Upload folder using huggingface_hub
Browse files- model-00001-of-00002.safetensors +1 -1
- model-00002-of-00002.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- tokenizer.json +6 -1
- trainer_state.json +20 -1518
- training_args.bin +1 -1
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4959881464
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:285654df02c31fd1d9c1dbcf18b0733f999ec31a0c5e3c47ec9a9ca1126d2fe5
|
| 3 |
size 4959881464
|
model-00002-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1270656176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f83f80d6a5a7504845e761798582602c6edb6da2a548406836f310400ba29368
|
| 3 |
size 1270656176
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 12461449840
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e6f869831ede3253bf8a0f04779389f6432b32a152cfc79063831fb8e9d9f02b
|
| 3 |
size 12461449840
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14575
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:143364c92d79d3d47c51a65c4994961f1b5b5faaf817f39fa0955404d28024a7
|
| 3 |
size 14575
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1c5417766e01489df8d4ab49b04f13b1361e304e6234d73da8f428f34e33380e
|
| 3 |
size 627
|
tokenizer.json
CHANGED
|
@@ -1,6 +1,11 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
"padding": null,
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": {
|
| 4 |
+
"direction": "Right",
|
| 5 |
+
"max_length": 1024,
|
| 6 |
+
"strategy": "LongestFirst",
|
| 7 |
+
"stride": 0
|
| 8 |
+
},
|
| 9 |
"padding": null,
|
| 10 |
"added_tokens": [
|
| 11 |
{
|
trainer_state.json
CHANGED
|
@@ -1,1542 +1,44 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "sean_test_out_large/checkpoint-
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 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": 0.0,
|
| 13 |
-
"grad_norm": 1387778.25,
|
| 14 |
-
"learning_rate": 5.972307862720845e-05,
|
| 15 |
-
"loss": 1.1897,
|
| 16 |
-
"step": 500
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.01,
|
| 20 |
-
"grad_norm": 664302.625,
|
| 21 |
-
"learning_rate": 5.94461572544169e-05,
|
| 22 |
-
"loss": 1.0759,
|
| 23 |
-
"step": 1000
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.01,
|
| 27 |
-
"grad_norm": 428486.84375,
|
| 28 |
-
"learning_rate": 5.9169235881625344e-05,
|
| 29 |
-
"loss": 1.0728,
|
| 30 |
-
"step": 1500
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.02,
|
| 34 |
-
"grad_norm": 299683.625,
|
| 35 |
-
"learning_rate": 5.8892314508833796e-05,
|
| 36 |
-
"loss": 1.0409,
|
| 37 |
-
"step": 2000
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.02,
|
| 41 |
-
"grad_norm": 503517.40625,
|
| 42 |
-
"learning_rate": 5.861539313604224e-05,
|
| 43 |
-
"loss": 1.0335,
|
| 44 |
-
"step": 2500
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.03,
|
| 48 |
-
"grad_norm": 657503.8125,
|
| 49 |
-
"learning_rate": 5.833847176325069e-05,
|
| 50 |
-
"loss": 0.9533,
|
| 51 |
-
"step": 3000
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.03,
|
| 55 |
-
"grad_norm": 533787.3125,
|
| 56 |
-
"learning_rate": 5.806155039045914e-05,
|
| 57 |
-
"loss": 0.9992,
|
| 58 |
-
"step": 3500
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.04,
|
| 62 |
-
"grad_norm": 913784.6875,
|
| 63 |
-
"learning_rate": 5.7784629017667584e-05,
|
| 64 |
-
"loss": 0.9657,
|
| 65 |
-
"step": 4000
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.04,
|
| 69 |
-
"grad_norm": 310152.21875,
|
| 70 |
-
"learning_rate": 5.7507707644876036e-05,
|
| 71 |
-
"loss": 0.9652,
|
| 72 |
-
"step": 4500
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.05,
|
| 76 |
-
"grad_norm": 309690.34375,
|
| 77 |
-
"learning_rate": 5.723078627208448e-05,
|
| 78 |
-
"loss": 0.969,
|
| 79 |
-
"step": 5000
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.05,
|
| 83 |
-
"grad_norm": 211456.5,
|
| 84 |
-
"learning_rate": 5.695386489929293e-05,
|
| 85 |
-
"loss": 0.9515,
|
| 86 |
-
"step": 5500
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.06,
|
| 90 |
-
"grad_norm": 391416.65625,
|
| 91 |
-
"learning_rate": 5.667694352650138e-05,
|
| 92 |
-
"loss": 0.9221,
|
| 93 |
-
"step": 6000
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.06,
|
| 97 |
-
"grad_norm": 221054.703125,
|
| 98 |
-
"learning_rate": 5.6400022153709824e-05,
|
| 99 |
-
"loss": 0.9679,
|
| 100 |
-
"step": 6500
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.06,
|
| 104 |
-
"grad_norm": 472076.15625,
|
| 105 |
-
"learning_rate": 5.612310078091827e-05,
|
| 106 |
-
"loss": 0.9519,
|
| 107 |
-
"step": 7000
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.07,
|
| 111 |
-
"grad_norm": 988913.375,
|
| 112 |
-
"learning_rate": 5.584617940812672e-05,
|
| 113 |
-
"loss": 0.9447,
|
| 114 |
-
"step": 7500
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.07,
|
| 118 |
-
"grad_norm": 661127.75,
|
| 119 |
-
"learning_rate": 5.556925803533517e-05,
|
| 120 |
-
"loss": 0.9009,
|
| 121 |
-
"step": 8000
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.08,
|
| 125 |
-
"grad_norm": 212504.640625,
|
| 126 |
-
"learning_rate": 5.529233666254362e-05,
|
| 127 |
-
"loss": 0.9456,
|
| 128 |
-
"step": 8500
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.08,
|
| 132 |
-
"grad_norm": 473405.96875,
|
| 133 |
-
"learning_rate": 5.5015415289752064e-05,
|
| 134 |
-
"loss": 0.9385,
|
| 135 |
-
"step": 9000
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.09,
|
| 139 |
-
"grad_norm": 266084.78125,
|
| 140 |
-
"learning_rate": 5.473849391696051e-05,
|
| 141 |
-
"loss": 1.0131,
|
| 142 |
-
"step": 9500
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.09,
|
| 146 |
-
"grad_norm": 491548.53125,
|
| 147 |
-
"learning_rate": 5.446157254416896e-05,
|
| 148 |
-
"loss": 0.9254,
|
| 149 |
-
"step": 10000
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.1,
|
| 153 |
-
"grad_norm": 267938.4375,
|
| 154 |
-
"learning_rate": 5.418465117137741e-05,
|
| 155 |
-
"loss": 0.9101,
|
| 156 |
-
"step": 10500
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.1,
|
| 160 |
-
"grad_norm": 258540.84375,
|
| 161 |
-
"learning_rate": 5.390772979858586e-05,
|
| 162 |
-
"loss": 0.9234,
|
| 163 |
-
"step": 11000
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.11,
|
| 167 |
-
"grad_norm": 200250.609375,
|
| 168 |
-
"learning_rate": 5.3630808425794305e-05,
|
| 169 |
-
"loss": 0.9157,
|
| 170 |
-
"step": 11500
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.11,
|
| 174 |
-
"grad_norm": 480926.25,
|
| 175 |
-
"learning_rate": 5.335388705300275e-05,
|
| 176 |
-
"loss": 0.9532,
|
| 177 |
-
"step": 12000
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.12,
|
| 181 |
-
"grad_norm": 593979.4375,
|
| 182 |
-
"learning_rate": 5.30769656802112e-05,
|
| 183 |
-
"loss": 0.9184,
|
| 184 |
-
"step": 12500
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.12,
|
| 188 |
-
"grad_norm": 686144.1875,
|
| 189 |
-
"learning_rate": 5.280004430741965e-05,
|
| 190 |
-
"loss": 0.9215,
|
| 191 |
-
"step": 13000
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.12,
|
| 195 |
-
"grad_norm": 582782.0,
|
| 196 |
-
"learning_rate": 5.25231229346281e-05,
|
| 197 |
-
"loss": 0.9151,
|
| 198 |
-
"step": 13500
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.13,
|
| 202 |
-
"grad_norm": 554729.875,
|
| 203 |
-
"learning_rate": 5.2246201561836545e-05,
|
| 204 |
-
"loss": 0.9042,
|
| 205 |
-
"step": 14000
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.13,
|
| 209 |
-
"grad_norm": 423190.0,
|
| 210 |
-
"learning_rate": 5.196928018904499e-05,
|
| 211 |
-
"loss": 0.9114,
|
| 212 |
-
"step": 14500
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.14,
|
| 216 |
-
"grad_norm": 724497.5,
|
| 217 |
-
"learning_rate": 5.169235881625344e-05,
|
| 218 |
-
"loss": 0.9123,
|
| 219 |
-
"step": 15000
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.14,
|
| 223 |
-
"grad_norm": 333863.78125,
|
| 224 |
-
"learning_rate": 5.141543744346189e-05,
|
| 225 |
-
"loss": 0.9237,
|
| 226 |
-
"step": 15500
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.15,
|
| 230 |
-
"grad_norm": 282522.5,
|
| 231 |
-
"learning_rate": 5.113851607067033e-05,
|
| 232 |
-
"loss": 0.8927,
|
| 233 |
-
"step": 16000
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.15,
|
| 237 |
-
"grad_norm": 326572.34375,
|
| 238 |
-
"learning_rate": 5.0861594697878785e-05,
|
| 239 |
-
"loss": 0.9256,
|
| 240 |
-
"step": 16500
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.16,
|
| 244 |
-
"grad_norm": 454635.34375,
|
| 245 |
-
"learning_rate": 5.058467332508723e-05,
|
| 246 |
-
"loss": 0.8991,
|
| 247 |
-
"step": 17000
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.16,
|
| 251 |
-
"grad_norm": 306963.25,
|
| 252 |
-
"learning_rate": 5.030775195229568e-05,
|
| 253 |
-
"loss": 0.8937,
|
| 254 |
-
"step": 17500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.17,
|
| 258 |
-
"grad_norm": 580224.0,
|
| 259 |
-
"learning_rate": 5.003083057950413e-05,
|
| 260 |
-
"loss": 0.9108,
|
| 261 |
-
"step": 18000
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.17,
|
| 265 |
-
"grad_norm": 549220.75,
|
| 266 |
-
"learning_rate": 4.975390920671257e-05,
|
| 267 |
-
"loss": 0.9307,
|
| 268 |
-
"step": 18500
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.18,
|
| 272 |
-
"grad_norm": 188370.0,
|
| 273 |
-
"learning_rate": 4.9476987833921025e-05,
|
| 274 |
-
"loss": 0.9009,
|
| 275 |
-
"step": 19000
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.18,
|
| 279 |
-
"grad_norm": 535151.8125,
|
| 280 |
-
"learning_rate": 4.920006646112947e-05,
|
| 281 |
-
"loss": 0.8773,
|
| 282 |
-
"step": 19500
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.18,
|
| 286 |
-
"grad_norm": 675718.1875,
|
| 287 |
-
"learning_rate": 4.892314508833792e-05,
|
| 288 |
-
"loss": 0.8984,
|
| 289 |
-
"step": 20000
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.19,
|
| 293 |
-
"grad_norm": 247562.0,
|
| 294 |
-
"learning_rate": 4.864622371554637e-05,
|
| 295 |
-
"loss": 0.8896,
|
| 296 |
-
"step": 20500
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.19,
|
| 300 |
-
"grad_norm": 412286.65625,
|
| 301 |
-
"learning_rate": 4.836930234275481e-05,
|
| 302 |
-
"loss": 0.8914,
|
| 303 |
-
"step": 21000
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.2,
|
| 307 |
-
"grad_norm": 362339.53125,
|
| 308 |
-
"learning_rate": 4.8092380969963265e-05,
|
| 309 |
-
"loss": 0.8906,
|
| 310 |
-
"step": 21500
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.2,
|
| 314 |
-
"grad_norm": 307259.8125,
|
| 315 |
-
"learning_rate": 4.781545959717171e-05,
|
| 316 |
-
"loss": 0.8615,
|
| 317 |
-
"step": 22000
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 0.21,
|
| 321 |
-
"grad_norm": 264488.625,
|
| 322 |
-
"learning_rate": 4.753853822438016e-05,
|
| 323 |
-
"loss": 0.896,
|
| 324 |
-
"step": 22500
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.21,
|
| 328 |
-
"grad_norm": 392361.15625,
|
| 329 |
-
"learning_rate": 4.726161685158861e-05,
|
| 330 |
-
"loss": 0.8802,
|
| 331 |
-
"step": 23000
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.22,
|
| 335 |
-
"grad_norm": 844823.375,
|
| 336 |
-
"learning_rate": 4.698469547879705e-05,
|
| 337 |
-
"loss": 0.8927,
|
| 338 |
-
"step": 23500
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.22,
|
| 342 |
-
"grad_norm": 252408.484375,
|
| 343 |
-
"learning_rate": 4.6707774106005505e-05,
|
| 344 |
-
"loss": 0.8942,
|
| 345 |
-
"step": 24000
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.23,
|
| 349 |
-
"grad_norm": 591983.0625,
|
| 350 |
-
"learning_rate": 4.643085273321395e-05,
|
| 351 |
-
"loss": 0.8859,
|
| 352 |
-
"step": 24500
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 0.23,
|
| 356 |
-
"grad_norm": 284245.46875,
|
| 357 |
-
"learning_rate": 4.6153931360422396e-05,
|
| 358 |
-
"loss": 0.8953,
|
| 359 |
-
"step": 25000
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.24,
|
| 363 |
-
"grad_norm": 288328.15625,
|
| 364 |
-
"learning_rate": 4.587700998763085e-05,
|
| 365 |
-
"loss": 0.8731,
|
| 366 |
-
"step": 25500
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.24,
|
| 370 |
-
"grad_norm": 422104.84375,
|
| 371 |
-
"learning_rate": 4.560008861483929e-05,
|
| 372 |
-
"loss": 0.8695,
|
| 373 |
-
"step": 26000
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.24,
|
| 377 |
-
"grad_norm": 450527.375,
|
| 378 |
-
"learning_rate": 4.5323167242047745e-05,
|
| 379 |
-
"loss": 0.8616,
|
| 380 |
-
"step": 26500
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.25,
|
| 384 |
-
"grad_norm": 440466.5,
|
| 385 |
-
"learning_rate": 4.504624586925619e-05,
|
| 386 |
-
"loss": 0.8871,
|
| 387 |
-
"step": 27000
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.25,
|
| 391 |
-
"grad_norm": 386183.34375,
|
| 392 |
-
"learning_rate": 4.4769324496464636e-05,
|
| 393 |
-
"loss": 0.8645,
|
| 394 |
-
"step": 27500
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.26,
|
| 398 |
-
"grad_norm": 542583.0,
|
| 399 |
-
"learning_rate": 4.449240312367309e-05,
|
| 400 |
-
"loss": 0.8755,
|
| 401 |
-
"step": 28000
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.26,
|
| 405 |
-
"grad_norm": 497213.90625,
|
| 406 |
-
"learning_rate": 4.421548175088153e-05,
|
| 407 |
-
"loss": 0.8663,
|
| 408 |
-
"step": 28500
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.27,
|
| 412 |
-
"grad_norm": 1084880.375,
|
| 413 |
-
"learning_rate": 4.3938560378089985e-05,
|
| 414 |
-
"loss": 0.8371,
|
| 415 |
-
"step": 29000
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.27,
|
| 419 |
-
"grad_norm": 365434.125,
|
| 420 |
-
"learning_rate": 4.366163900529843e-05,
|
| 421 |
-
"loss": 0.8791,
|
| 422 |
-
"step": 29500
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.28,
|
| 426 |
-
"grad_norm": 310639.5,
|
| 427 |
-
"learning_rate": 4.3384717632506876e-05,
|
| 428 |
-
"loss": 0.8298,
|
| 429 |
-
"step": 30000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.28,
|
| 433 |
-
"grad_norm": 1322288.0,
|
| 434 |
-
"learning_rate": 4.310779625971533e-05,
|
| 435 |
-
"loss": 0.8658,
|
| 436 |
-
"step": 30500
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.29,
|
| 440 |
-
"grad_norm": 488033.0625,
|
| 441 |
-
"learning_rate": 4.283087488692377e-05,
|
| 442 |
-
"loss": 0.8556,
|
| 443 |
-
"step": 31000
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.29,
|
| 447 |
-
"grad_norm": 404639.3125,
|
| 448 |
-
"learning_rate": 4.2553953514132225e-05,
|
| 449 |
-
"loss": 0.8547,
|
| 450 |
-
"step": 31500
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.3,
|
| 454 |
-
"grad_norm": 324229.59375,
|
| 455 |
-
"learning_rate": 4.227703214134067e-05,
|
| 456 |
-
"loss": 0.8853,
|
| 457 |
-
"step": 32000
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.3,
|
| 461 |
-
"grad_norm": 391828.6875,
|
| 462 |
-
"learning_rate": 4.2000110768549116e-05,
|
| 463 |
-
"loss": 0.8725,
|
| 464 |
-
"step": 32500
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.3,
|
| 468 |
-
"grad_norm": 481346.0,
|
| 469 |
-
"learning_rate": 4.172318939575757e-05,
|
| 470 |
-
"loss": 0.8484,
|
| 471 |
-
"step": 33000
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.31,
|
| 475 |
-
"grad_norm": 500705.90625,
|
| 476 |
-
"learning_rate": 4.144626802296601e-05,
|
| 477 |
-
"loss": 0.8415,
|
| 478 |
-
"step": 33500
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.31,
|
| 482 |
-
"grad_norm": 484916.0625,
|
| 483 |
-
"learning_rate": 4.116934665017446e-05,
|
| 484 |
-
"loss": 0.8464,
|
| 485 |
-
"step": 34000
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.32,
|
| 489 |
-
"grad_norm": 305320.15625,
|
| 490 |
-
"learning_rate": 4.089242527738291e-05,
|
| 491 |
-
"loss": 0.8439,
|
| 492 |
-
"step": 34500
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 0.32,
|
| 496 |
-
"grad_norm": 275214.6875,
|
| 497 |
-
"learning_rate": 4.0615503904591356e-05,
|
| 498 |
-
"loss": 0.8257,
|
| 499 |
-
"step": 35000
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.33,
|
| 503 |
-
"grad_norm": 364166.0,
|
| 504 |
-
"learning_rate": 4.033858253179981e-05,
|
| 505 |
-
"loss": 0.8763,
|
| 506 |
-
"step": 35500
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.33,
|
| 510 |
-
"grad_norm": 285496.46875,
|
| 511 |
-
"learning_rate": 4.006166115900825e-05,
|
| 512 |
-
"loss": 0.856,
|
| 513 |
-
"step": 36000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.34,
|
| 517 |
-
"grad_norm": 959005.0625,
|
| 518 |
-
"learning_rate": 3.97847397862167e-05,
|
| 519 |
-
"loss": 0.8306,
|
| 520 |
-
"step": 36500
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.34,
|
| 524 |
-
"grad_norm": 357090.28125,
|
| 525 |
-
"learning_rate": 3.950781841342515e-05,
|
| 526 |
-
"loss": 0.8564,
|
| 527 |
-
"step": 37000
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 0.35,
|
| 531 |
-
"grad_norm": 605763.8125,
|
| 532 |
-
"learning_rate": 3.9230897040633596e-05,
|
| 533 |
-
"loss": 0.8476,
|
| 534 |
-
"step": 37500
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.35,
|
| 538 |
-
"grad_norm": 418555.625,
|
| 539 |
-
"learning_rate": 3.895397566784205e-05,
|
| 540 |
-
"loss": 0.857,
|
| 541 |
-
"step": 38000
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.36,
|
| 545 |
-
"grad_norm": 726787.125,
|
| 546 |
-
"learning_rate": 3.8677054295050493e-05,
|
| 547 |
-
"loss": 0.8627,
|
| 548 |
-
"step": 38500
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.36,
|
| 552 |
-
"grad_norm": 212226.21875,
|
| 553 |
-
"learning_rate": 3.840013292225894e-05,
|
| 554 |
-
"loss": 0.8172,
|
| 555 |
-
"step": 39000
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.36,
|
| 559 |
-
"grad_norm": 219878.765625,
|
| 560 |
-
"learning_rate": 3.812321154946739e-05,
|
| 561 |
-
"loss": 0.8476,
|
| 562 |
-
"step": 39500
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 0.37,
|
| 566 |
-
"grad_norm": 520633.34375,
|
| 567 |
-
"learning_rate": 3.7846290176675836e-05,
|
| 568 |
-
"loss": 0.8226,
|
| 569 |
-
"step": 40000
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.37,
|
| 573 |
-
"grad_norm": 317160.71875,
|
| 574 |
-
"learning_rate": 3.756936880388429e-05,
|
| 575 |
-
"loss": 0.8235,
|
| 576 |
-
"step": 40500
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.38,
|
| 580 |
-
"grad_norm": 409949.8125,
|
| 581 |
-
"learning_rate": 3.7292447431092734e-05,
|
| 582 |
-
"loss": 0.8112,
|
| 583 |
-
"step": 41000
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.38,
|
| 587 |
-
"grad_norm": 287649.84375,
|
| 588 |
-
"learning_rate": 3.701552605830118e-05,
|
| 589 |
-
"loss": 0.8357,
|
| 590 |
-
"step": 41500
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.39,
|
| 594 |
-
"grad_norm": 386562.09375,
|
| 595 |
-
"learning_rate": 3.673860468550963e-05,
|
| 596 |
-
"loss": 0.8425,
|
| 597 |
-
"step": 42000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 0.39,
|
| 601 |
-
"grad_norm": 454850.125,
|
| 602 |
-
"learning_rate": 3.6461683312718076e-05,
|
| 603 |
-
"loss": 0.8347,
|
| 604 |
-
"step": 42500
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.4,
|
| 608 |
-
"grad_norm": 239857.5,
|
| 609 |
-
"learning_rate": 3.618476193992653e-05,
|
| 610 |
-
"loss": 0.8619,
|
| 611 |
-
"step": 43000
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.4,
|
| 615 |
-
"grad_norm": 227077.171875,
|
| 616 |
-
"learning_rate": 3.5907840567134974e-05,
|
| 617 |
-
"loss": 0.847,
|
| 618 |
-
"step": 43500
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.41,
|
| 622 |
-
"grad_norm": 465131.4375,
|
| 623 |
-
"learning_rate": 3.563091919434342e-05,
|
| 624 |
-
"loss": 0.8333,
|
| 625 |
-
"step": 44000
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.41,
|
| 629 |
-
"grad_norm": 329121.21875,
|
| 630 |
-
"learning_rate": 3.535399782155187e-05,
|
| 631 |
-
"loss": 0.8466,
|
| 632 |
-
"step": 44500
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 0.42,
|
| 636 |
-
"grad_norm": 362261.40625,
|
| 637 |
-
"learning_rate": 3.5077076448760316e-05,
|
| 638 |
-
"loss": 0.8076,
|
| 639 |
-
"step": 45000
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 0.42,
|
| 643 |
-
"grad_norm": 729279.75,
|
| 644 |
-
"learning_rate": 3.480015507596876e-05,
|
| 645 |
-
"loss": 0.8015,
|
| 646 |
-
"step": 45500
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 0.42,
|
| 650 |
-
"grad_norm": 311664.15625,
|
| 651 |
-
"learning_rate": 3.4523233703177214e-05,
|
| 652 |
-
"loss": 0.838,
|
| 653 |
-
"step": 46000
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 0.43,
|
| 657 |
-
"grad_norm": 267956.34375,
|
| 658 |
-
"learning_rate": 3.424631233038566e-05,
|
| 659 |
-
"loss": 0.8229,
|
| 660 |
-
"step": 46500
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 0.43,
|
| 664 |
-
"grad_norm": 393154.1875,
|
| 665 |
-
"learning_rate": 3.396939095759411e-05,
|
| 666 |
-
"loss": 0.7973,
|
| 667 |
-
"step": 47000
|
| 668 |
-
},
|
| 669 |
-
{
|
| 670 |
-
"epoch": 0.44,
|
| 671 |
-
"grad_norm": 298007.96875,
|
| 672 |
-
"learning_rate": 3.3692469584802556e-05,
|
| 673 |
-
"loss": 0.8309,
|
| 674 |
-
"step": 47500
|
| 675 |
-
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 0.44,
|
| 678 |
-
"grad_norm": 235954.09375,
|
| 679 |
-
"learning_rate": 3.3415548212011e-05,
|
| 680 |
-
"loss": 0.808,
|
| 681 |
-
"step": 48000
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 0.45,
|
| 685 |
-
"grad_norm": 565394.3125,
|
| 686 |
-
"learning_rate": 3.3138626839219454e-05,
|
| 687 |
-
"loss": 0.8123,
|
| 688 |
-
"step": 48500
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 0.45,
|
| 692 |
-
"grad_norm": 228396.78125,
|
| 693 |
-
"learning_rate": 3.28617054664279e-05,
|
| 694 |
-
"loss": 0.8456,
|
| 695 |
-
"step": 49000
|
| 696 |
-
},
|
| 697 |
-
{
|
| 698 |
-
"epoch": 0.46,
|
| 699 |
-
"grad_norm": 314704.09375,
|
| 700 |
-
"learning_rate": 3.258478409363635e-05,
|
| 701 |
-
"loss": 0.8116,
|
| 702 |
-
"step": 49500
|
| 703 |
-
},
|
| 704 |
-
{
|
| 705 |
-
"epoch": 0.46,
|
| 706 |
-
"grad_norm": 250648.609375,
|
| 707 |
-
"learning_rate": 3.2307862720844797e-05,
|
| 708 |
-
"loss": 0.8284,
|
| 709 |
-
"step": 50000
|
| 710 |
-
},
|
| 711 |
-
{
|
| 712 |
-
"epoch": 0.47,
|
| 713 |
-
"grad_norm": 409734.96875,
|
| 714 |
-
"learning_rate": 3.203094134805324e-05,
|
| 715 |
-
"loss": 0.8015,
|
| 716 |
-
"step": 50500
|
| 717 |
-
},
|
| 718 |
-
{
|
| 719 |
-
"epoch": 0.47,
|
| 720 |
-
"grad_norm": 171156.265625,
|
| 721 |
-
"learning_rate": 3.1754019975261694e-05,
|
| 722 |
-
"loss": 0.8322,
|
| 723 |
-
"step": 51000
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 0.48,
|
| 727 |
-
"grad_norm": 300950.65625,
|
| 728 |
-
"learning_rate": 3.147709860247014e-05,
|
| 729 |
-
"loss": 0.8215,
|
| 730 |
-
"step": 51500
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 0.48,
|
| 734 |
-
"grad_norm": 223661.734375,
|
| 735 |
-
"learning_rate": 3.120017722967859e-05,
|
| 736 |
-
"loss": 0.8004,
|
| 737 |
-
"step": 52000
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 0.48,
|
| 741 |
-
"grad_norm": 303242.25,
|
| 742 |
-
"learning_rate": 3.0923255856887037e-05,
|
| 743 |
-
"loss": 0.8109,
|
| 744 |
-
"step": 52500
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.49,
|
| 748 |
-
"grad_norm": 190445.625,
|
| 749 |
-
"learning_rate": 3.064633448409548e-05,
|
| 750 |
-
"loss": 0.8126,
|
| 751 |
-
"step": 53000
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.49,
|
| 755 |
-
"grad_norm": 312735.21875,
|
| 756 |
-
"learning_rate": 3.036941311130393e-05,
|
| 757 |
-
"loss": 0.8186,
|
| 758 |
-
"step": 53500
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.5,
|
| 762 |
-
"grad_norm": 492028.34375,
|
| 763 |
-
"learning_rate": 3.009249173851238e-05,
|
| 764 |
-
"loss": 0.7993,
|
| 765 |
-
"step": 54000
|
| 766 |
-
},
|
| 767 |
{
|
| 768 |
"epoch": 0.5,
|
| 769 |
-
"grad_norm":
|
| 770 |
-
"learning_rate":
|
| 771 |
-
"loss":
|
| 772 |
-
"step":
|
| 773 |
-
},
|
| 774 |
-
{
|
| 775 |
-
"epoch": 0.51,
|
| 776 |
-
"grad_norm": 361054.71875,
|
| 777 |
-
"learning_rate": 2.9538648992929277e-05,
|
| 778 |
-
"loss": 0.8034,
|
| 779 |
-
"step": 55000
|
| 780 |
-
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 0.51,
|
| 783 |
-
"grad_norm": 333324.875,
|
| 784 |
-
"learning_rate": 2.9261727620137725e-05,
|
| 785 |
-
"loss": 0.7747,
|
| 786 |
-
"step": 55500
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 0.52,
|
| 790 |
-
"grad_norm": 220997.15625,
|
| 791 |
-
"learning_rate": 2.898480624734617e-05,
|
| 792 |
-
"loss": 0.7898,
|
| 793 |
-
"step": 56000
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 0.52,
|
| 797 |
-
"grad_norm": 427477.28125,
|
| 798 |
-
"learning_rate": 2.870788487455462e-05,
|
| 799 |
-
"loss": 0.8487,
|
| 800 |
-
"step": 56500
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 0.53,
|
| 804 |
-
"grad_norm": 494515.15625,
|
| 805 |
-
"learning_rate": 2.8430963501763068e-05,
|
| 806 |
-
"loss": 0.7867,
|
| 807 |
-
"step": 57000
|
| 808 |
-
},
|
| 809 |
-
{
|
| 810 |
-
"epoch": 0.53,
|
| 811 |
-
"grad_norm": 131962.8125,
|
| 812 |
-
"learning_rate": 2.8154042128971517e-05,
|
| 813 |
-
"loss": 0.8257,
|
| 814 |
-
"step": 57500
|
| 815 |
-
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 0.54,
|
| 818 |
-
"grad_norm": 360923.75,
|
| 819 |
-
"learning_rate": 2.7877120756179962e-05,
|
| 820 |
-
"loss": 0.824,
|
| 821 |
-
"step": 58000
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 0.54,
|
| 825 |
-
"grad_norm": 537102.3125,
|
| 826 |
-
"learning_rate": 2.760019938338841e-05,
|
| 827 |
-
"loss": 0.7947,
|
| 828 |
-
"step": 58500
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 0.54,
|
| 832 |
-
"grad_norm": 185466.890625,
|
| 833 |
-
"learning_rate": 2.732327801059686e-05,
|
| 834 |
-
"loss": 0.7965,
|
| 835 |
-
"step": 59000
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 0.55,
|
| 839 |
-
"grad_norm": 449036.0,
|
| 840 |
-
"learning_rate": 2.7046356637805308e-05,
|
| 841 |
-
"loss": 0.8359,
|
| 842 |
-
"step": 59500
|
| 843 |
-
},
|
| 844 |
-
{
|
| 845 |
-
"epoch": 0.55,
|
| 846 |
-
"grad_norm": 302378.875,
|
| 847 |
-
"learning_rate": 2.6769435265013757e-05,
|
| 848 |
-
"loss": 0.791,
|
| 849 |
-
"step": 60000
|
| 850 |
-
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 0.56,
|
| 853 |
-
"grad_norm": 170428.640625,
|
| 854 |
-
"learning_rate": 2.6492513892222202e-05,
|
| 855 |
-
"loss": 0.804,
|
| 856 |
-
"step": 60500
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 0.56,
|
| 860 |
-
"grad_norm": 486227.0,
|
| 861 |
-
"learning_rate": 2.621559251943065e-05,
|
| 862 |
-
"loss": 0.7807,
|
| 863 |
-
"step": 61000
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 0.57,
|
| 867 |
-
"grad_norm": 657373.0,
|
| 868 |
-
"learning_rate": 2.59386711466391e-05,
|
| 869 |
-
"loss": 0.8004,
|
| 870 |
-
"step": 61500
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 0.57,
|
| 874 |
-
"grad_norm": 386137.0625,
|
| 875 |
-
"learning_rate": 2.5661749773847548e-05,
|
| 876 |
-
"loss": 0.7669,
|
| 877 |
-
"step": 62000
|
| 878 |
-
},
|
| 879 |
-
{
|
| 880 |
-
"epoch": 0.58,
|
| 881 |
-
"grad_norm": 451425.8125,
|
| 882 |
-
"learning_rate": 2.5384828401055994e-05,
|
| 883 |
-
"loss": 0.7948,
|
| 884 |
-
"step": 62500
|
| 885 |
-
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 0.58,
|
| 888 |
-
"grad_norm": 188410.265625,
|
| 889 |
-
"learning_rate": 2.5107907028264442e-05,
|
| 890 |
-
"loss": 0.7988,
|
| 891 |
-
"step": 63000
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 0.59,
|
| 895 |
-
"grad_norm": 543696.4375,
|
| 896 |
-
"learning_rate": 2.483098565547289e-05,
|
| 897 |
-
"loss": 0.7927,
|
| 898 |
-
"step": 63500
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 0.59,
|
| 902 |
-
"grad_norm": 317845.90625,
|
| 903 |
-
"learning_rate": 2.455406428268134e-05,
|
| 904 |
-
"loss": 0.8021,
|
| 905 |
-
"step": 64000
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 0.6,
|
| 909 |
-
"grad_norm": 567647.5,
|
| 910 |
-
"learning_rate": 2.427714290988979e-05,
|
| 911 |
-
"loss": 0.7853,
|
| 912 |
-
"step": 64500
|
| 913 |
-
},
|
| 914 |
-
{
|
| 915 |
-
"epoch": 0.6,
|
| 916 |
-
"grad_norm": 413894.4375,
|
| 917 |
-
"learning_rate": 2.4000221537098234e-05,
|
| 918 |
-
"loss": 0.7973,
|
| 919 |
-
"step": 65000
|
| 920 |
-
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 0.6,
|
| 923 |
-
"grad_norm": 519043.9375,
|
| 924 |
-
"learning_rate": 2.3723300164306682e-05,
|
| 925 |
-
"loss": 0.7516,
|
| 926 |
-
"step": 65500
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 0.61,
|
| 930 |
-
"grad_norm": 129269.5234375,
|
| 931 |
-
"learning_rate": 2.344637879151513e-05,
|
| 932 |
-
"loss": 0.8054,
|
| 933 |
-
"step": 66000
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 0.61,
|
| 937 |
-
"grad_norm": 272380.0625,
|
| 938 |
-
"learning_rate": 2.316945741872358e-05,
|
| 939 |
-
"loss": 0.8112,
|
| 940 |
-
"step": 66500
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 0.62,
|
| 944 |
-
"grad_norm": 299374.875,
|
| 945 |
-
"learning_rate": 2.2892536045932025e-05,
|
| 946 |
-
"loss": 0.7687,
|
| 947 |
-
"step": 67000
|
| 948 |
-
},
|
| 949 |
-
{
|
| 950 |
-
"epoch": 0.62,
|
| 951 |
-
"grad_norm": 597694.5,
|
| 952 |
-
"learning_rate": 2.2615614673140474e-05,
|
| 953 |
-
"loss": 0.8072,
|
| 954 |
-
"step": 67500
|
| 955 |
-
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 0.63,
|
| 958 |
-
"grad_norm": 273690.375,
|
| 959 |
-
"learning_rate": 2.2338693300348922e-05,
|
| 960 |
-
"loss": 0.7895,
|
| 961 |
-
"step": 68000
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 0.63,
|
| 965 |
-
"grad_norm": 237268.21875,
|
| 966 |
-
"learning_rate": 2.206177192755737e-05,
|
| 967 |
-
"loss": 0.7802,
|
| 968 |
-
"step": 68500
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 0.64,
|
| 972 |
-
"grad_norm": 347646.09375,
|
| 973 |
-
"learning_rate": 2.178485055476582e-05,
|
| 974 |
-
"loss": 0.7991,
|
| 975 |
-
"step": 69000
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 0.64,
|
| 979 |
-
"grad_norm": 266270.28125,
|
| 980 |
-
"learning_rate": 2.1507929181974265e-05,
|
| 981 |
-
"loss": 0.7979,
|
| 982 |
-
"step": 69500
|
| 983 |
-
},
|
| 984 |
-
{
|
| 985 |
-
"epoch": 0.65,
|
| 986 |
-
"grad_norm": 371650.84375,
|
| 987 |
-
"learning_rate": 2.1231007809182714e-05,
|
| 988 |
-
"loss": 0.7616,
|
| 989 |
-
"step": 70000
|
| 990 |
-
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 0.65,
|
| 993 |
-
"grad_norm": 389250.75,
|
| 994 |
-
"learning_rate": 2.0954086436391163e-05,
|
| 995 |
-
"loss": 0.8056,
|
| 996 |
-
"step": 70500
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 0.66,
|
| 1000 |
-
"grad_norm": 472425.71875,
|
| 1001 |
-
"learning_rate": 2.067716506359961e-05,
|
| 1002 |
-
"loss": 0.7632,
|
| 1003 |
-
"step": 71000
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 0.66,
|
| 1007 |
-
"grad_norm": 341248.3125,
|
| 1008 |
-
"learning_rate": 2.0400243690808057e-05,
|
| 1009 |
-
"loss": 0.7648,
|
| 1010 |
-
"step": 71500
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 0.66,
|
| 1014 |
-
"grad_norm": 470803.96875,
|
| 1015 |
-
"learning_rate": 2.0123322318016505e-05,
|
| 1016 |
-
"loss": 0.7699,
|
| 1017 |
-
"step": 72000
|
| 1018 |
-
},
|
| 1019 |
-
{
|
| 1020 |
-
"epoch": 0.67,
|
| 1021 |
-
"grad_norm": 363643.71875,
|
| 1022 |
-
"learning_rate": 1.9846400945224954e-05,
|
| 1023 |
-
"loss": 0.7929,
|
| 1024 |
-
"step": 72500
|
| 1025 |
-
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 0.67,
|
| 1028 |
-
"grad_norm": 380603.65625,
|
| 1029 |
-
"learning_rate": 1.9569479572433403e-05,
|
| 1030 |
-
"loss": 0.7614,
|
| 1031 |
-
"step": 73000
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 0.68,
|
| 1035 |
-
"grad_norm": 282191.78125,
|
| 1036 |
-
"learning_rate": 1.929255819964185e-05,
|
| 1037 |
-
"loss": 0.7508,
|
| 1038 |
-
"step": 73500
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 0.68,
|
| 1042 |
-
"grad_norm": 270537.5625,
|
| 1043 |
-
"learning_rate": 1.9015636826850297e-05,
|
| 1044 |
-
"loss": 0.7724,
|
| 1045 |
-
"step": 74000
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 0.69,
|
| 1049 |
-
"grad_norm": 786935.9375,
|
| 1050 |
-
"learning_rate": 1.8738715454058745e-05,
|
| 1051 |
-
"loss": 0.7903,
|
| 1052 |
-
"step": 74500
|
| 1053 |
-
},
|
| 1054 |
-
{
|
| 1055 |
-
"epoch": 0.69,
|
| 1056 |
-
"grad_norm": 276354.875,
|
| 1057 |
-
"learning_rate": 1.8461794081267194e-05,
|
| 1058 |
-
"loss": 0.8034,
|
| 1059 |
-
"step": 75000
|
| 1060 |
-
},
|
| 1061 |
-
{
|
| 1062 |
-
"epoch": 0.7,
|
| 1063 |
-
"grad_norm": 488051.25,
|
| 1064 |
-
"learning_rate": 1.8184872708475643e-05,
|
| 1065 |
-
"loss": 0.7988,
|
| 1066 |
-
"step": 75500
|
| 1067 |
-
},
|
| 1068 |
-
{
|
| 1069 |
-
"epoch": 0.7,
|
| 1070 |
-
"grad_norm": 268592.53125,
|
| 1071 |
-
"learning_rate": 1.7907951335684088e-05,
|
| 1072 |
-
"loss": 0.771,
|
| 1073 |
-
"step": 76000
|
| 1074 |
-
},
|
| 1075 |
-
{
|
| 1076 |
-
"epoch": 0.71,
|
| 1077 |
-
"grad_norm": 286408.90625,
|
| 1078 |
-
"learning_rate": 1.7631029962892537e-05,
|
| 1079 |
-
"loss": 0.7732,
|
| 1080 |
-
"step": 76500
|
| 1081 |
-
},
|
| 1082 |
-
{
|
| 1083 |
-
"epoch": 0.71,
|
| 1084 |
-
"grad_norm": 287447.375,
|
| 1085 |
-
"learning_rate": 1.7354108590100985e-05,
|
| 1086 |
-
"loss": 0.7788,
|
| 1087 |
-
"step": 77000
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 0.72,
|
| 1091 |
-
"grad_norm": 427488.15625,
|
| 1092 |
-
"learning_rate": 1.7077187217309434e-05,
|
| 1093 |
-
"loss": 0.7763,
|
| 1094 |
-
"step": 77500
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 0.72,
|
| 1098 |
-
"grad_norm": 327204.96875,
|
| 1099 |
-
"learning_rate": 1.6800265844517883e-05,
|
| 1100 |
-
"loss": 0.7814,
|
| 1101 |
-
"step": 78000
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 0.72,
|
| 1105 |
-
"grad_norm": 446818.0,
|
| 1106 |
-
"learning_rate": 1.6523344471726328e-05,
|
| 1107 |
-
"loss": 0.773,
|
| 1108 |
-
"step": 78500
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 0.73,
|
| 1112 |
-
"grad_norm": 367178.46875,
|
| 1113 |
-
"learning_rate": 1.6246423098934777e-05,
|
| 1114 |
-
"loss": 0.7784,
|
| 1115 |
-
"step": 79000
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 0.73,
|
| 1119 |
-
"grad_norm": 564747.75,
|
| 1120 |
-
"learning_rate": 1.5969501726143226e-05,
|
| 1121 |
-
"loss": 0.7751,
|
| 1122 |
-
"step": 79500
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 0.74,
|
| 1126 |
-
"grad_norm": 373075.34375,
|
| 1127 |
-
"learning_rate": 1.5692580353351674e-05,
|
| 1128 |
-
"loss": 0.7701,
|
| 1129 |
-
"step": 80000
|
| 1130 |
-
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 0.74,
|
| 1133 |
-
"grad_norm": 220253.890625,
|
| 1134 |
-
"learning_rate": 1.541565898056012e-05,
|
| 1135 |
-
"loss": 0.7717,
|
| 1136 |
-
"step": 80500
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 0.75,
|
| 1140 |
-
"grad_norm": 442242.46875,
|
| 1141 |
-
"learning_rate": 1.513873760776857e-05,
|
| 1142 |
-
"loss": 0.7699,
|
| 1143 |
-
"step": 81000
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 0.75,
|
| 1147 |
-
"grad_norm": 281603.1875,
|
| 1148 |
-
"learning_rate": 1.4861816234977015e-05,
|
| 1149 |
-
"loss": 0.7626,
|
| 1150 |
-
"step": 81500
|
| 1151 |
-
},
|
| 1152 |
-
{
|
| 1153 |
-
"epoch": 0.76,
|
| 1154 |
-
"grad_norm": 420246.21875,
|
| 1155 |
-
"learning_rate": 1.4584894862185464e-05,
|
| 1156 |
-
"loss": 0.7719,
|
| 1157 |
-
"step": 82000
|
| 1158 |
-
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 0.76,
|
| 1161 |
-
"grad_norm": 347264.375,
|
| 1162 |
-
"learning_rate": 1.4307973489393911e-05,
|
| 1163 |
-
"loss": 0.764,
|
| 1164 |
-
"step": 82500
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 0.77,
|
| 1168 |
-
"grad_norm": 326494.21875,
|
| 1169 |
-
"learning_rate": 1.403105211660236e-05,
|
| 1170 |
-
"loss": 0.7617,
|
| 1171 |
-
"step": 83000
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 0.77,
|
| 1175 |
-
"grad_norm": 193308.671875,
|
| 1176 |
-
"learning_rate": 1.3754130743810807e-05,
|
| 1177 |
-
"loss": 0.7599,
|
| 1178 |
-
"step": 83500
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 0.78,
|
| 1182 |
-
"grad_norm": 283088.8125,
|
| 1183 |
-
"learning_rate": 1.3477209371019255e-05,
|
| 1184 |
-
"loss": 0.7628,
|
| 1185 |
-
"step": 84000
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 0.78,
|
| 1189 |
-
"grad_norm": 539284.0,
|
| 1190 |
-
"learning_rate": 1.3200287998227702e-05,
|
| 1191 |
-
"loss": 0.7358,
|
| 1192 |
-
"step": 84500
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 0.78,
|
| 1196 |
-
"grad_norm": 473964.84375,
|
| 1197 |
-
"learning_rate": 1.2923366625436151e-05,
|
| 1198 |
-
"loss": 0.755,
|
| 1199 |
-
"step": 85000
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 0.79,
|
| 1203 |
-
"grad_norm": 327867.3125,
|
| 1204 |
-
"learning_rate": 1.26464452526446e-05,
|
| 1205 |
-
"loss": 0.7718,
|
| 1206 |
-
"step": 85500
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 0.79,
|
| 1210 |
-
"grad_norm": 263024.28125,
|
| 1211 |
-
"learning_rate": 1.2369523879853047e-05,
|
| 1212 |
-
"loss": 0.7722,
|
| 1213 |
-
"step": 86000
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 0.8,
|
| 1217 |
-
"grad_norm": 266528.5625,
|
| 1218 |
-
"learning_rate": 1.2092602507061495e-05,
|
| 1219 |
-
"loss": 0.7635,
|
| 1220 |
-
"step": 86500
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 0.8,
|
| 1224 |
-
"grad_norm": 126082.46875,
|
| 1225 |
-
"learning_rate": 1.1815681134269942e-05,
|
| 1226 |
-
"loss": 0.7508,
|
| 1227 |
-
"step": 87000
|
| 1228 |
-
},
|
| 1229 |
-
{
|
| 1230 |
-
"epoch": 0.81,
|
| 1231 |
-
"grad_norm": 300010.3125,
|
| 1232 |
-
"learning_rate": 1.1538759761478391e-05,
|
| 1233 |
-
"loss": 0.7755,
|
| 1234 |
-
"step": 87500
|
| 1235 |
-
},
|
| 1236 |
-
{
|
| 1237 |
-
"epoch": 0.81,
|
| 1238 |
-
"grad_norm": 471967.5625,
|
| 1239 |
-
"learning_rate": 1.1261838388686838e-05,
|
| 1240 |
-
"loss": 0.7416,
|
| 1241 |
-
"step": 88000
|
| 1242 |
-
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 0.82,
|
| 1245 |
-
"grad_norm": 261213.0625,
|
| 1246 |
-
"learning_rate": 1.0984917015895287e-05,
|
| 1247 |
-
"loss": 0.7705,
|
| 1248 |
-
"step": 88500
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 0.82,
|
| 1252 |
-
"grad_norm": 192903.46875,
|
| 1253 |
-
"learning_rate": 1.0707995643103734e-05,
|
| 1254 |
-
"loss": 0.7392,
|
| 1255 |
-
"step": 89000
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 0.83,
|
| 1259 |
-
"grad_norm": 682632.8125,
|
| 1260 |
-
"learning_rate": 1.0431074270312182e-05,
|
| 1261 |
-
"loss": 0.7577,
|
| 1262 |
-
"step": 89500
|
| 1263 |
-
},
|
| 1264 |
-
{
|
| 1265 |
-
"epoch": 0.83,
|
| 1266 |
-
"grad_norm": 218102.4375,
|
| 1267 |
-
"learning_rate": 1.0154152897520631e-05,
|
| 1268 |
-
"loss": 0.7586,
|
| 1269 |
-
"step": 90000
|
| 1270 |
-
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 0.84,
|
| 1273 |
-
"grad_norm": 212267.5625,
|
| 1274 |
-
"learning_rate": 9.877231524729078e-06,
|
| 1275 |
-
"loss": 0.7472,
|
| 1276 |
-
"step": 90500
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 0.84,
|
| 1280 |
-
"grad_norm": 320534.59375,
|
| 1281 |
-
"learning_rate": 9.600310151937527e-06,
|
| 1282 |
-
"loss": 0.7602,
|
| 1283 |
-
"step": 91000
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 0.84,
|
| 1287 |
-
"grad_norm": 295994.96875,
|
| 1288 |
-
"learning_rate": 9.323388779145974e-06,
|
| 1289 |
-
"loss": 0.7368,
|
| 1290 |
-
"step": 91500
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 0.85,
|
| 1294 |
-
"grad_norm": 221153.6875,
|
| 1295 |
-
"learning_rate": 9.046467406354423e-06,
|
| 1296 |
-
"loss": 0.7492,
|
| 1297 |
-
"step": 92000
|
| 1298 |
-
},
|
| 1299 |
-
{
|
| 1300 |
-
"epoch": 0.85,
|
| 1301 |
-
"grad_norm": 250322.6875,
|
| 1302 |
-
"learning_rate": 8.76954603356287e-06,
|
| 1303 |
-
"loss": 0.7582,
|
| 1304 |
-
"step": 92500
|
| 1305 |
-
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 0.86,
|
| 1308 |
-
"grad_norm": 364772.46875,
|
| 1309 |
-
"learning_rate": 8.492624660771318e-06,
|
| 1310 |
-
"loss": 0.747,
|
| 1311 |
-
"step": 93000
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 0.86,
|
| 1315 |
-
"grad_norm": 226077.453125,
|
| 1316 |
-
"learning_rate": 8.215703287979767e-06,
|
| 1317 |
-
"loss": 0.7515,
|
| 1318 |
-
"step": 93500
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 0.87,
|
| 1322 |
-
"grad_norm": 260477.265625,
|
| 1323 |
-
"learning_rate": 7.938781915188214e-06,
|
| 1324 |
-
"loss": 0.7402,
|
| 1325 |
-
"step": 94000
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 0.87,
|
| 1329 |
-
"grad_norm": 380977.46875,
|
| 1330 |
-
"learning_rate": 7.661860542396663e-06,
|
| 1331 |
-
"loss": 0.7512,
|
| 1332 |
-
"step": 94500
|
| 1333 |
-
},
|
| 1334 |
-
{
|
| 1335 |
-
"epoch": 0.88,
|
| 1336 |
-
"grad_norm": 1052145.875,
|
| 1337 |
-
"learning_rate": 7.38493916960511e-06,
|
| 1338 |
-
"loss": 0.7589,
|
| 1339 |
-
"step": 95000
|
| 1340 |
-
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 0.88,
|
| 1343 |
-
"grad_norm": 213601.171875,
|
| 1344 |
-
"learning_rate": 7.1080177968135575e-06,
|
| 1345 |
-
"loss": 0.7581,
|
| 1346 |
-
"step": 95500
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 0.89,
|
| 1350 |
-
"grad_norm": 467877.71875,
|
| 1351 |
-
"learning_rate": 6.831096424022006e-06,
|
| 1352 |
-
"loss": 0.7464,
|
| 1353 |
-
"step": 96000
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 0.89,
|
| 1357 |
-
"grad_norm": 566204.3125,
|
| 1358 |
-
"learning_rate": 6.554175051230454e-06,
|
| 1359 |
-
"loss": 0.7557,
|
| 1360 |
-
"step": 96500
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 0.9,
|
| 1364 |
-
"grad_norm": 330177.5625,
|
| 1365 |
-
"learning_rate": 6.277253678438902e-06,
|
| 1366 |
-
"loss": 0.7304,
|
| 1367 |
-
"step": 97000
|
| 1368 |
-
},
|
| 1369 |
-
{
|
| 1370 |
-
"epoch": 0.9,
|
| 1371 |
-
"grad_norm": 319206.65625,
|
| 1372 |
-
"learning_rate": 6.00033230564735e-06,
|
| 1373 |
-
"loss": 0.736,
|
| 1374 |
-
"step": 97500
|
| 1375 |
-
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 0.9,
|
| 1378 |
-
"grad_norm": 410904.90625,
|
| 1379 |
-
"learning_rate": 5.723410932855798e-06,
|
| 1380 |
-
"loss": 0.752,
|
| 1381 |
-
"step": 98000
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 0.91,
|
| 1385 |
-
"grad_norm": 375357.15625,
|
| 1386 |
-
"learning_rate": 5.4464895600642454e-06,
|
| 1387 |
-
"loss": 0.7358,
|
| 1388 |
-
"step": 98500
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 0.91,
|
| 1392 |
-
"grad_norm": 307483.375,
|
| 1393 |
-
"learning_rate": 5.169568187272693e-06,
|
| 1394 |
-
"loss": 0.7749,
|
| 1395 |
-
"step": 99000
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 0.92,
|
| 1399 |
-
"grad_norm": 434196.59375,
|
| 1400 |
-
"learning_rate": 4.892646814481141e-06,
|
| 1401 |
-
"loss": 0.7324,
|
| 1402 |
-
"step": 99500
|
| 1403 |
-
},
|
| 1404 |
-
{
|
| 1405 |
-
"epoch": 0.92,
|
| 1406 |
-
"grad_norm": 554119.4375,
|
| 1407 |
-
"learning_rate": 4.61572544168959e-06,
|
| 1408 |
-
"loss": 0.7096,
|
| 1409 |
-
"step": 100000
|
| 1410 |
-
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 0.93,
|
| 1413 |
-
"grad_norm": 224006.78125,
|
| 1414 |
-
"learning_rate": 4.338804068898038e-06,
|
| 1415 |
-
"loss": 0.7205,
|
| 1416 |
-
"step": 100500
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 0.93,
|
| 1420 |
-
"grad_norm": 215164.78125,
|
| 1421 |
-
"learning_rate": 4.0618826961064855e-06,
|
| 1422 |
-
"loss": 0.732,
|
| 1423 |
-
"step": 101000
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 0.94,
|
| 1427 |
-
"grad_norm": 391820.53125,
|
| 1428 |
-
"learning_rate": 3.7849613233149334e-06,
|
| 1429 |
-
"loss": 0.7228,
|
| 1430 |
-
"step": 101500
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 0.94,
|
| 1434 |
-
"grad_norm": 306694.15625,
|
| 1435 |
-
"learning_rate": 3.5080399505233816e-06,
|
| 1436 |
-
"loss": 0.7507,
|
| 1437 |
-
"step": 102000
|
| 1438 |
-
},
|
| 1439 |
-
{
|
| 1440 |
-
"epoch": 0.95,
|
| 1441 |
-
"grad_norm": 467995.03125,
|
| 1442 |
-
"learning_rate": 3.2311185777318295e-06,
|
| 1443 |
-
"loss": 0.7433,
|
| 1444 |
-
"step": 102500
|
| 1445 |
-
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 0.95,
|
| 1448 |
-
"grad_norm": 284427.625,
|
| 1449 |
-
"learning_rate": 2.9541972049402773e-06,
|
| 1450 |
-
"loss": 0.7047,
|
| 1451 |
-
"step": 103000
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 0.96,
|
| 1455 |
-
"grad_norm": 193486.921875,
|
| 1456 |
-
"learning_rate": 2.6772758321487256e-06,
|
| 1457 |
-
"loss": 0.7646,
|
| 1458 |
-
"step": 103500
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 0.96,
|
| 1462 |
-
"grad_norm": 262856.15625,
|
| 1463 |
-
"learning_rate": 2.4003544593571735e-06,
|
| 1464 |
-
"loss": 0.7306,
|
| 1465 |
-
"step": 104000
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 0.96,
|
| 1469 |
-
"grad_norm": 267263.5,
|
| 1470 |
-
"learning_rate": 2.1234330865656213e-06,
|
| 1471 |
-
"loss": 0.7631,
|
| 1472 |
-
"step": 104500
|
| 1473 |
-
},
|
| 1474 |
-
{
|
| 1475 |
-
"epoch": 0.97,
|
| 1476 |
-
"grad_norm": 229738.515625,
|
| 1477 |
-
"learning_rate": 1.8465117137740692e-06,
|
| 1478 |
-
"loss": 0.731,
|
| 1479 |
-
"step": 105000
|
| 1480 |
-
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 0.97,
|
| 1483 |
-
"grad_norm": 528997.6875,
|
| 1484 |
-
"learning_rate": 1.569590340982517e-06,
|
| 1485 |
-
"loss": 0.7393,
|
| 1486 |
-
"step": 105500
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 0.98,
|
| 1490 |
-
"grad_norm": 346399.28125,
|
| 1491 |
-
"learning_rate": 1.292668968190965e-06,
|
| 1492 |
-
"loss": 0.7304,
|
| 1493 |
-
"step": 106000
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 0.98,
|
| 1497 |
-
"grad_norm": 315753.28125,
|
| 1498 |
-
"learning_rate": 1.015747595399413e-06,
|
| 1499 |
-
"loss": 0.7239,
|
| 1500 |
-
"step": 106500
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 0.99,
|
| 1504 |
-
"grad_norm": 245733.109375,
|
| 1505 |
-
"learning_rate": 7.388262226078609e-07,
|
| 1506 |
-
"loss": 0.7067,
|
| 1507 |
-
"step": 107000
|
| 1508 |
-
},
|
| 1509 |
-
{
|
| 1510 |
-
"epoch": 0.99,
|
| 1511 |
-
"grad_norm": 423298.15625,
|
| 1512 |
-
"learning_rate": 4.619048498163089e-07,
|
| 1513 |
-
"loss": 0.734,
|
| 1514 |
-
"step": 107500
|
| 1515 |
},
|
| 1516 |
{
|
| 1517 |
"epoch": 1.0,
|
| 1518 |
-
"grad_norm":
|
| 1519 |
-
"learning_rate":
|
| 1520 |
-
"loss":
|
| 1521 |
-
"step":
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 1.0,
|
| 1525 |
-
"eval_accuracy": 0.
|
| 1526 |
-
"eval_loss":
|
| 1527 |
-
"eval_runtime":
|
| 1528 |
-
"eval_samples_per_second": 17.
|
| 1529 |
-
"eval_steps_per_second":
|
| 1530 |
-
"step":
|
| 1531 |
}
|
| 1532 |
],
|
| 1533 |
"logging_steps": 500,
|
| 1534 |
-
"max_steps":
|
| 1535 |
"num_input_tokens_seen": 0,
|
| 1536 |
"num_train_epochs": 1,
|
| 1537 |
"save_steps": 500,
|
| 1538 |
-
"total_flos":
|
| 1539 |
-
"train_batch_size":
|
| 1540 |
"trial_name": null,
|
| 1541 |
"trial_params": null
|
| 1542 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 1.240515112876892,
|
| 3 |
+
"best_model_checkpoint": "sean_test_out_large/checkpoint-1000",
|
| 4 |
"epoch": 1.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.5,
|
| 13 |
+
"grad_norm": 971037.375,
|
| 14 |
+
"learning_rate": 3e-05,
|
| 15 |
+
"loss": 1.7354,
|
| 16 |
+
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 1.0,
|
| 20 |
+
"grad_norm": 247651.375,
|
| 21 |
+
"learning_rate": 0.0,
|
| 22 |
+
"loss": 1.3372,
|
| 23 |
+
"step": 1000
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 1.0,
|
| 27 |
+
"eval_accuracy": 0.4668,
|
| 28 |
+
"eval_loss": 1.240515112876892,
|
| 29 |
+
"eval_runtime": 573.2037,
|
| 30 |
+
"eval_samples_per_second": 17.446,
|
| 31 |
+
"eval_steps_per_second": 8.723,
|
| 32 |
+
"step": 1000
|
| 33 |
}
|
| 34 |
],
|
| 35 |
"logging_steps": 500,
|
| 36 |
+
"max_steps": 1000,
|
| 37 |
"num_input_tokens_seen": 0,
|
| 38 |
"num_train_epochs": 1,
|
| 39 |
"save_steps": 500,
|
| 40 |
+
"total_flos": 4315155738240000.0,
|
| 41 |
+
"train_batch_size": 2,
|
| 42 |
"trial_name": null,
|
| 43 |
"trial_params": null
|
| 44 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4603
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45ad6de8855d7a0a11d6a75da560249dad40620168d7086da168acfd7b2a271d
|
| 3 |
size 4603
|