Training in progress, step 50, checkpoint
Browse files- last-checkpoint/adapter_config.json +5 -5
- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/global_step50/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +1 -1
- last-checkpoint/global_step50/mp_rank_00_model_states.pt +1 -1
- last-checkpoint/latest +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/trainer_state.json +103 -453
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/adapter_config.json
CHANGED
|
@@ -20,13 +20,13 @@
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
-
"down_proj",
|
| 24 |
-
"v_proj",
|
| 25 |
-
"o_proj",
|
| 26 |
-
"k_proj",
|
| 27 |
"q_proj",
|
|
|
|
| 28 |
"up_proj",
|
| 29 |
-
"
|
|
|
|
|
|
|
|
|
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
|
|
|
| 20 |
"rank_pattern": {},
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
|
|
|
|
|
|
|
|
|
|
|
|
| 23 |
"q_proj",
|
| 24 |
+
"gate_proj",
|
| 25 |
"up_proj",
|
| 26 |
+
"o_proj",
|
| 27 |
+
"down_proj",
|
| 28 |
+
"k_proj",
|
| 29 |
+
"v_proj"
|
| 30 |
],
|
| 31 |
"task_type": "CAUSAL_LM",
|
| 32 |
"use_dora": false,
|
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 42002584
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90e6f7f1355e0b50e6787ad798e52ba61bc8a3c2fe93f8979c59fce5aad3fa7b
|
| 3 |
size 42002584
|
last-checkpoint/global_step50/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 251710672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72f56582a1543996046e5f08a23017fc4cf300f72ec2f806e0e9187834c04390
|
| 3 |
size 251710672
|
last-checkpoint/global_step50/mp_rank_00_model_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 153726841
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c1a0f522ad56a8fd99f7087ab3d4067072ddcf9bee87a31166e48f08109c5d86
|
| 3 |
size 153726841
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step50
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -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 sha256:1b5815cfe0a312f88454f2cd698ee65c6bbb0eadab8665cc149e714a36be932a
|
| 3 |
size 14244
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,712 +1,362 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0061162079510703364,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 0.0,
|
| 15 |
-
"loss": 3.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.012232415902140673,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 8.613531161467861e-05,
|
| 22 |
-
"loss": 3.
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01834862385321101,
|
| 27 |
-
"grad_norm":
|
| 28 |
"learning_rate": 0.00013652123889719707,
|
| 29 |
-
"loss": 3.
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.024464831804281346,
|
| 34 |
-
"grad_norm":
|
| 35 |
"learning_rate": 0.00017227062322935723,
|
| 36 |
-
"loss":
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.03058103975535168,
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 0.0002,
|
| 43 |
-
"loss": 2.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03669724770642202,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 0.0002,
|
| 50 |
-
"loss":
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04281345565749235,
|
| 55 |
-
"grad_norm":
|
| 56 |
"learning_rate": 0.0001979166666666667,
|
| 57 |
-
"loss": 2.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04892966360856269,
|
| 62 |
-
"grad_norm":
|
| 63 |
"learning_rate": 0.00019583333333333334,
|
| 64 |
-
"loss": 3.
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05504587155963303,
|
| 69 |
-
"grad_norm":
|
| 70 |
"learning_rate": 0.00019375000000000002,
|
| 71 |
-
"loss": 2.
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06116207951070336,
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 0.00019166666666666667,
|
| 78 |
-
"loss":
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.0672782874617737,
|
| 83 |
-
"grad_norm":
|
| 84 |
"learning_rate": 0.00018958333333333332,
|
| 85 |
-
"loss":
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.07339449541284404,
|
| 90 |
-
"grad_norm":
|
| 91 |
"learning_rate": 0.0001875,
|
| 92 |
-
"loss": 2.
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.07951070336391437,
|
| 97 |
-
"grad_norm": 9.
|
| 98 |
"learning_rate": 0.00018541666666666668,
|
| 99 |
-
"loss": 2.
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.0856269113149847,
|
| 104 |
-
"grad_norm":
|
| 105 |
"learning_rate": 0.00018333333333333334,
|
| 106 |
-
"loss":
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09174311926605505,
|
| 111 |
-
"grad_norm": 8.
|
| 112 |
"learning_rate": 0.00018125000000000001,
|
| 113 |
-
"loss": 2.
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.09785932721712538,
|
| 118 |
-
"grad_norm":
|
| 119 |
"learning_rate": 0.0001791666666666667,
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.10397553516819572,
|
| 125 |
-
"grad_norm": 10.
|
| 126 |
"learning_rate": 0.00017708333333333335,
|
| 127 |
-
"loss": 2.
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11009174311926606,
|
| 132 |
-
"grad_norm":
|
| 133 |
"learning_rate": 0.000175,
|
| 134 |
-
"loss": 2.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.1162079510703364,
|
| 139 |
-
"grad_norm":
|
| 140 |
"learning_rate": 0.00017291666666666668,
|
| 141 |
-
"loss": 2.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12232415902140673,
|
| 146 |
-
"grad_norm": 12.
|
| 147 |
"learning_rate": 0.00017083333333333333,
|
| 148 |
-
"loss": 2.
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.12844036697247707,
|
| 153 |
-
"grad_norm": 8.
|
| 154 |
"learning_rate": 0.00016875,
|
| 155 |
-
"loss":
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.1345565749235474,
|
| 160 |
-
"grad_norm":
|
| 161 |
"learning_rate": 0.0001666666666666667,
|
| 162 |
-
"loss": 2.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.14067278287461774,
|
| 167 |
-
"grad_norm":
|
| 168 |
"learning_rate": 0.00016458333333333334,
|
| 169 |
-
"loss":
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.14678899082568808,
|
| 174 |
-
"grad_norm": 8.
|
| 175 |
"learning_rate": 0.00016250000000000002,
|
| 176 |
-
"loss":
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.1529051987767584,
|
| 181 |
-
"grad_norm":
|
| 182 |
"learning_rate": 0.00016041666666666667,
|
| 183 |
-
"loss": 1.
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.15902140672782875,
|
| 188 |
-
"grad_norm": 7.
|
| 189 |
"learning_rate": 0.00015833333333333332,
|
| 190 |
-
"loss": 2.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.1651376146788991,
|
| 195 |
-
"grad_norm":
|
| 196 |
"learning_rate": 0.00015625,
|
| 197 |
-
"loss":
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1712538226299694,
|
| 202 |
-
"grad_norm":
|
| 203 |
"learning_rate": 0.00015416666666666668,
|
| 204 |
-
"loss": 2.
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.17737003058103976,
|
| 209 |
-
"grad_norm": 6.
|
| 210 |
"learning_rate": 0.00015208333333333333,
|
| 211 |
-
"loss": 2.
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1834862385321101,
|
| 216 |
-
"grad_norm": 6.
|
| 217 |
"learning_rate": 0.00015000000000000001,
|
| 218 |
-
"loss": 2.
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.18960244648318042,
|
| 223 |
-
"grad_norm": 5.
|
| 224 |
"learning_rate": 0.0001479166666666667,
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.19571865443425077,
|
| 230 |
-
"grad_norm":
|
| 231 |
"learning_rate": 0.00014583333333333335,
|
| 232 |
-
"loss":
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2018348623853211,
|
| 237 |
-
"grad_norm":
|
| 238 |
"learning_rate": 0.00014375,
|
| 239 |
-
"loss":
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.20795107033639143,
|
| 244 |
-
"grad_norm":
|
| 245 |
"learning_rate": 0.00014166666666666668,
|
| 246 |
-
"loss": 1.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.21406727828746178,
|
| 251 |
-
"grad_norm": 4.
|
| 252 |
"learning_rate": 0.00013958333333333333,
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.22018348623853212,
|
| 258 |
-
"grad_norm":
|
| 259 |
"learning_rate": 0.0001375,
|
| 260 |
-
"loss": 2.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.22629969418960244,
|
| 265 |
-
"grad_norm": 8.
|
| 266 |
"learning_rate": 0.0001354166666666667,
|
| 267 |
-
"loss": 2.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.2324159021406728,
|
| 272 |
-
"grad_norm": 6.
|
| 273 |
"learning_rate": 0.00013333333333333334,
|
| 274 |
-
"loss": 2.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.23853211009174313,
|
| 279 |
-
"grad_norm":
|
| 280 |
"learning_rate": 0.00013125000000000002,
|
| 281 |
-
"loss":
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.24464831804281345,
|
| 286 |
-
"grad_norm":
|
| 287 |
"learning_rate": 0.00012916666666666667,
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.25076452599388377,
|
| 293 |
-
"grad_norm":
|
| 294 |
"learning_rate": 0.00012708333333333332,
|
| 295 |
-
"loss": 2.
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.25688073394495414,
|
| 300 |
-
"grad_norm":
|
| 301 |
"learning_rate": 0.000125,
|
| 302 |
-
"loss":
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.26299694189602446,
|
| 307 |
-
"grad_norm":
|
| 308 |
"learning_rate": 0.00012291666666666668,
|
| 309 |
-
"loss": 1.
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2691131498470948,
|
| 314 |
-
"grad_norm":
|
| 315 |
"learning_rate": 0.00012083333333333333,
|
| 316 |
-
"loss":
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.27522935779816515,
|
| 321 |
-
"grad_norm":
|
| 322 |
"learning_rate": 0.00011875,
|
| 323 |
-
"loss":
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.28134556574923547,
|
| 328 |
-
"grad_norm":
|
| 329 |
"learning_rate": 0.00011666666666666668,
|
| 330 |
-
"loss":
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.2874617737003058,
|
| 335 |
-
"grad_norm":
|
| 336 |
"learning_rate": 0.00011458333333333333,
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.29357798165137616,
|
| 342 |
-
"grad_norm":
|
| 343 |
"learning_rate": 0.00011250000000000001,
|
| 344 |
-
"loss": 1.
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.2996941896024465,
|
| 349 |
-
"grad_norm":
|
| 350 |
"learning_rate": 0.00011041666666666668,
|
| 351 |
-
"loss": 2.
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3058103975535168,
|
| 356 |
-
"grad_norm": 5.
|
| 357 |
"learning_rate": 0.00010833333333333333,
|
| 358 |
-
"loss":
|
| 359 |
"step": 50
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.3119266055045872,
|
| 363 |
-
"grad_norm": 6.529269695281982,
|
| 364 |
-
"learning_rate": 0.00010625000000000001,
|
| 365 |
-
"loss": 2.7562,
|
| 366 |
-
"step": 51
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.3180428134556575,
|
| 370 |
-
"grad_norm": 6.519890308380127,
|
| 371 |
-
"learning_rate": 0.00010416666666666667,
|
| 372 |
-
"loss": 1.6187,
|
| 373 |
-
"step": 52
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.3241590214067278,
|
| 377 |
-
"grad_norm": 5.84199333190918,
|
| 378 |
-
"learning_rate": 0.00010208333333333333,
|
| 379 |
-
"loss": 1.4098,
|
| 380 |
-
"step": 53
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.3302752293577982,
|
| 384 |
-
"grad_norm": 5.9514336585998535,
|
| 385 |
-
"learning_rate": 0.0001,
|
| 386 |
-
"loss": 1.3005,
|
| 387 |
-
"step": 54
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.3363914373088685,
|
| 391 |
-
"grad_norm": 5.569624423980713,
|
| 392 |
-
"learning_rate": 9.791666666666667e-05,
|
| 393 |
-
"loss": 1.8324,
|
| 394 |
-
"step": 55
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.3425076452599388,
|
| 398 |
-
"grad_norm": 6.462126731872559,
|
| 399 |
-
"learning_rate": 9.583333333333334e-05,
|
| 400 |
-
"loss": 1.5223,
|
| 401 |
-
"step": 56
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.3486238532110092,
|
| 405 |
-
"grad_norm": 5.906198501586914,
|
| 406 |
-
"learning_rate": 9.375e-05,
|
| 407 |
-
"loss": 1.8126,
|
| 408 |
-
"step": 57
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.3547400611620795,
|
| 412 |
-
"grad_norm": 8.181982040405273,
|
| 413 |
-
"learning_rate": 9.166666666666667e-05,
|
| 414 |
-
"loss": 1.5992,
|
| 415 |
-
"step": 58
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.36085626911314983,
|
| 419 |
-
"grad_norm": 4.730391979217529,
|
| 420 |
-
"learning_rate": 8.958333333333335e-05,
|
| 421 |
-
"loss": 1.6709,
|
| 422 |
-
"step": 59
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.3669724770642202,
|
| 426 |
-
"grad_norm": 5.18928337097168,
|
| 427 |
-
"learning_rate": 8.75e-05,
|
| 428 |
-
"loss": 1.7399,
|
| 429 |
-
"step": 60
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.3730886850152905,
|
| 433 |
-
"grad_norm": 7.227060317993164,
|
| 434 |
-
"learning_rate": 8.541666666666666e-05,
|
| 435 |
-
"loss": 2.1354,
|
| 436 |
-
"step": 61
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.37920489296636084,
|
| 440 |
-
"grad_norm": 6.325191497802734,
|
| 441 |
-
"learning_rate": 8.333333333333334e-05,
|
| 442 |
-
"loss": 2.2917,
|
| 443 |
-
"step": 62
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.3853211009174312,
|
| 447 |
-
"grad_norm": 7.269418239593506,
|
| 448 |
-
"learning_rate": 8.125000000000001e-05,
|
| 449 |
-
"loss": 2.4691,
|
| 450 |
-
"step": 63
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.39143730886850153,
|
| 454 |
-
"grad_norm": 9.848830223083496,
|
| 455 |
-
"learning_rate": 7.916666666666666e-05,
|
| 456 |
-
"loss": 2.9251,
|
| 457 |
-
"step": 64
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.39755351681957185,
|
| 461 |
-
"grad_norm": 2.8502776622772217,
|
| 462 |
-
"learning_rate": 7.708333333333334e-05,
|
| 463 |
-
"loss": 0.1757,
|
| 464 |
-
"step": 65
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.4036697247706422,
|
| 468 |
-
"grad_norm": 4.723325252532959,
|
| 469 |
-
"learning_rate": 7.500000000000001e-05,
|
| 470 |
-
"loss": 1.3464,
|
| 471 |
-
"step": 66
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.40978593272171254,
|
| 475 |
-
"grad_norm": 8.410296440124512,
|
| 476 |
-
"learning_rate": 7.291666666666667e-05,
|
| 477 |
-
"loss": 2.6273,
|
| 478 |
-
"step": 67
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.41590214067278286,
|
| 482 |
-
"grad_norm": 7.364680290222168,
|
| 483 |
-
"learning_rate": 7.083333333333334e-05,
|
| 484 |
-
"loss": 2.2223,
|
| 485 |
-
"step": 68
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.42201834862385323,
|
| 489 |
-
"grad_norm": 7.319981575012207,
|
| 490 |
-
"learning_rate": 6.875e-05,
|
| 491 |
-
"loss": 1.9165,
|
| 492 |
-
"step": 69
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.42813455657492355,
|
| 496 |
-
"grad_norm": 6.94813346862793,
|
| 497 |
-
"learning_rate": 6.666666666666667e-05,
|
| 498 |
-
"loss": 2.2407,
|
| 499 |
-
"step": 70
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.43425076452599387,
|
| 503 |
-
"grad_norm": 6.473064422607422,
|
| 504 |
-
"learning_rate": 6.458333333333334e-05,
|
| 505 |
-
"loss": 1.212,
|
| 506 |
-
"step": 71
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.44036697247706424,
|
| 510 |
-
"grad_norm": 8.280125617980957,
|
| 511 |
-
"learning_rate": 6.25e-05,
|
| 512 |
-
"loss": 2.5599,
|
| 513 |
-
"step": 72
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.44648318042813456,
|
| 517 |
-
"grad_norm": 7.279771327972412,
|
| 518 |
-
"learning_rate": 6.041666666666667e-05,
|
| 519 |
-
"loss": 2.2878,
|
| 520 |
-
"step": 73
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.4525993883792049,
|
| 524 |
-
"grad_norm": 8.446040153503418,
|
| 525 |
-
"learning_rate": 5.833333333333334e-05,
|
| 526 |
-
"loss": 2.9793,
|
| 527 |
-
"step": 74
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.45871559633027525,
|
| 531 |
-
"grad_norm": 4.083460330963135,
|
| 532 |
-
"learning_rate": 5.6250000000000005e-05,
|
| 533 |
-
"loss": 0.8241,
|
| 534 |
-
"step": 75
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.4648318042813456,
|
| 538 |
-
"grad_norm": 4.364982604980469,
|
| 539 |
-
"learning_rate": 5.4166666666666664e-05,
|
| 540 |
-
"loss": 0.3777,
|
| 541 |
-
"step": 76
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.4709480122324159,
|
| 545 |
-
"grad_norm": 5.940834045410156,
|
| 546 |
-
"learning_rate": 5.208333333333334e-05,
|
| 547 |
-
"loss": 2.491,
|
| 548 |
-
"step": 77
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.47706422018348627,
|
| 552 |
-
"grad_norm": 9.047613143920898,
|
| 553 |
-
"learning_rate": 5e-05,
|
| 554 |
-
"loss": 2.649,
|
| 555 |
-
"step": 78
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.4831804281345566,
|
| 559 |
-
"grad_norm": 8.733380317687988,
|
| 560 |
-
"learning_rate": 4.791666666666667e-05,
|
| 561 |
-
"loss": 2.8966,
|
| 562 |
-
"step": 79
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.4892966360856269,
|
| 566 |
-
"grad_norm": 8.215818405151367,
|
| 567 |
-
"learning_rate": 4.5833333333333334e-05,
|
| 568 |
-
"loss": 2.5587,
|
| 569 |
-
"step": 80
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.4954128440366973,
|
| 573 |
-
"grad_norm": 6.593921661376953,
|
| 574 |
-
"learning_rate": 4.375e-05,
|
| 575 |
-
"loss": 2.1509,
|
| 576 |
-
"step": 81
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.5015290519877675,
|
| 580 |
-
"grad_norm": 7.0422563552856445,
|
| 581 |
-
"learning_rate": 4.166666666666667e-05,
|
| 582 |
-
"loss": 2.6517,
|
| 583 |
-
"step": 82
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.5076452599388379,
|
| 587 |
-
"grad_norm": 6.228955268859863,
|
| 588 |
-
"learning_rate": 3.958333333333333e-05,
|
| 589 |
-
"loss": 1.8935,
|
| 590 |
-
"step": 83
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.5137614678899083,
|
| 594 |
-
"grad_norm": 5.051086902618408,
|
| 595 |
-
"learning_rate": 3.7500000000000003e-05,
|
| 596 |
-
"loss": 1.1341,
|
| 597 |
-
"step": 84
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.5198776758409785,
|
| 601 |
-
"grad_norm": 6.446547985076904,
|
| 602 |
-
"learning_rate": 3.541666666666667e-05,
|
| 603 |
-
"loss": 2.4077,
|
| 604 |
-
"step": 85
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.5259938837920489,
|
| 608 |
-
"grad_norm": 8.105069160461426,
|
| 609 |
-
"learning_rate": 3.3333333333333335e-05,
|
| 610 |
-
"loss": 1.3238,
|
| 611 |
-
"step": 86
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.5321100917431193,
|
| 615 |
-
"grad_norm": 7.07869291305542,
|
| 616 |
-
"learning_rate": 3.125e-05,
|
| 617 |
-
"loss": 1.8556,
|
| 618 |
-
"step": 87
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.5382262996941896,
|
| 622 |
-
"grad_norm": 7.414394378662109,
|
| 623 |
-
"learning_rate": 2.916666666666667e-05,
|
| 624 |
-
"loss": 2.7577,
|
| 625 |
-
"step": 88
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.5443425076452599,
|
| 629 |
-
"grad_norm": 3.7846713066101074,
|
| 630 |
-
"learning_rate": 2.7083333333333332e-05,
|
| 631 |
-
"loss": 0.9188,
|
| 632 |
-
"step": 89
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.5504587155963303,
|
| 636 |
-
"grad_norm": 6.049407482147217,
|
| 637 |
-
"learning_rate": 2.5e-05,
|
| 638 |
-
"loss": 1.6882,
|
| 639 |
-
"step": 90
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.5565749235474006,
|
| 643 |
-
"grad_norm": 6.262319564819336,
|
| 644 |
-
"learning_rate": 2.2916666666666667e-05,
|
| 645 |
-
"loss": 2.7452,
|
| 646 |
-
"step": 91
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 0.5626911314984709,
|
| 650 |
-
"grad_norm": 6.638827800750732,
|
| 651 |
-
"learning_rate": 2.0833333333333336e-05,
|
| 652 |
-
"loss": 1.9348,
|
| 653 |
-
"step": 92
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.5688073394495413,
|
| 657 |
-
"grad_norm": 9.3283052444458,
|
| 658 |
-
"learning_rate": 1.8750000000000002e-05,
|
| 659 |
-
"loss": 3.2134,
|
| 660 |
-
"step": 93
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 0.5749235474006116,
|
| 664 |
-
"grad_norm": 5.360898017883301,
|
| 665 |
-
"learning_rate": 1.6666666666666667e-05,
|
| 666 |
-
"loss": 2.019,
|
| 667 |
-
"step": 94
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.581039755351682,
|
| 671 |
-
"grad_norm": 8.911893844604492,
|
| 672 |
-
"learning_rate": 1.4583333333333335e-05,
|
| 673 |
-
"loss": 3.1926,
|
| 674 |
-
"step": 95
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 0.5871559633027523,
|
| 678 |
-
"grad_norm": 7.837462425231934,
|
| 679 |
-
"learning_rate": 1.25e-05,
|
| 680 |
-
"loss": 2.5144,
|
| 681 |
-
"step": 96
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.5932721712538226,
|
| 685 |
-
"grad_norm": 5.172741889953613,
|
| 686 |
-
"learning_rate": 1.0416666666666668e-05,
|
| 687 |
-
"loss": 1.7045,
|
| 688 |
-
"step": 97
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 0.599388379204893,
|
| 692 |
-
"grad_norm": 7.193814277648926,
|
| 693 |
-
"learning_rate": 8.333333333333334e-06,
|
| 694 |
-
"loss": 2.8239,
|
| 695 |
-
"step": 98
|
| 696 |
-
},
|
| 697 |
-
{
|
| 698 |
-
"epoch": 0.6055045871559633,
|
| 699 |
-
"grad_norm": 5.897807598114014,
|
| 700 |
-
"learning_rate": 6.25e-06,
|
| 701 |
-
"loss": 1.1928,
|
| 702 |
-
"step": 99
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 0.6116207951070336,
|
| 706 |
-
"grad_norm": 5.558406352996826,
|
| 707 |
-
"learning_rate": 4.166666666666667e-06,
|
| 708 |
-
"loss": 1.8766,
|
| 709 |
-
"step": 100
|
| 710 |
}
|
| 711 |
],
|
| 712 |
"logging_steps": 1,
|
|
@@ -726,7 +376,7 @@
|
|
| 726 |
"attributes": {}
|
| 727 |
}
|
| 728 |
},
|
| 729 |
-
"total_flos":
|
| 730 |
"train_batch_size": 2,
|
| 731 |
"trial_name": null,
|
| 732 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3058103975535168,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 50,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0061162079510703364,
|
| 13 |
+
"grad_norm": 14.838244438171387,
|
| 14 |
"learning_rate": 0.0,
|
| 15 |
+
"loss": 3.3874,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.012232415902140673,
|
| 20 |
+
"grad_norm": 9.148641586303711,
|
| 21 |
"learning_rate": 8.613531161467861e-05,
|
| 22 |
+
"loss": 3.3261,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.01834862385321101,
|
| 27 |
+
"grad_norm": 11.121794700622559,
|
| 28 |
"learning_rate": 0.00013652123889719707,
|
| 29 |
+
"loss": 3.5395,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.024464831804281346,
|
| 34 |
+
"grad_norm": 15.398090362548828,
|
| 35 |
"learning_rate": 0.00017227062322935723,
|
| 36 |
+
"loss": 2.7237,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.03058103975535168,
|
| 41 |
+
"grad_norm": 12.812384605407715,
|
| 42 |
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 2.5766,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03669724770642202,
|
| 48 |
+
"grad_norm": 13.770212173461914,
|
| 49 |
"learning_rate": 0.0002,
|
| 50 |
+
"loss": 2.9812,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04281345565749235,
|
| 55 |
+
"grad_norm": 9.31240177154541,
|
| 56 |
"learning_rate": 0.0001979166666666667,
|
| 57 |
+
"loss": 2.2007,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04892966360856269,
|
| 62 |
+
"grad_norm": 8.453168869018555,
|
| 63 |
"learning_rate": 0.00019583333333333334,
|
| 64 |
+
"loss": 3.0672,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05504587155963303,
|
| 69 |
+
"grad_norm": 6.629086971282959,
|
| 70 |
"learning_rate": 0.00019375000000000002,
|
| 71 |
+
"loss": 2.3837,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06116207951070336,
|
| 76 |
+
"grad_norm": 8.609308242797852,
|
| 77 |
"learning_rate": 0.00019166666666666667,
|
| 78 |
+
"loss": 2.8275,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.0672782874617737,
|
| 83 |
+
"grad_norm": 8.45747184753418,
|
| 84 |
"learning_rate": 0.00018958333333333332,
|
| 85 |
+
"loss": 2.7901,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.07339449541284404,
|
| 90 |
+
"grad_norm": 7.778749942779541,
|
| 91 |
"learning_rate": 0.0001875,
|
| 92 |
+
"loss": 2.1263,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.07951070336391437,
|
| 97 |
+
"grad_norm": 9.23363208770752,
|
| 98 |
"learning_rate": 0.00018541666666666668,
|
| 99 |
+
"loss": 2.7385,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.0856269113149847,
|
| 104 |
+
"grad_norm": 10.394725799560547,
|
| 105 |
"learning_rate": 0.00018333333333333334,
|
| 106 |
+
"loss": 1.9376,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09174311926605505,
|
| 111 |
+
"grad_norm": 8.084063529968262,
|
| 112 |
"learning_rate": 0.00018125000000000001,
|
| 113 |
+
"loss": 2.5122,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.09785932721712538,
|
| 118 |
+
"grad_norm": 7.217422008514404,
|
| 119 |
"learning_rate": 0.0001791666666666667,
|
| 120 |
+
"loss": 1.8823,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.10397553516819572,
|
| 125 |
+
"grad_norm": 10.879976272583008,
|
| 126 |
"learning_rate": 0.00017708333333333335,
|
| 127 |
+
"loss": 2.3302,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11009174311926606,
|
| 132 |
+
"grad_norm": 9.06281566619873,
|
| 133 |
"learning_rate": 0.000175,
|
| 134 |
+
"loss": 2.0973,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.1162079510703364,
|
| 139 |
+
"grad_norm": 9.083320617675781,
|
| 140 |
"learning_rate": 0.00017291666666666668,
|
| 141 |
+
"loss": 2.5653,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12232415902140673,
|
| 146 |
+
"grad_norm": 12.028092384338379,
|
| 147 |
"learning_rate": 0.00017083333333333333,
|
| 148 |
+
"loss": 2.1673,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.12844036697247707,
|
| 153 |
+
"grad_norm": 8.217774391174316,
|
| 154 |
"learning_rate": 0.00016875,
|
| 155 |
+
"loss": 2.9435,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.1345565749235474,
|
| 160 |
+
"grad_norm": 7.708691596984863,
|
| 161 |
"learning_rate": 0.0001666666666666667,
|
| 162 |
+
"loss": 2.5439,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.14067278287461774,
|
| 167 |
+
"grad_norm": 9.682010650634766,
|
| 168 |
"learning_rate": 0.00016458333333333334,
|
| 169 |
+
"loss": 2.5956,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.14678899082568808,
|
| 174 |
+
"grad_norm": 8.222599029541016,
|
| 175 |
"learning_rate": 0.00016250000000000002,
|
| 176 |
+
"loss": 1.9396,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.1529051987767584,
|
| 181 |
+
"grad_norm": 5.596776008605957,
|
| 182 |
"learning_rate": 0.00016041666666666667,
|
| 183 |
+
"loss": 1.6316,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.15902140672782875,
|
| 188 |
+
"grad_norm": 7.1853766441345215,
|
| 189 |
"learning_rate": 0.00015833333333333332,
|
| 190 |
+
"loss": 2.2344,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.1651376146788991,
|
| 195 |
+
"grad_norm": 8.985827445983887,
|
| 196 |
"learning_rate": 0.00015625,
|
| 197 |
+
"loss": 1.8781,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1712538226299694,
|
| 202 |
+
"grad_norm": 7.93477201461792,
|
| 203 |
"learning_rate": 0.00015416666666666668,
|
| 204 |
+
"loss": 2.6036,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.17737003058103976,
|
| 209 |
+
"grad_norm": 6.477381229400635,
|
| 210 |
"learning_rate": 0.00015208333333333333,
|
| 211 |
+
"loss": 2.4767,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.1834862385321101,
|
| 216 |
+
"grad_norm": 6.518289566040039,
|
| 217 |
"learning_rate": 0.00015000000000000001,
|
| 218 |
+
"loss": 2.4941,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.18960244648318042,
|
| 223 |
+
"grad_norm": 5.40209436416626,
|
| 224 |
"learning_rate": 0.0001479166666666667,
|
| 225 |
+
"loss": 0.7781,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.19571865443425077,
|
| 230 |
+
"grad_norm": 10.794727325439453,
|
| 231 |
"learning_rate": 0.00014583333333333335,
|
| 232 |
+
"loss": 1.9824,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2018348623853211,
|
| 237 |
+
"grad_norm": 7.486544609069824,
|
| 238 |
"learning_rate": 0.00014375,
|
| 239 |
+
"loss": 2.679,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.20795107033639143,
|
| 244 |
+
"grad_norm": 6.3984832763671875,
|
| 245 |
"learning_rate": 0.00014166666666666668,
|
| 246 |
+
"loss": 1.1488,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.21406727828746178,
|
| 251 |
+
"grad_norm": 4.752143859863281,
|
| 252 |
"learning_rate": 0.00013958333333333333,
|
| 253 |
+
"loss": 0.2673,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.22018348623853212,
|
| 258 |
+
"grad_norm": 7.158022403717041,
|
| 259 |
"learning_rate": 0.0001375,
|
| 260 |
+
"loss": 2.448,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.22629969418960244,
|
| 265 |
+
"grad_norm": 8.356815338134766,
|
| 266 |
"learning_rate": 0.0001354166666666667,
|
| 267 |
+
"loss": 2.293,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.2324159021406728,
|
| 272 |
+
"grad_norm": 6.427772521972656,
|
| 273 |
"learning_rate": 0.00013333333333333334,
|
| 274 |
+
"loss": 2.0871,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.23853211009174313,
|
| 279 |
+
"grad_norm": 8.426436424255371,
|
| 280 |
"learning_rate": 0.00013125000000000002,
|
| 281 |
+
"loss": 2.6529,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.24464831804281345,
|
| 286 |
+
"grad_norm": 6.148336887359619,
|
| 287 |
"learning_rate": 0.00012916666666666667,
|
| 288 |
+
"loss": 1.0728,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.25076452599388377,
|
| 293 |
+
"grad_norm": 7.108105182647705,
|
| 294 |
"learning_rate": 0.00012708333333333332,
|
| 295 |
+
"loss": 2.3828,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.25688073394495414,
|
| 300 |
+
"grad_norm": 7.101551055908203,
|
| 301 |
"learning_rate": 0.000125,
|
| 302 |
+
"loss": 1.5256,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.26299694189602446,
|
| 307 |
+
"grad_norm": 5.024567127227783,
|
| 308 |
"learning_rate": 0.00012291666666666668,
|
| 309 |
+
"loss": 1.165,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2691131498470948,
|
| 314 |
+
"grad_norm": 5.79965877532959,
|
| 315 |
"learning_rate": 0.00012083333333333333,
|
| 316 |
+
"loss": 2.6776,
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.27522935779816515,
|
| 321 |
+
"grad_norm": 14.309733390808105,
|
| 322 |
"learning_rate": 0.00011875,
|
| 323 |
+
"loss": 2.7559,
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.28134556574923547,
|
| 328 |
+
"grad_norm": 10.46374797821045,
|
| 329 |
"learning_rate": 0.00011666666666666668,
|
| 330 |
+
"loss": 2.9232,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.2874617737003058,
|
| 335 |
+
"grad_norm": 3.6996495723724365,
|
| 336 |
"learning_rate": 0.00011458333333333333,
|
| 337 |
+
"loss": 0.654,
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.29357798165137616,
|
| 342 |
+
"grad_norm": 5.34403657913208,
|
| 343 |
"learning_rate": 0.00011250000000000001,
|
| 344 |
+
"loss": 1.4462,
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.2996941896024465,
|
| 349 |
+
"grad_norm": 5.660646438598633,
|
| 350 |
"learning_rate": 0.00011041666666666668,
|
| 351 |
+
"loss": 2.1049,
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3058103975535168,
|
| 356 |
+
"grad_norm": 5.277816295623779,
|
| 357 |
"learning_rate": 0.00010833333333333333,
|
| 358 |
+
"loss": 1.859,
|
| 359 |
"step": 50
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 360 |
}
|
| 361 |
],
|
| 362 |
"logging_steps": 1,
|
|
|
|
| 376 |
"attributes": {}
|
| 377 |
}
|
| 378 |
},
|
| 379 |
+
"total_flos": 537183016976384.0,
|
| 380 |
"train_batch_size": 2,
|
| 381 |
"trial_name": null,
|
| 382 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 6904
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:700e2ffca061353ef9e5698b2619b2dcba0c46609c85e640fbe8ae9cd993bfaa
|
| 3 |
size 6904
|