6346185207366943,
|
| 470 |
+
"learning_rate": 1.9042792792792794e-05,
|
| 471 |
+
"loss": 0.5342,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2575075075075075,
|
| 476 |
+
"grad_norm": 2.0800909996032715,
|
| 477 |
+
"learning_rate": 1.8573573573573575e-05,
|
| 478 |
+
"loss": 0.5268,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2762762762762763,
|
| 483 |
+
"grad_norm": 2.693878650665283,
|
| 484 |
+
"learning_rate": 1.8105292792792795e-05,
|
| 485 |
+
"loss": 0.5074,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.295045045045045,
|
| 490 |
+
"grad_norm": 2.1206555366516113,
|
| 491 |
+
"learning_rate": 1.7636073573573576e-05,
|
| 492 |
+
"loss": 0.5003,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3138138138138138,
|
| 497 |
+
"grad_norm": 1.7897528409957886,
|
| 498 |
+
"learning_rate": 1.7166854354354357e-05,
|
| 499 |
+
"loss": 0.4749,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3325825825825826,
|
| 504 |
+
"grad_norm": 1.4757843017578125,
|
| 505 |
+
"learning_rate": 1.6697635135135138e-05,
|
| 506 |
+
"loss": 0.4688,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3513513513513513,
|
| 511 |
+
"grad_norm": 2.02535343170166,
|
| 512 |
+
"learning_rate": 1.6229354354354355e-05,
|
| 513 |
+
"loss": 0.4554,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.37012012012012,
|
| 518 |
+
"grad_norm": 1.7602412700653076,
|
| 519 |
+
"learning_rate": 1.5760135135135136e-05,
|
| 520 |
+
"loss": 0.4421,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3888888888888888,
|
| 525 |
+
"grad_norm": 1.8384273052215576,
|
| 526 |
+
"learning_rate": 1.5290915915915917e-05,
|
| 527 |
+
"loss": 0.4288,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4076576576576576,
|
| 532 |
+
"grad_norm": 2.2813003063201904,
|
| 533 |
+
"learning_rate": 1.4821696696696696e-05,
|
| 534 |
+
"loss": 0.4223,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.4264264264264264,
|
| 539 |
+
"grad_norm": 1.6081308126449585,
|
| 540 |
+
"learning_rate": 1.4352477477477477e-05,
|
| 541 |
+
"loss": 0.4116,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4451951951951951,
|
| 546 |
+
"grad_norm": 1.9083577394485474,
|
| 547 |
+
"learning_rate": 1.3884196696696697e-05,
|
| 548 |
+
"loss": 0.3985,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4639639639639639,
|
| 553 |
+
"grad_norm": 2.3188250064849854,
|
| 554 |
+
"learning_rate": 1.3414977477477478e-05,
|
| 555 |
+
"loss": 0.3967,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4827327327327327,
|
| 560 |
+
"grad_norm": 2.214327096939087,
|
| 561 |
+
"learning_rate": 1.294575825825826e-05,
|
| 562 |
+
"loss": 0.3906,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5015015015015014,
|
| 567 |
+
"grad_norm": 1.9394819736480713,
|
| 568 |
+
"learning_rate": 1.247653903903904e-05,
|
| 569 |
+
"loss": 0.3749,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5202702702702702,
|
| 574 |
+
"grad_norm": 2.485877275466919,
|
| 575 |
+
"learning_rate": 1.2008258258258259e-05,
|
| 576 |
+
"loss": 0.3724,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.539039039039039,
|
| 581 |
+
"grad_norm": 1.7980527877807617,
|
| 582 |
+
"learning_rate": 1.153903903903904e-05,
|
| 583 |
+
"loss": 0.3657,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5578078078078077,
|
| 588 |
+
"grad_norm": 2.025686740875244,
|
| 589 |
+
"learning_rate": 1.1069819819819819e-05,
|
| 590 |
+
"loss": 0.3545,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5765765765765765,
|
| 595 |
+
"grad_norm": 1.5588300228118896,
|
| 596 |
+
"learning_rate": 1.06006006006006e-05,
|
| 597 |
+
"loss": 0.3497,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5953453453453452,
|
| 602 |
+
"grad_norm": 2.663644790649414,
|
| 603 |
+
"learning_rate": 1.013231981981982e-05,
|
| 604 |
+
"loss": 0.35,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.614114114114114,
|
| 609 |
+
"grad_norm": 2.828707695007324,
|
| 610 |
+
"learning_rate": 9.663100600600601e-06,
|
| 611 |
+
"loss": 0.3412,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6328828828828827,
|
| 616 |
+
"grad_norm": 1.8409276008605957,
|
| 617 |
+
"learning_rate": 9.193881381381382e-06,
|
| 618 |
+
"loss": 0.3326,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6516516516516515,
|
| 623 |
+
"grad_norm": 1.8295179605484009,
|
| 624 |
+
"learning_rate": 8.724662162162163e-06,
|
| 625 |
+
"loss": 0.3315,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6704204204204203,
|
| 630 |
+
"grad_norm": 1.9252082109451294,
|
| 631 |
+
"learning_rate": 8.255442942942944e-06,
|
| 632 |
+
"loss": 0.3264,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.689189189189189,
|
| 637 |
+
"grad_norm": 2.1823854446411133,
|
| 638 |
+
"learning_rate": 7.786223723723723e-06,
|
| 639 |
+
"loss": 0.3239,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7079579579579578,
|
| 644 |
+
"grad_norm": 1.524289846420288,
|
| 645 |
+
"learning_rate": 7.3170045045045045e-06,
|
| 646 |
+
"loss": 0.3128,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7267267267267268,
|
| 651 |
+
"grad_norm": 1.5701310634613037,
|
| 652 |
+
"learning_rate": 6.8477852852852856e-06,
|
| 653 |
+
"loss": 0.3098,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7454954954954955,
|
| 658 |
+
"grad_norm": 2.198347330093384,
|
| 659 |
+
"learning_rate": 6.378566066066067e-06,
|
| 660 |
+
"loss": 0.313,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7642642642642643,
|
| 665 |
+
"grad_norm": 1.6907720565795898,
|
| 666 |
+
"learning_rate": 5.911223723723724e-06,
|
| 667 |
+
"loss": 0.3051,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.783033033033033,
|
| 672 |
+
"grad_norm": 2.1231484413146973,
|
| 673 |
+
"learning_rate": 5.442004504504505e-06,
|
| 674 |
+
"loss": 0.3069,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.8018018018018018,
|
| 679 |
+
"grad_norm": 2.2364184856414795,
|
| 680 |
+
"learning_rate": 4.972785285285286e-06,
|
| 681 |
+
"loss": 0.298,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8205705705705706,
|
| 686 |
+
"grad_norm": 2.3122572898864746,
|
| 687 |
+
"learning_rate": 4.503566066066067e-06,
|
| 688 |
+
"loss": 0.2982,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8393393393393394,
|
| 693 |
+
"grad_norm": 1.4891096353530884,
|
| 694 |
+
"learning_rate": 4.034346846846847e-06,
|
| 695 |
+
"loss": 0.2924,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8581081081081081,
|
| 700 |
+
"grad_norm": 1.6942968368530273,
|
| 701 |
+
"learning_rate": 3.5651276276276276e-06,
|
| 702 |
+
"loss": 0.2903,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8768768768768769,
|
| 707 |
+
"grad_norm": 1.5200239419937134,
|
| 708 |
+
"learning_rate": 3.096846846846847e-06,
|
| 709 |
+
"loss": 0.2914,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8956456456456456,
|
| 714 |
+
"grad_norm": 1.5315965414047241,
|
| 715 |
+
"learning_rate": 2.6276276276276277e-06,
|
| 716 |
+
"loss": 0.286,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9144144144144144,
|
| 721 |
+
"grad_norm": 1.921713948249817,
|
| 722 |
+
"learning_rate": 2.1584084084084083e-06,
|
| 723 |
+
"loss": 0.2841,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9331831831831832,
|
| 728 |
+
"grad_norm": 1.4052722454071045,
|
| 729 |
+
"learning_rate": 1.6891891891891894e-06,
|
| 730 |
+
"loss": 0.2916,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.951951951951952,
|
| 735 |
+
"grad_norm": 1.8074671030044556,
|
| 736 |
+
"learning_rate": 1.2209084084084084e-06,
|
| 737 |
+
"loss": 0.2806,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9707207207207207,
|
| 742 |
+
"grad_norm": 1.321871042251587,
|
| 743 |
+
"learning_rate": 7.516891891891893e-07,
|
| 744 |
+
"loss": 0.2801,
|
| 745 |
+
"step": 52500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9894894894894894,
|
| 749 |
+
"grad_norm": 1.944077491760254,
|
| 750 |
+
"learning_rate": 2.8246996996997e-07,
|
| 751 |
+
"loss": 0.2869,
|
| 752 |
+
"step": 53000
|
| 753 |
+
}
|
| 754 |
+
],
|
| 755 |
+
"logging_steps": 500,
|
| 756 |
+
"max_steps": 53280,
|
| 757 |
+
"num_input_tokens_seen": 0,
|
| 758 |
+
"num_train_epochs": 2,
|
| 759 |
+
"save_steps": 500,
|
| 760 |
+
"stateful_callbacks": {
|
| 761 |
+
"TrainerControl": {
|
| 762 |
+
"args": {
|
| 763 |
+
"should_epoch_stop": false,
|
| 764 |
+
"should_evaluate": false,
|
| 765 |
+
"should_log": false,
|
| 766 |
+
"should_save": true,
|
| 767 |
+
"should_training_stop": false
|
| 768 |
+
},
|
| 769 |
+
"attributes": {}
|
| 770 |
+
}
|
| 771 |
+
},
|
| 772 |
+
"total_flos": 1.1476913710379827e+17,
|
| 773 |
+
"train_batch_size": 32,
|
| 774 |
+
"trial_name": null,
|
| 775 |
+
"trial_params": null
|
| 776 |
+
}
|
checkpoints/{checkpoint-46500 → checkpoint-53000}/training_args.bin
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/generation_config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/model.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 258368552
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0808fe9e6ee3e9921f8fd3a2388b98e6f54d13f0c87655bb092ad05328fec645
|
| 3 |
size 258368552
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 516816826
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fdc41dd20bdf2348667a36cc4bbb4bc92ea63bdb9f005dfbcaa5a90767ad7e3a
|
| 3 |
size 516816826
|
checkpoints/{checkpoint-46000 → checkpoint-53280}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2fe5f07d61e98d2a968a39b27d61606c3d0b18ad53c2a044e3c462b092ae5845
|
| 3 |
size 14244
|
checkpoints/{checkpoint-46000 → checkpoint-53280}/scaler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 988
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b590339cd90b6f0bae6f38747b4f1234fa93fdb389c4c1a1d71c6d03865a59e
|
| 3 |
size 988
|
checkpoints/{checkpoint-46500 → checkpoint-53280}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72def428996ce73e58c7892ae80d897b71c1abe271bf6b8d6edf277a73d3d657
|
| 3 |
size 1064
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/special_tokens_map.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/spiece.model
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/tokenizer.json
RENAMED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/tokenizer_config.json
RENAMED
|
File without changes
|
checkpoints/checkpoint-53280/trainer_state.json
ADDED
|
@@ -0,0 +1,776 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 53280,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01876876876876877,
|
| 14 |
+
"grad_norm": 0.431496798992157,
|
| 15 |
+
"learning_rate": 4.953171921921922e-05,
|
| 16 |
+
"loss": 2.2319,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03753753753753754,
|
| 21 |
+
"grad_norm": 0.5443540811538696,
|
| 22 |
+
"learning_rate": 4.90625e-05,
|
| 23 |
+
"loss": 1.9685,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05630630630630631,
|
| 28 |
+
"grad_norm": 0.4891342520713806,
|
| 29 |
+
"learning_rate": 4.859328078078078e-05,
|
| 30 |
+
"loss": 1.9115,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07507507507507508,
|
| 35 |
+
"grad_norm": 0.5062255263328552,
|
| 36 |
+
"learning_rate": 4.812406156156156e-05,
|
| 37 |
+
"loss": 1.835,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09384384384384384,
|
| 42 |
+
"grad_norm": 0.5304695963859558,
|
| 43 |
+
"learning_rate": 4.765484234234234e-05,
|
| 44 |
+
"loss": 1.7728,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.11261261261261261,
|
| 49 |
+
"grad_norm": 0.5526185631752014,
|
| 50 |
+
"learning_rate": 4.7185623123123124e-05,
|
| 51 |
+
"loss": 1.7251,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.1313813813813814,
|
| 56 |
+
"grad_norm": 0.5777105689048767,
|
| 57 |
+
"learning_rate": 4.6716403903903905e-05,
|
| 58 |
+
"loss": 1.6773,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.15015015015015015,
|
| 63 |
+
"grad_norm": 0.49299120903015137,
|
| 64 |
+
"learning_rate": 4.6247184684684686e-05,
|
| 65 |
+
"loss": 1.653,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.16891891891891891,
|
| 70 |
+
"grad_norm": 0.5166372060775757,
|
| 71 |
+
"learning_rate": 4.577796546546547e-05,
|
| 72 |
+
"loss": 1.6323,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.18768768768768768,
|
| 77 |
+
"grad_norm": 0.6083309054374695,
|
| 78 |
+
"learning_rate": 4.530874624624625e-05,
|
| 79 |
+
"loss": 1.5978,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.20645645645645647,
|
| 84 |
+
"grad_norm": 0.6018310189247131,
|
| 85 |
+
"learning_rate": 4.483952702702703e-05,
|
| 86 |
+
"loss": 1.5637,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.22522522522522523,
|
| 91 |
+
"grad_norm": 0.6258436441421509,
|
| 92 |
+
"learning_rate": 4.437030780780781e-05,
|
| 93 |
+
"loss": 1.5415,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.243993993993994,
|
| 98 |
+
"grad_norm": 0.589625895023346,
|
| 99 |
+
"learning_rate": 4.390108858858859e-05,
|
| 100 |
+
"loss": 1.5393,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2627627627627628,
|
| 105 |
+
"grad_norm": 0.714053213596344,
|
| 106 |
+
"learning_rate": 4.343186936936937e-05,
|
| 107 |
+
"loss": 1.5081,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.28153153153153154,
|
| 112 |
+
"grad_norm": 0.6066734194755554,
|
| 113 |
+
"learning_rate": 4.2962650150150154e-05,
|
| 114 |
+
"loss": 1.5088,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3003003003003003,
|
| 119 |
+
"grad_norm": 0.6267812848091125,
|
| 120 |
+
"learning_rate": 4.2493430930930935e-05,
|
| 121 |
+
"loss": 1.4817,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.31906906906906907,
|
| 126 |
+
"grad_norm": 0.78885817527771,
|
| 127 |
+
"learning_rate": 4.2024211711711716e-05,
|
| 128 |
+
"loss": 1.4647,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.33783783783783783,
|
| 133 |
+
"grad_norm": 0.7244852185249329,
|
| 134 |
+
"learning_rate": 4.15549924924925e-05,
|
| 135 |
+
"loss": 1.4486,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.3566066066066066,
|
| 140 |
+
"grad_norm": 0.8008115291595459,
|
| 141 |
+
"learning_rate": 4.108577327327328e-05,
|
| 142 |
+
"loss": 1.4265,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.37537537537537535,
|
| 147 |
+
"grad_norm": 0.8214899301528931,
|
| 148 |
+
"learning_rate": 4.061655405405406e-05,
|
| 149 |
+
"loss": 1.3988,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.39414414414414417,
|
| 154 |
+
"grad_norm": 0.824074387550354,
|
| 155 |
+
"learning_rate": 4.014733483483484e-05,
|
| 156 |
+
"loss": 1.3838,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.41291291291291293,
|
| 161 |
+
"grad_norm": 0.957482099533081,
|
| 162 |
+
"learning_rate": 3.967811561561562e-05,
|
| 163 |
+
"loss": 1.3667,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.4316816816816817,
|
| 168 |
+
"grad_norm": 0.9371333718299866,
|
| 169 |
+
"learning_rate": 3.92088963963964e-05,
|
| 170 |
+
"loss": 1.3309,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.45045045045045046,
|
| 175 |
+
"grad_norm": 1.1298383474349976,
|
| 176 |
+
"learning_rate": 3.873967717717718e-05,
|
| 177 |
+
"loss": 1.3116,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.4692192192192192,
|
| 182 |
+
"grad_norm": 0.9902416467666626,
|
| 183 |
+
"learning_rate": 3.8270457957957964e-05,
|
| 184 |
+
"loss": 1.2831,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.487987987987988,
|
| 189 |
+
"grad_norm": 1.0585743188858032,
|
| 190 |
+
"learning_rate": 3.7801238738738745e-05,
|
| 191 |
+
"loss": 1.2783,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5067567567567568,
|
| 196 |
+
"grad_norm": 1.0553358793258667,
|
| 197 |
+
"learning_rate": 3.733295795795796e-05,
|
| 198 |
+
"loss": 1.2451,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5255255255255256,
|
| 203 |
+
"grad_norm": 1.130255103111267,
|
| 204 |
+
"learning_rate": 3.686373873873874e-05,
|
| 205 |
+
"loss": 1.2277,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5442942942942943,
|
| 210 |
+
"grad_norm": 1.5941433906555176,
|
| 211 |
+
"learning_rate": 3.6394519519519524e-05,
|
| 212 |
+
"loss": 1.1952,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.5630630630630631,
|
| 217 |
+
"grad_norm": 1.3499375581741333,
|
| 218 |
+
"learning_rate": 3.592623873873874e-05,
|
| 219 |
+
"loss": 1.2071,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.5818318318318318,
|
| 224 |
+
"grad_norm": 1.6456067562103271,
|
| 225 |
+
"learning_rate": 3.545701951951952e-05,
|
| 226 |
+
"loss": 1.175,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6006006006006006,
|
| 231 |
+
"grad_norm": 1.5855329036712646,
|
| 232 |
+
"learning_rate": 3.49878003003003e-05,
|
| 233 |
+
"loss": 1.1348,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6193693693693694,
|
| 238 |
+
"grad_norm": 1.4513354301452637,
|
| 239 |
+
"learning_rate": 3.4518581081081084e-05,
|
| 240 |
+
"loss": 1.1411,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6381381381381381,
|
| 245 |
+
"grad_norm": 1.5460500717163086,
|
| 246 |
+
"learning_rate": 3.4049361861861865e-05,
|
| 247 |
+
"loss": 1.118,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.6569069069069069,
|
| 252 |
+
"grad_norm": 1.18765389919281,
|
| 253 |
+
"learning_rate": 3.358108108108108e-05,
|
| 254 |
+
"loss": 1.0947,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.6756756756756757,
|
| 259 |
+
"grad_norm": 2.74556040763855,
|
| 260 |
+
"learning_rate": 3.311186186186186e-05,
|
| 261 |
+
"loss": 1.0946,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.6944444444444444,
|
| 266 |
+
"grad_norm": 1.7159161567687988,
|
| 267 |
+
"learning_rate": 3.264264264264264e-05,
|
| 268 |
+
"loss": 1.0557,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7132132132132132,
|
| 273 |
+
"grad_norm": 1.7812073230743408,
|
| 274 |
+
"learning_rate": 3.217436186186186e-05,
|
| 275 |
+
"loss": 1.0353,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7319819819819819,
|
| 280 |
+
"grad_norm": 1.616114854812622,
|
| 281 |
+
"learning_rate": 3.170514264264264e-05,
|
| 282 |
+
"loss": 1.0303,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.7507507507507507,
|
| 287 |
+
"grad_norm": 1.718697190284729,
|
| 288 |
+
"learning_rate": 3.123592342342342e-05,
|
| 289 |
+
"loss": 1.0243,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.7695195195195195,
|
| 294 |
+
"grad_norm": 1.679237961769104,
|
| 295 |
+
"learning_rate": 3.07667042042042e-05,
|
| 296 |
+
"loss": 0.9882,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.7882882882882883,
|
| 301 |
+
"grad_norm": 2.228395938873291,
|
| 302 |
+
"learning_rate": 3.0298423423423423e-05,
|
| 303 |
+
"loss": 0.9913,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.8070570570570571,
|
| 308 |
+
"grad_norm": 4.01596736907959,
|
| 309 |
+
"learning_rate": 2.9829204204204204e-05,
|
| 310 |
+
"loss": 0.9756,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.8258258258258259,
|
| 315 |
+
"grad_norm": 2.9847664833068848,
|
| 316 |
+
"learning_rate": 2.9359984984984985e-05,
|
| 317 |
+
"loss": 0.9362,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.8445945945945946,
|
| 322 |
+
"grad_norm": 2.6638920307159424,
|
| 323 |
+
"learning_rate": 2.8890765765765766e-05,
|
| 324 |
+
"loss": 0.9295,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.8633633633633634,
|
| 329 |
+
"grad_norm": 2.0596117973327637,
|
| 330 |
+
"learning_rate": 2.8423423423423425e-05,
|
| 331 |
+
"loss": 0.9021,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.8821321321321322,
|
| 336 |
+
"grad_norm": 2.9540657997131348,
|
| 337 |
+
"learning_rate": 2.7954204204204206e-05,
|
| 338 |
+
"loss": 0.8909,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9009009009009009,
|
| 343 |
+
"grad_norm": 3.7612063884735107,
|
| 344 |
+
"learning_rate": 2.7484984984984983e-05,
|
| 345 |
+
"loss": 0.8637,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9196696696696697,
|
| 350 |
+
"grad_norm": 4.234411716461182,
|
| 351 |
+
"learning_rate": 2.7015765765765764e-05,
|
| 352 |
+
"loss": 0.8465,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.9384384384384384,
|
| 357 |
+
"grad_norm": 2.0504462718963623,
|
| 358 |
+
"learning_rate": 2.6547484984984988e-05,
|
| 359 |
+
"loss": 0.82,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.9572072072072072,
|
| 364 |
+
"grad_norm": 4.168994426727295,
|
| 365 |
+
"learning_rate": 2.607826576576577e-05,
|
| 366 |
+
"loss": 0.8003,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.975975975975976,
|
| 371 |
+
"grad_norm": 1.9281890392303467,
|
| 372 |
+
"learning_rate": 2.560904654654655e-05,
|
| 373 |
+
"loss": 0.7895,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.9947447447447447,
|
| 378 |
+
"grad_norm": 2.6740379333496094,
|
| 379 |
+
"learning_rate": 2.513982732732733e-05,
|
| 380 |
+
"loss": 0.772,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0135135135135136,
|
| 385 |
+
"grad_norm": 3.448998212814331,
|
| 386 |
+
"learning_rate": 2.4671546546546547e-05,
|
| 387 |
+
"loss": 0.7612,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.0322822822822824,
|
| 392 |
+
"grad_norm": 1.8563182353973389,
|
| 393 |
+
"learning_rate": 2.420232732732733e-05,
|
| 394 |
+
"loss": 0.7195,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.0510510510510511,
|
| 399 |
+
"grad_norm": 2.4365921020507812,
|
| 400 |
+
"learning_rate": 2.373310810810811e-05,
|
| 401 |
+
"loss": 0.6996,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.0698198198198199,
|
| 406 |
+
"grad_norm": 2.343167781829834,
|
| 407 |
+
"learning_rate": 2.326388888888889e-05,
|
| 408 |
+
"loss": 0.6947,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.0885885885885886,
|
| 413 |
+
"grad_norm": 3.7776143550872803,
|
| 414 |
+
"learning_rate": 2.279560810810811e-05,
|
| 415 |
+
"loss": 0.6685,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1073573573573574,
|
| 420 |
+
"grad_norm": 6.413450717926025,
|
| 421 |
+
"learning_rate": 2.232638888888889e-05,
|
| 422 |
+
"loss": 0.6401,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.1261261261261262,
|
| 427 |
+
"grad_norm": 2.64066743850708,
|
| 428 |
+
"learning_rate": 2.1857169669669672e-05,
|
| 429 |
+
"loss": 0.6276,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.144894894894895,
|
| 434 |
+
"grad_norm": 1.6746728420257568,
|
| 435 |
+
"learning_rate": 2.138795045045045e-05,
|
| 436 |
+
"loss": 0.6072,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.1636636636636637,
|
| 441 |
+
"grad_norm": 5.016355514526367,
|
| 442 |
+
"learning_rate": 2.091873123123123e-05,
|
| 443 |
+
"loss": 0.5864,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.1824324324324325,
|
| 448 |
+
"grad_norm": 2.3659191131591797,
|
| 449 |
+
"learning_rate": 2.0449512012012012e-05,
|
| 450 |
+
"loss": 0.57,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2012012012012012,
|
| 455 |
+
"grad_norm": 2.68090558052063,
|
| 456 |
+
"learning_rate": 1.9981231231231232e-05,
|
| 457 |
+
"loss": 0.5612,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.21996996996997,
|
| 462 |
+
"grad_norm": 3.082831621170044,
|
| 463 |
+
"learning_rate": 1.9512012012012013e-05,
|
| 464 |
+
"loss": 0.5476,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.2387387387387387,
|
| 469 |
+
"grad_norm": 2.6346185207366943,
|
| 470 |
+
"learning_rate": 1.9042792792792794e-05,
|
| 471 |
+
"loss": 0.5342,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.2575075075075075,
|
| 476 |
+
"grad_norm": 2.0800909996032715,
|
| 477 |
+
"learning_rate": 1.8573573573573575e-05,
|
| 478 |
+
"loss": 0.5268,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.2762762762762763,
|
| 483 |
+
"grad_norm": 2.693878650665283,
|
| 484 |
+
"learning_rate": 1.8105292792792795e-05,
|
| 485 |
+
"loss": 0.5074,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.295045045045045,
|
| 490 |
+
"grad_norm": 2.1206555366516113,
|
| 491 |
+
"learning_rate": 1.7636073573573576e-05,
|
| 492 |
+
"loss": 0.5003,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.3138138138138138,
|
| 497 |
+
"grad_norm": 1.7897528409957886,
|
| 498 |
+
"learning_rate": 1.7166854354354357e-05,
|
| 499 |
+
"loss": 0.4749,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.3325825825825826,
|
| 504 |
+
"grad_norm": 1.4757843017578125,
|
| 505 |
+
"learning_rate": 1.6697635135135138e-05,
|
| 506 |
+
"loss": 0.4688,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.3513513513513513,
|
| 511 |
+
"grad_norm": 2.02535343170166,
|
| 512 |
+
"learning_rate": 1.6229354354354355e-05,
|
| 513 |
+
"loss": 0.4554,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.37012012012012,
|
| 518 |
+
"grad_norm": 1.7602412700653076,
|
| 519 |
+
"learning_rate": 1.5760135135135136e-05,
|
| 520 |
+
"loss": 0.4421,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.3888888888888888,
|
| 525 |
+
"grad_norm": 1.8384273052215576,
|
| 526 |
+
"learning_rate": 1.5290915915915917e-05,
|
| 527 |
+
"loss": 0.4288,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.4076576576576576,
|
| 532 |
+
"grad_norm": 2.2813003063201904,
|
| 533 |
+
"learning_rate": 1.4821696696696696e-05,
|
| 534 |
+
"loss": 0.4223,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.4264264264264264,
|
| 539 |
+
"grad_norm": 1.6081308126449585,
|
| 540 |
+
"learning_rate": 1.4352477477477477e-05,
|
| 541 |
+
"loss": 0.4116,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.4451951951951951,
|
| 546 |
+
"grad_norm": 1.9083577394485474,
|
| 547 |
+
"learning_rate": 1.3884196696696697e-05,
|
| 548 |
+
"loss": 0.3985,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.4639639639639639,
|
| 553 |
+
"grad_norm": 2.3188250064849854,
|
| 554 |
+
"learning_rate": 1.3414977477477478e-05,
|
| 555 |
+
"loss": 0.3967,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.4827327327327327,
|
| 560 |
+
"grad_norm": 2.214327096939087,
|
| 561 |
+
"learning_rate": 1.294575825825826e-05,
|
| 562 |
+
"loss": 0.3906,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.5015015015015014,
|
| 567 |
+
"grad_norm": 1.9394819736480713,
|
| 568 |
+
"learning_rate": 1.247653903903904e-05,
|
| 569 |
+
"loss": 0.3749,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.5202702702702702,
|
| 574 |
+
"grad_norm": 2.485877275466919,
|
| 575 |
+
"learning_rate": 1.2008258258258259e-05,
|
| 576 |
+
"loss": 0.3724,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.539039039039039,
|
| 581 |
+
"grad_norm": 1.7980527877807617,
|
| 582 |
+
"learning_rate": 1.153903903903904e-05,
|
| 583 |
+
"loss": 0.3657,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.5578078078078077,
|
| 588 |
+
"grad_norm": 2.025686740875244,
|
| 589 |
+
"learning_rate": 1.1069819819819819e-05,
|
| 590 |
+
"loss": 0.3545,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.5765765765765765,
|
| 595 |
+
"grad_norm": 1.5588300228118896,
|
| 596 |
+
"learning_rate": 1.06006006006006e-05,
|
| 597 |
+
"loss": 0.3497,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.5953453453453452,
|
| 602 |
+
"grad_norm": 2.663644790649414,
|
| 603 |
+
"learning_rate": 1.013231981981982e-05,
|
| 604 |
+
"loss": 0.35,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.614114114114114,
|
| 609 |
+
"grad_norm": 2.828707695007324,
|
| 610 |
+
"learning_rate": 9.663100600600601e-06,
|
| 611 |
+
"loss": 0.3412,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.6328828828828827,
|
| 616 |
+
"grad_norm": 1.8409276008605957,
|
| 617 |
+
"learning_rate": 9.193881381381382e-06,
|
| 618 |
+
"loss": 0.3326,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.6516516516516515,
|
| 623 |
+
"grad_norm": 1.8295179605484009,
|
| 624 |
+
"learning_rate": 8.724662162162163e-06,
|
| 625 |
+
"loss": 0.3315,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.6704204204204203,
|
| 630 |
+
"grad_norm": 1.9252082109451294,
|
| 631 |
+
"learning_rate": 8.255442942942944e-06,
|
| 632 |
+
"loss": 0.3264,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.689189189189189,
|
| 637 |
+
"grad_norm": 2.1823854446411133,
|
| 638 |
+
"learning_rate": 7.786223723723723e-06,
|
| 639 |
+
"loss": 0.3239,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.7079579579579578,
|
| 644 |
+
"grad_norm": 1.524289846420288,
|
| 645 |
+
"learning_rate": 7.3170045045045045e-06,
|
| 646 |
+
"loss": 0.3128,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.7267267267267268,
|
| 651 |
+
"grad_norm": 1.5701310634613037,
|
| 652 |
+
"learning_rate": 6.8477852852852856e-06,
|
| 653 |
+
"loss": 0.3098,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.7454954954954955,
|
| 658 |
+
"grad_norm": 2.198347330093384,
|
| 659 |
+
"learning_rate": 6.378566066066067e-06,
|
| 660 |
+
"loss": 0.313,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.7642642642642643,
|
| 665 |
+
"grad_norm": 1.6907720565795898,
|
| 666 |
+
"learning_rate": 5.911223723723724e-06,
|
| 667 |
+
"loss": 0.3051,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.783033033033033,
|
| 672 |
+
"grad_norm": 2.1231484413146973,
|
| 673 |
+
"learning_rate": 5.442004504504505e-06,
|
| 674 |
+
"loss": 0.3069,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.8018018018018018,
|
| 679 |
+
"grad_norm": 2.2364184856414795,
|
| 680 |
+
"learning_rate": 4.972785285285286e-06,
|
| 681 |
+
"loss": 0.298,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.8205705705705706,
|
| 686 |
+
"grad_norm": 2.3122572898864746,
|
| 687 |
+
"learning_rate": 4.503566066066067e-06,
|
| 688 |
+
"loss": 0.2982,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.8393393393393394,
|
| 693 |
+
"grad_norm": 1.4891096353530884,
|
| 694 |
+
"learning_rate": 4.034346846846847e-06,
|
| 695 |
+
"loss": 0.2924,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.8581081081081081,
|
| 700 |
+
"grad_norm": 1.6942968368530273,
|
| 701 |
+
"learning_rate": 3.5651276276276276e-06,
|
| 702 |
+
"loss": 0.2903,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.8768768768768769,
|
| 707 |
+
"grad_norm": 1.5200239419937134,
|
| 708 |
+
"learning_rate": 3.096846846846847e-06,
|
| 709 |
+
"loss": 0.2914,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.8956456456456456,
|
| 714 |
+
"grad_norm": 1.5315965414047241,
|
| 715 |
+
"learning_rate": 2.6276276276276277e-06,
|
| 716 |
+
"loss": 0.286,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.9144144144144144,
|
| 721 |
+
"grad_norm": 1.921713948249817,
|
| 722 |
+
"learning_rate": 2.1584084084084083e-06,
|
| 723 |
+
"loss": 0.2841,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.9331831831831832,
|
| 728 |
+
"grad_norm": 1.4052722454071045,
|
| 729 |
+
"learning_rate": 1.6891891891891894e-06,
|
| 730 |
+
"loss": 0.2916,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.951951951951952,
|
| 735 |
+
"grad_norm": 1.8074671030044556,
|
| 736 |
+
"learning_rate": 1.2209084084084084e-06,
|
| 737 |
+
"loss": 0.2806,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.9707207207207207,
|
| 742 |
+
"grad_norm": 1.321871042251587,
|
| 743 |
+
"learning_rate": 7.516891891891893e-07,
|
| 744 |
+
"loss": 0.2801,
|
| 745 |
+
"step": 52500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.9894894894894894,
|
| 749 |
+
"grad_norm": 1.944077491760254,
|
| 750 |
+
"learning_rate": 2.8246996996997e-07,
|
| 751 |
+
"loss": 0.2869,
|
| 752 |
+
"step": 53000
|
| 753 |
+
}
|
| 754 |
+
],
|
| 755 |
+
"logging_steps": 500,
|
| 756 |
+
"max_steps": 53280,
|
| 757 |
+
"num_input_tokens_seen": 0,
|
| 758 |
+
"num_train_epochs": 2,
|
| 759 |
+
"save_steps": 500,
|
| 760 |
+
"stateful_callbacks": {
|
| 761 |
+
"TrainerControl": {
|
| 762 |
+
"args": {
|
| 763 |
+
"should_epoch_stop": false,
|
| 764 |
+
"should_evaluate": false,
|
| 765 |
+
"should_log": false,
|
| 766 |
+
"should_save": true,
|
| 767 |
+
"should_training_stop": true
|
| 768 |
+
},
|
| 769 |
+
"attributes": {}
|
| 770 |
+
}
|
| 771 |
+
},
|
| 772 |
+
"total_flos": 1.153747578299351e+17,
|
| 773 |
+
"train_batch_size": 32,
|
| 774 |
+
"trial_name": null,
|
| 775 |
+
"trial_params": null
|
| 776 |
+
}
|
checkpoints/{checkpoint-46670 → checkpoint-53280}/training_args.bin
RENAMED
|
File without changes
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 258368552
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0808fe9e6ee3e9921f8fd3a2388b98e6f54d13f0c87655bb092ad05328fec645
|
| 3 |
size 258368552
|
spiece.model
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
src/data/generate_cyr_lat_pairs.py
CHANGED
|
@@ -110,7 +110,7 @@ with open(output_path, 'w', encoding = "utf-8") as out_file:
|
|
| 110 |
|
| 111 |
with open(output_path, 'a', encoding = "utf-8") as out_file:
|
| 112 |
with open("src/data/kk.txt", 'r', encoding = "utf-8") as f:
|
| 113 |
-
for line in tqdm(islice(f,
|
| 114 |
try:
|
| 115 |
cyr_text = line.strip()
|
| 116 |
lat_text = convert_to_latin(cyr_text).strip()
|
|
|
|
| 110 |
|
| 111 |
with open(output_path, 'a', encoding = "utf-8") as out_file:
|
| 112 |
with open("src/data/kk.txt", 'r', encoding = "utf-8") as f:
|
| 113 |
+
for line in tqdm(islice(f, 1_800_000), total = 1_800_000, desc = "Lines in CC100-Kazakh"):
|
| 114 |
try:
|
| 115 |
cyr_text = line.strip()
|
| 116 |
lat_text = convert_to_latin(cyr_text).strip()
|
src/tokeniser/dalat5_sp.model
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
src/tokeniser/dalat5_sp.vocab
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
src/tokeniser/spiece.model
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c450f9e659db8204e00c7bd17df6a9ab3147d891d3d9ba765937c8912ca7f79
|
| 3 |
+
size 1040866
|
src/train_t5.py
CHANGED
|
@@ -79,6 +79,7 @@ training_args = TrainingArguments(
|
|
| 79 |
save_strategy = "steps",
|
| 80 |
save_steps = 500,
|
| 81 |
save_total_limit = 3,
|
|
|
|
| 82 |
logging_dir = "logs",
|
| 83 |
fp16 = torch.cuda.is_available()
|
| 84 |
)
|
|
|
|
| 79 |
save_strategy = "steps",
|
| 80 |
save_steps = 500,
|
| 81 |
save_total_limit = 3,
|
| 82 |
+
eval_strategy = "steps",
|
| 83 |
logging_dir = "logs",
|
| 84 |
fp16 = torch.cuda.is_available()
|
| 85 |
)
|
src/train_tokeniser.py
CHANGED
|
@@ -9,7 +9,7 @@ corpus = []
|
|
| 9 |
with open("src/data/tokeniser_corpus.txt", "w", encoding = "utf-8") as f_out:
|
| 10 |
with open("src/data/clean_corpus.jsonl", "r", encoding = "utf-8") as f_in:
|
| 11 |
for i, line in enumerate(f_in):
|
| 12 |
-
if i >=
|
| 13 |
break
|
| 14 |
|
| 15 |
item = json.loads(line)
|
|
|
|
| 9 |
with open("src/data/tokeniser_corpus.txt", "w", encoding = "utf-8") as f_out:
|
| 10 |
with open("src/data/clean_corpus.jsonl", "r", encoding = "utf-8") as f_in:
|
| 11 |
for i, line in enumerate(f_in):
|
| 12 |
+
if i >= 1000000: # take 1000000 records for the tokeniser (no need to load everything in the corpus)
|
| 13 |
break
|
| 14 |
|
| 15 |
item = json.loads(line)
|