Model save
Browse files- README.md +3 -3
- all_results.json +8 -8
- eval_results.json +4 -4
- train_results.json +4 -4
- trainer_state.json +457 -457
README.md
CHANGED
|
@@ -4,8 +4,8 @@ base_model: yanolja/EEVE-Korean-Instruct-2.8B-v1.0
|
|
| 4 |
tags:
|
| 5 |
- trl
|
| 6 |
- sft
|
| 7 |
-
- generated_from_trainer
|
| 8 |
- alignment-handbook
|
|
|
|
| 9 |
model-index:
|
| 10 |
- name: eeve-2.8B_C
|
| 11 |
results: []
|
|
@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [yanolja/EEVE-Korean-Instruct-2.8B-v1.0](https://huggingface.co/yanolja/EEVE-Korean-Instruct-2.8B-v1.0) on the None dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -54,7 +54,7 @@ The following hyperparameters were used during training:
|
|
| 54 |
|
| 55 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 56 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 57 |
-
| 0.
|
| 58 |
|
| 59 |
|
| 60 |
### Framework versions
|
|
|
|
| 4 |
tags:
|
| 5 |
- trl
|
| 6 |
- sft
|
|
|
|
| 7 |
- alignment-handbook
|
| 8 |
+
- generated_from_trainer
|
| 9 |
model-index:
|
| 10 |
- name: eeve-2.8B_C
|
| 11 |
results: []
|
|
|
|
| 18 |
|
| 19 |
This model is a fine-tuned version of [yanolja/EEVE-Korean-Instruct-2.8B-v1.0](https://huggingface.co/yanolja/EEVE-Korean-Instruct-2.8B-v1.0) on the None dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.3932
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 54 |
|
| 55 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 56 |
|:-------------:|:-----:|:----:|:---------------:|
|
| 57 |
+
| 0.3021 | 1.0 | 233 | 0.3932 |
|
| 58 |
|
| 59 |
|
| 60 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime": 1.
|
| 5 |
"eval_samples": 230,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 6.
|
| 8 |
"total_flos": 1.933383782157517e+16,
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime": 303.
|
| 11 |
"train_samples": 11140,
|
| 12 |
-
"train_samples_per_second": 36.
|
| 13 |
-
"train_steps_per_second": 0.
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 0.3927557170391083,
|
| 4 |
+
"eval_runtime": 1.5271,
|
| 5 |
"eval_samples": 230,
|
| 6 |
+
"eval_samples_per_second": 150.611,
|
| 7 |
+
"eval_steps_per_second": 6.548,
|
| 8 |
"total_flos": 1.933383782157517e+16,
|
| 9 |
+
"train_loss": 0.43314852683840904,
|
| 10 |
+
"train_runtime": 303.7389,
|
| 11 |
"train_samples": 11140,
|
| 12 |
+
"train_samples_per_second": 36.676,
|
| 13 |
+
"train_steps_per_second": 0.767
|
| 14 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
-
"eval_loss": 0.
|
| 4 |
-
"eval_runtime": 1.
|
| 5 |
"eval_samples": 230,
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 6.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
+
"eval_loss": 0.3927557170391083,
|
| 4 |
+
"eval_runtime": 1.5271,
|
| 5 |
"eval_samples": 230,
|
| 6 |
+
"eval_samples_per_second": 150.611,
|
| 7 |
+
"eval_steps_per_second": 6.548
|
| 8 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"total_flos": 1.933383782157517e+16,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime": 303.
|
| 6 |
"train_samples": 11140,
|
| 7 |
-
"train_samples_per_second": 36.
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.0,
|
| 3 |
"total_flos": 1.933383782157517e+16,
|
| 4 |
+
"train_loss": 0.43314852683840904,
|
| 5 |
+
"train_runtime": 303.7389,
|
| 6 |
"train_samples": 11140,
|
| 7 |
+
"train_samples_per_second": 36.676,
|
| 8 |
+
"train_steps_per_second": 0.767
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -10,1651 +10,1651 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004291845493562232,
|
| 13 |
-
"grad_norm": 6.
|
| 14 |
"learning_rate": 4.1666666666666667e-07,
|
| 15 |
"loss": 2.3401,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008583690987124463,
|
| 20 |
-
"grad_norm": 7.
|
| 21 |
"learning_rate": 8.333333333333333e-07,
|
| 22 |
"loss": 2.3774,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.012875536480686695,
|
| 27 |
-
"grad_norm":
|
| 28 |
"learning_rate": 1.25e-06,
|
| 29 |
-
"loss": 2.
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.017167381974248927,
|
| 34 |
-
"grad_norm": 6.
|
| 35 |
"learning_rate": 1.6666666666666667e-06,
|
| 36 |
-
"loss": 2.
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02145922746781116,
|
| 41 |
-
"grad_norm": 6.
|
| 42 |
"learning_rate": 2.0833333333333334e-06,
|
| 43 |
-
"loss": 2.
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02575107296137339,
|
| 48 |
-
"grad_norm": 6.
|
| 49 |
"learning_rate": 2.5e-06,
|
| 50 |
-
"loss": 2.
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.030042918454935622,
|
| 55 |
-
"grad_norm": 6.
|
| 56 |
"learning_rate": 2.916666666666667e-06,
|
| 57 |
-
"loss": 2.
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.034334763948497854,
|
| 62 |
-
"grad_norm": 5.
|
| 63 |
"learning_rate": 3.3333333333333333e-06,
|
| 64 |
-
"loss": 2.
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03862660944206009,
|
| 69 |
-
"grad_norm": 5.
|
| 70 |
"learning_rate": 3.7500000000000005e-06,
|
| 71 |
-
"loss": 2.
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04291845493562232,
|
| 76 |
-
"grad_norm": 5.
|
| 77 |
"learning_rate": 4.166666666666667e-06,
|
| 78 |
-
"loss": 2.
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04721030042918455,
|
| 83 |
-
"grad_norm": 4.
|
| 84 |
"learning_rate": 4.583333333333333e-06,
|
| 85 |
-
"loss": 1.
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.05150214592274678,
|
| 90 |
-
"grad_norm": 4.
|
| 91 |
"learning_rate": 5e-06,
|
| 92 |
-
"loss": 1.
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.055793991416309016,
|
| 97 |
-
"grad_norm": 4.
|
| 98 |
"learning_rate": 5.416666666666667e-06,
|
| 99 |
-
"loss": 1.
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.060085836909871244,
|
| 104 |
-
"grad_norm": 3.
|
| 105 |
"learning_rate": 5.833333333333334e-06,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06437768240343347,
|
| 111 |
-
"grad_norm": 3.
|
| 112 |
"learning_rate": 6.25e-06,
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.06866952789699571,
|
| 118 |
-
"grad_norm": 3.
|
| 119 |
"learning_rate": 6.666666666666667e-06,
|
| 120 |
-
"loss": 1.
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07296137339055794,
|
| 125 |
-
"grad_norm": 2.
|
| 126 |
"learning_rate": 7.083333333333335e-06,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07725321888412018,
|
| 132 |
-
"grad_norm": 2.
|
| 133 |
"learning_rate": 7.500000000000001e-06,
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0815450643776824,
|
| 139 |
-
"grad_norm": 2.
|
| 140 |
"learning_rate": 7.916666666666667e-06,
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08583690987124463,
|
| 146 |
-
"grad_norm": 1.
|
| 147 |
"learning_rate": 8.333333333333334e-06,
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09012875536480687,
|
| 153 |
-
"grad_norm": 1.
|
| 154 |
"learning_rate": 8.750000000000001e-06,
|
| 155 |
-
"loss": 0.
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.0944206008583691,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
"learning_rate": 9.166666666666666e-06,
|
| 162 |
-
"loss": 0.
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09871244635193133,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 9.583333333333335e-06,
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.10300429184549356,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
"learning_rate": 1e-05,
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.1072961373390558,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
"learning_rate": 9.999435142363484e-06,
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.11158798283261803,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
"learning_rate": 9.997740697079595e-06,
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.11587982832618025,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
"learning_rate": 9.994917046996472e-06,
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.12017167381974249,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
"learning_rate": 9.990964830098246e-06,
|
| 204 |
-
"loss": 0.
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.12446351931330472,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
"learning_rate": 9.985884939360873e-06,
|
| 211 |
-
"loss": 0.
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.12875536480686695,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
"learning_rate": 9.979678522550382e-06,
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.13304721030042918,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
"learning_rate": 9.972346981963546e-06,
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.13733905579399142,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
"learning_rate": 9.963891974111042e-06,
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.14163090128755365,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
"learning_rate": 9.95431540934317e-06,
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.1459227467811159,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
"learning_rate": 9.943619451418225e-06,
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.15021459227467812,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
"learning_rate": 9.931806517013612e-06,
|
| 253 |
-
"loss": 0.
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.15450643776824036,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 9.918879275179819e-06,
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15879828326180256,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
"learning_rate": 9.904840646737346e-06,
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1630901287553648,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
"learning_rate": 9.889693803616793e-06,
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.16738197424892703,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
"learning_rate": 9.873442168142158e-06,
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.17167381974248927,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
"learning_rate": 9.856089412257605e-06,
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.1759656652360515,
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
"learning_rate": 9.837639456697802e-06,
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.18025751072961374,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
"learning_rate": 9.818096470102067e-06,
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.18454935622317598,
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
"learning_rate": 9.797464868072489e-06,
|
| 309 |
-
"loss": 0.
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.1888412017167382,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
"learning_rate": 9.775749312176249e-06,
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.19313304721030042,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
"learning_rate": 9.752954708892379e-06,
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.19742489270386265,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
"learning_rate": 9.729086208503174e-06,
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.2017167381974249,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
"learning_rate": 9.704149203930522e-06,
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.20600858369098712,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
"learning_rate": 9.67814932951741e-06,
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.21030042918454936,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
"learning_rate": 9.651092459754879e-06,
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.2145922746781116,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
"learning_rate": 9.622984707954732e-06,
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.21888412017167383,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
"learning_rate": 9.593832424868271e-06,
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.22317596566523606,
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
"learning_rate": 9.563642197251382e-06,
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.22746781115879827,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
"learning_rate": 9.532420846376316e-06,
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.2317596566523605,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
"learning_rate": 9.500175426490455e-06,
|
| 386 |
-
"loss": 0.
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.23605150214592274,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
"learning_rate": 9.466913223222467e-06,
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.24034334763948498,
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
"learning_rate": 9.432641751936162e-06,
|
| 400 |
-
"loss": 0.
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2446351931330472,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
"learning_rate": 9.397368756032445e-06,
|
| 407 |
"loss": 0.3435,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.24892703862660945,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
"learning_rate": 9.361102205199762e-06,
|
| 414 |
-
"loss": 0.
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.2532188841201717,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
"learning_rate": 9.32385029361338e-06,
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2575107296137339,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
"learning_rate": 9.285621438083997e-06,
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.26180257510729615,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
"learning_rate": 9.246424276156008e-06,
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.26609442060085836,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
"learning_rate": 9.206267664155906e-06,
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.2703862660944206,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
"learning_rate": 9.165160675191272e-06,
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.27467811158798283,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
"learning_rate": 9.123112597100759e-06,
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.27896995708154504,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
"learning_rate": 9.080132930355567e-06,
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2832618025751073,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
"learning_rate": 9.03623138591289e-06,
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.2875536480686695,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 8.99141788302178e-06,
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2918454935622318,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
"learning_rate": 8.94570254698197e-06,
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.296137339055794,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 8.899095706856122e-06,
|
| 491 |
-
"loss": 0.
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.30042918454935624,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
"learning_rate": 8.851607893136065e-06,
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.30472103004291845,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
"learning_rate": 8.803249835363486e-06,
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.3090128755364807,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
"learning_rate": 8.754032459705672e-06,
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.3133047210300429,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 8.703966886486819e-06,
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.31759656652360513,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
"learning_rate": 8.65306442767547e-06,
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.3218884120171674,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
"learning_rate": 8.601336584328659e-06,
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3261802575107296,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 8.548795043993316e-06,
|
| 540 |
-
"loss": 0.
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.33047210300429186,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
"learning_rate": 8.495451678065563e-06,
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.33476394849785407,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
"learning_rate": 8.441318539108433e-06,
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.33905579399141633,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
"learning_rate": 8.386407858128707e-06,
|
| 561 |
-
"loss": 0.
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.34334763948497854,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
"learning_rate": 8.330732041813367e-06,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.34763948497854075,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 8.274303669726427e-06,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.351931330472103,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 8.217135491466636e-06,
|
| 582 |
-
"loss": 0.
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.3562231759656652,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 8.15924042378682e-06,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.3605150214592275,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 8.100631547675417e-06,
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3648068669527897,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 8.041322105400923e-06,
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.36909871244635195,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 7.981325497519892e-06,
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.37339055793991416,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 7.920655279849173e-06,
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.3776824034334764,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
"learning_rate": 7.859325160403073e-06,
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.38197424892703863,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
"learning_rate": 7.797348996296116e-06,
|
| 631 |
-
"loss": 0.
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.38626609442060084,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
"learning_rate": 7.734740790612137e-06,
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.3905579399141631,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
"learning_rate": 7.671514689240366e-06,
|
| 645 |
-
"loss": 0.
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.3948497854077253,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
"learning_rate": 7.607684977679284e-06,
|
| 652 |
-
"loss": 0.
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.39914163090128757,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
"learning_rate": 7.543266077808893e-06,
|
| 659 |
-
"loss": 0.
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.4034334763948498,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
"learning_rate": 7.478272544632204e-06,
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.40772532188841204,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
"learning_rate": 7.412719062986632e-06,
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.41201716738197425,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
"learning_rate": 7.3466204442260605e-06,
|
| 680 |
"loss": 0.276,
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.41630901287553645,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
"learning_rate": 7.279991622874319e-06,
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.4206008583690987,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
"learning_rate": 7.212847653250828e-06,
|
| 694 |
-
"loss": 0.
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.4248927038626609,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
"learning_rate": 7.145203706069183e-06,
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4291845493562232,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
"learning_rate": 7.0770750650094335e-06,
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.4334763948497854,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
"learning_rate": 7.008477123264849e-06,
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.43776824034334766,
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
"learning_rate": 6.939425380063924e-06,
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.44206008583690987,
|
| 727 |
-
"grad_norm": 0.
|
| 728 |
"learning_rate": 6.869935437168449e-06,
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.44635193133047213,
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
"learning_rate": 6.800022995348381e-06,
|
| 736 |
"loss": 0.3182,
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.45064377682403434,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
"learning_rate": 6.729703850834381e-06,
|
| 743 |
-
"loss": 0.
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.45493562231759654,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
"learning_rate": 6.65899389174876e-06,
|
| 750 |
-
"loss": 0.
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.4592274678111588,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 6.587909094515663e-06,
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.463519313304721,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
"learning_rate": 6.5164655202513135e-06,
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4678111587982833,
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
"learning_rate": 6.444679311135112e-06,
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.4721030042918455,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
"learning_rate": 6.372566686762427e-06,
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.47639484978540775,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
"learning_rate": 6.300143940479881e-06,
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.48068669527896996,
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
"learning_rate": 6.227427435703997e-06,
|
| 792 |
"loss": 0.2892,
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.48497854077253216,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
"learning_rate": 6.154433602223979e-06,
|
| 799 |
-
"loss": 0.
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4892703862660944,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
"learning_rate": 6.0811789324895365e-06,
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.49356223175965663,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
"learning_rate": 6.0076799778845105e-06,
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4978540772532189,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
"learning_rate": 5.933953344987215e-06,
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.5021459227467812,
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
"learning_rate": 5.860015691818292e-06,
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.5064377682403434,
|
| 832 |
-
"grad_norm": 0.
|
| 833 |
"learning_rate": 5.78588372407695e-06,
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.5107296137339056,
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
"learning_rate": 5.711574191366427e-06,
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.5150214592274678,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
"learning_rate": 5.637103883409525e-06,
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.51931330472103,
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
"learning_rate": 5.562489626255104e-06,
|
| 855 |
-
"loss": 0.
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.5236051502145923,
|
| 860 |
-
"grad_norm": 0.
|
| 861 |
"learning_rate": 5.487748278476342e-06,
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.5278969957081545,
|
| 867 |
-
"grad_norm": 0.
|
| 868 |
"learning_rate": 5.412896727361663e-06,
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.5321888412017167,
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
"learning_rate": 5.337951885099167e-06,
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.5364806866952789,
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
"learning_rate": 5.262930684955439e-06,
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.5407725321888412,
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
"learning_rate": 5.187850077449604e-06,
|
| 890 |
"loss": 0.317,
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.5450643776824035,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
"learning_rate": 5.112727026523461e-06,
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.5493562231759657,
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
"learning_rate": 5.03757850570861e-06,
|
| 904 |
"loss": 0.3066,
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.5536480686695279,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
"learning_rate": 4.9624214942913916e-06,
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.5579399141630901,
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
"learning_rate": 4.88727297347654e-06,
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.5622317596566524,
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
"learning_rate": 4.8121499225503974e-06,
|
| 925 |
"loss": 0.2874,
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.5665236051502146,
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
"learning_rate": 4.737069315044562e-06,
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.5708154506437768,
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
"learning_rate": 4.662048114900837e-06,
|
| 939 |
-
"loss": 0.
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.575107296137339,
|
| 944 |
-
"grad_norm": 0.
|
| 945 |
"learning_rate": 4.587103272638339e-06,
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.5793991416309013,
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
"learning_rate": 4.512251721523659e-06,
|
| 953 |
"loss": 0.2671,
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.5836909871244635,
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
"learning_rate": 4.437510373744897e-06,
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.5879828326180258,
|
| 965 |
-
"grad_norm": 0.
|
| 966 |
"learning_rate": 4.362896116590475e-06,
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.592274678111588,
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
"learning_rate": 4.2884258086335755e-06,
|
| 974 |
-
"loss": 0.
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.5965665236051502,
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
"learning_rate": 4.214116275923051e-06,
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.6008583690987125,
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
"learning_rate": 4.1399843081817085e-06,
|
| 988 |
-
"loss": 0.
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.6051502145922747,
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
"learning_rate": 4.066046655012786e-06,
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.6094420600858369,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
"learning_rate": 3.992320022115492e-06,
|
| 1002 |
-
"loss": 0.
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.6137339055793991,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
"learning_rate": 3.918821067510464e-06,
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.6180257510729614,
|
| 1014 |
-
"grad_norm": 0.
|
| 1015 |
"learning_rate": 3.845566397776022e-06,
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.6223175965665236,
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
"learning_rate": 3.7725725642960047e-06,
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.6266094420600858,
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
"learning_rate": 3.6998560595201188e-06,
|
| 1030 |
-
"loss": 0.
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.630901287553648,
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
"learning_rate": 3.627433313237576e-06,
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.6351931330472103,
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
"learning_rate": 3.555320688864889e-06,
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.6394849785407726,
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
"learning_rate": 3.483534479748688e-06,
|
| 1051 |
-
"loss": 0.
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6437768240343348,
|
| 1056 |
-
"grad_norm": 0.
|
| 1057 |
"learning_rate": 3.4120909054843375e-06,
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.648068669527897,
|
| 1063 |
-
"grad_norm": 0.
|
| 1064 |
"learning_rate": 3.3410061082512422e-06,
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6523605150214592,
|
| 1070 |
-
"grad_norm": 0.
|
| 1071 |
"learning_rate": 3.2702961491656197e-06,
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6566523605150214,
|
| 1077 |
-
"grad_norm": 0.
|
| 1078 |
"learning_rate": 3.1999770046516198e-06,
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6609442060085837,
|
| 1084 |
-
"grad_norm": 0.
|
| 1085 |
"learning_rate": 3.130064562831553e-06,
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6652360515021459,
|
| 1091 |
-
"grad_norm": 0.
|
| 1092 |
"learning_rate": 3.0605746199360755e-06,
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6695278969957081,
|
| 1098 |
-
"grad_norm": 0.
|
| 1099 |
"learning_rate": 2.991522876735154e-06,
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6738197424892703,
|
| 1105 |
-
"grad_norm": 0.
|
| 1106 |
"learning_rate": 2.9229249349905686e-06,
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.6781115879828327,
|
| 1112 |
-
"grad_norm": 0.
|
| 1113 |
"learning_rate": 2.8547962939308187e-06,
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6824034334763949,
|
| 1119 |
-
"grad_norm": 0.
|
| 1120 |
"learning_rate": 2.787152346749173e-06,
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6866952789699571,
|
| 1126 |
-
"grad_norm": 0.
|
| 1127 |
"learning_rate": 2.720008377125682e-06,
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.6909871244635193,
|
| 1133 |
-
"grad_norm": 0.
|
| 1134 |
"learning_rate": 2.6533795557739407e-06,
|
| 1135 |
"loss": 0.3238,
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.6952789699570815,
|
| 1140 |
-
"grad_norm": 0.
|
| 1141 |
"learning_rate": 2.5872809370133704e-06,
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.6995708154506438,
|
| 1147 |
-
"grad_norm": 0.
|
| 1148 |
"learning_rate": 2.5217274553677975e-06,
|
| 1149 |
"loss": 0.2985,
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.703862660944206,
|
| 1154 |
-
"grad_norm": 0.
|
| 1155 |
"learning_rate": 2.4567339221911086e-06,
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7081545064377682,
|
| 1161 |
-
"grad_norm": 0.
|
| 1162 |
"learning_rate": 2.3923150223207176e-06,
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7124463519313304,
|
| 1168 |
-
"grad_norm": 0.
|
| 1169 |
"learning_rate": 2.328485310759635e-06,
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7167381974248928,
|
| 1175 |
-
"grad_norm": 0.
|
| 1176 |
"learning_rate": 2.265259209387867e-06,
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.721030042918455,
|
| 1182 |
-
"grad_norm": 0.
|
| 1183 |
"learning_rate": 2.202651003703885e-06,
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7253218884120172,
|
| 1189 |
-
"grad_norm": 0.
|
| 1190 |
"learning_rate": 2.140674839596931e-06,
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7296137339055794,
|
| 1196 |
-
"grad_norm": 0.
|
| 1197 |
"learning_rate": 2.0793447201508288e-06,
|
| 1198 |
-
"loss": 0.
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.7339055793991416,
|
| 1203 |
-
"grad_norm": 0.
|
| 1204 |
"learning_rate": 2.01867450248011e-06,
|
| 1205 |
-
"loss": 0.
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.7381974248927039,
|
| 1210 |
-
"grad_norm": 0.
|
| 1211 |
"learning_rate": 1.9586778945990785e-06,
|
| 1212 |
-
"loss": 0.
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.7424892703862661,
|
| 1217 |
-
"grad_norm": 0.
|
| 1218 |
"learning_rate": 1.8993684523245842e-06,
|
| 1219 |
-
"loss": 0.
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
"epoch": 0.7467811158798283,
|
| 1224 |
-
"grad_norm": 0.
|
| 1225 |
"learning_rate": 1.8407595762131814e-06,
|
| 1226 |
-
"loss": 0.
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.7510729613733905,
|
| 1231 |
-
"grad_norm": 0.
|
| 1232 |
"learning_rate": 1.7828645085333645e-06,
|
| 1233 |
-
"loss": 0.
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
"epoch": 0.7553648068669528,
|
| 1238 |
-
"grad_norm": 0.
|
| 1239 |
"learning_rate": 1.7256963302735752e-06,
|
| 1240 |
-
"loss": 0.
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.759656652360515,
|
| 1245 |
-
"grad_norm": 0.
|
| 1246 |
"learning_rate": 1.6692679581866334e-06,
|
| 1247 |
-
"loss": 0.
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
"epoch": 0.7639484978540773,
|
| 1252 |
-
"grad_norm": 0.
|
| 1253 |
"learning_rate": 1.6135921418712959e-06,
|
| 1254 |
-
"loss": 0.
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.7682403433476395,
|
| 1259 |
-
"grad_norm": 0.
|
| 1260 |
"learning_rate": 1.5586814608915673e-06,
|
| 1261 |
-
"loss": 0.
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"epoch": 0.7725321888412017,
|
| 1266 |
-
"grad_norm": 0.
|
| 1267 |
"learning_rate": 1.5045483219344387e-06,
|
| 1268 |
-
"loss": 0.
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.776824034334764,
|
| 1273 |
-
"grad_norm": 0.
|
| 1274 |
"learning_rate": 1.4512049560066837e-06,
|
| 1275 |
"loss": 0.2758,
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.7811158798283262,
|
| 1280 |
-
"grad_norm": 0.
|
| 1281 |
"learning_rate": 1.3986634156713418e-06,
|
| 1282 |
-
"loss": 0.
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.7854077253218884,
|
| 1287 |
-
"grad_norm": 0.
|
| 1288 |
"learning_rate": 1.3469355723245303e-06,
|
| 1289 |
-
"loss": 0.
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
"epoch": 0.7896995708154506,
|
| 1294 |
-
"grad_norm": 0.
|
| 1295 |
"learning_rate": 1.2960331135131826e-06,
|
| 1296 |
-
"loss": 0.
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.7939914163090128,
|
| 1301 |
-
"grad_norm": 0.
|
| 1302 |
"learning_rate": 1.245967540294329e-06,
|
| 1303 |
-
"loss": 0.
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
"epoch": 0.7982832618025751,
|
| 1308 |
-
"grad_norm": 0.
|
| 1309 |
"learning_rate": 1.1967501646365147e-06,
|
| 1310 |
-
"loss": 0.
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.8025751072961373,
|
| 1315 |
-
"grad_norm": 0.
|
| 1316 |
"learning_rate": 1.1483921068639353e-06,
|
| 1317 |
-
"loss": 0.
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
"epoch": 0.8068669527896996,
|
| 1322 |
-
"grad_norm": 0.
|
| 1323 |
"learning_rate": 1.1009042931438784e-06,
|
| 1324 |
-
"loss": 0.
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.8111587982832618,
|
| 1329 |
-
"grad_norm": 0.
|
| 1330 |
"learning_rate": 1.0542974530180327e-06,
|
| 1331 |
-
"loss": 0.
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.8154506437768241,
|
| 1336 |
-
"grad_norm": 0.
|
| 1337 |
"learning_rate": 1.00858211697822e-06,
|
| 1338 |
-
"loss": 0.
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 0.8197424892703863,
|
| 1343 |
-
"grad_norm": 0.
|
| 1344 |
"learning_rate": 9.637686140871121e-07,
|
| 1345 |
-
"loss": 0.
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
"epoch": 0.8240343347639485,
|
| 1350 |
-
"grad_norm": 0.
|
| 1351 |
"learning_rate": 9.198670696444339e-07,
|
| 1352 |
-
"loss": 0.
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.8283261802575107,
|
| 1357 |
-
"grad_norm": 0.
|
| 1358 |
"learning_rate": 8.768874028992431e-07,
|
| 1359 |
-
"loss": 0.
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 0.8326180257510729,
|
| 1364 |
-
"grad_norm": 0.
|
| 1365 |
"learning_rate": 8.348393248087289e-07,
|
| 1366 |
-
"loss": 0.
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 0.8369098712446352,
|
| 1371 |
-
"grad_norm": 0.
|
| 1372 |
"learning_rate": 7.937323358440935e-07,
|
| 1373 |
-
"loss": 0.
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
"epoch": 0.8412017167381974,
|
| 1378 |
-
"grad_norm": 0.
|
| 1379 |
"learning_rate": 7.535757238439939e-07,
|
| 1380 |
-
"loss": 0.
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
"epoch": 0.8454935622317596,
|
| 1385 |
-
"grad_norm": 0.
|
| 1386 |
"learning_rate": 7.143785619160026e-07,
|
| 1387 |
-
"loss": 0.
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
"epoch": 0.8497854077253219,
|
| 1392 |
-
"grad_norm": 0.
|
| 1393 |
"learning_rate": 6.761497063866207e-07,
|
| 1394 |
-
"loss": 0.
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 0.8540772532188842,
|
| 1399 |
-
"grad_norm": 0.
|
| 1400 |
"learning_rate": 6.388977948002406e-07,
|
| 1401 |
-
"loss": 0.
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
"epoch": 0.8583690987124464,
|
| 1406 |
-
"grad_norm": 0.
|
| 1407 |
"learning_rate": 6.026312439675553e-07,
|
| 1408 |
-
"loss": 0.
|
| 1409 |
"step": 200
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 0.8626609442060086,
|
| 1413 |
-
"grad_norm": 0.
|
| 1414 |
"learning_rate": 5.673582480638395e-07,
|
| 1415 |
-
"loss": 0.
|
| 1416 |
"step": 201
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 0.8669527896995708,
|
| 1420 |
-
"grad_norm": 0.
|
| 1421 |
"learning_rate": 5.330867767775333e-07,
|
| 1422 |
-
"loss": 0.
|
| 1423 |
"step": 202
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 0.871244635193133,
|
| 1427 |
-
"grad_norm": 0.
|
| 1428 |
"learning_rate": 4.998245735095459e-07,
|
| 1429 |
-
"loss": 0.
|
| 1430 |
"step": 203
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 0.8755364806866953,
|
| 1434 |
-
"grad_norm": 0.
|
| 1435 |
"learning_rate": 4.6757915362368567e-07,
|
| 1436 |
-
"loss": 0.
|
| 1437 |
"step": 204
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 0.8798283261802575,
|
| 1441 |
-
"grad_norm": 0.
|
| 1442 |
"learning_rate": 4.363578027486187e-07,
|
| 1443 |
-
"loss": 0.
|
| 1444 |
"step": 205
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 0.8841201716738197,
|
| 1448 |
-
"grad_norm": 0.
|
| 1449 |
"learning_rate": 4.0616757513173123e-07,
|
| 1450 |
"loss": 0.2828,
|
| 1451 |
"step": 206
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 0.8884120171673819,
|
| 1455 |
-
"grad_norm": 0.
|
| 1456 |
"learning_rate": 3.7701529204526856e-07,
|
| 1457 |
-
"loss": 0.
|
| 1458 |
"step": 207
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 0.8927038626609443,
|
| 1462 |
-
"grad_norm": 0.
|
| 1463 |
"learning_rate": 3.4890754024512254e-07,
|
| 1464 |
-
"loss": 0.
|
| 1465 |
"step": 208
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 0.8969957081545065,
|
| 1469 |
-
"grad_norm": 0.
|
| 1470 |
"learning_rate": 3.2185067048259245e-07,
|
| 1471 |
"loss": 0.2926,
|
| 1472 |
"step": 209
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 0.9012875536480687,
|
| 1476 |
-
"grad_norm": 0.
|
| 1477 |
"learning_rate": 2.9585079606947843e-07,
|
| 1478 |
-
"loss": 0.
|
| 1479 |
"step": 210
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 0.9055793991416309,
|
| 1483 |
-
"grad_norm": 0.
|
| 1484 |
"learning_rate": 2.7091379149682683e-07,
|
| 1485 |
-
"loss": 0.
|
| 1486 |
"step": 211
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 0.9098712446351931,
|
| 1490 |
-
"grad_norm": 0.
|
| 1491 |
"learning_rate": 2.470452911076227e-07,
|
| 1492 |
-
"loss": 0.
|
| 1493 |
"step": 212
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 0.9141630901287554,
|
| 1497 |
-
"grad_norm": 0.
|
| 1498 |
"learning_rate": 2.242506878237538e-07,
|
| 1499 |
-
"loss": 0.
|
| 1500 |
"step": 213
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 0.9184549356223176,
|
| 1504 |
-
"grad_norm": 0.
|
| 1505 |
"learning_rate": 2.0253513192751374e-07,
|
| 1506 |
-
"loss": 0.
|
| 1507 |
"step": 214
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
"epoch": 0.9227467811158798,
|
| 1511 |
-
"grad_norm": 0.
|
| 1512 |
"learning_rate": 1.8190352989793325e-07,
|
| 1513 |
"loss": 0.2807,
|
| 1514 |
"step": 215
|
| 1515 |
},
|
| 1516 |
{
|
| 1517 |
"epoch": 0.927038626609442,
|
| 1518 |
-
"grad_norm": 0.
|
| 1519 |
"learning_rate": 1.6236054330219853e-07,
|
| 1520 |
-
"loss": 0.
|
| 1521 |
"step": 216
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 0.9313304721030042,
|
| 1525 |
-
"grad_norm": 0.
|
| 1526 |
"learning_rate": 1.439105877423963e-07,
|
| 1527 |
-
"loss": 0.
|
| 1528 |
"step": 217
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
"epoch": 0.9356223175965666,
|
| 1532 |
-
"grad_norm": 0.
|
| 1533 |
"learning_rate": 1.2655783185784253e-07,
|
| 1534 |
-
"loss": 0.
|
| 1535 |
"step": 218
|
| 1536 |
},
|
| 1537 |
{
|
| 1538 |
"epoch": 0.9399141630901288,
|
| 1539 |
-
"grad_norm": 0.
|
| 1540 |
"learning_rate": 1.1030619638320805e-07,
|
| 1541 |
-
"loss": 0.
|
| 1542 |
"step": 219
|
| 1543 |
},
|
| 1544 |
{
|
| 1545 |
"epoch": 0.944206008583691,
|
| 1546 |
-
"grad_norm": 0.
|
| 1547 |
"learning_rate": 9.51593532626538e-08,
|
| 1548 |
-
"loss": 0.
|
| 1549 |
"step": 220
|
| 1550 |
},
|
| 1551 |
{
|
| 1552 |
"epoch": 0.9484978540772532,
|
| 1553 |
-
"grad_norm": 0.
|
| 1554 |
"learning_rate": 8.11207248201834e-08,
|
| 1555 |
-
"loss": 0.
|
| 1556 |
"step": 221
|
| 1557 |
},
|
| 1558 |
{
|
| 1559 |
"epoch": 0.9527896995708155,
|
| 1560 |
-
"grad_norm": 0.
|
| 1561 |
"learning_rate": 6.819348298638839e-08,
|
| 1562 |
-
"loss": 0.
|
| 1563 |
"step": 222
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 0.9570815450643777,
|
| 1567 |
-
"grad_norm": 0.
|
| 1568 |
"learning_rate": 5.638054858177644e-08,
|
| 1569 |
-
"loss": 0.
|
| 1570 |
"step": 223
|
| 1571 |
},
|
| 1572 |
{
|
| 1573 |
"epoch": 0.9613733905579399,
|
| 1574 |
-
"grad_norm": 0.
|
| 1575 |
"learning_rate": 4.568459065683206e-08,
|
| 1576 |
-
"loss": 0.
|
| 1577 |
"step": 224
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"epoch": 0.9656652360515021,
|
| 1581 |
-
"grad_norm": 0.
|
| 1582 |
"learning_rate": 3.610802588895845e-08,
|
| 1583 |
-
"loss": 0.
|
| 1584 |
"step": 225
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 0.9699570815450643,
|
| 1588 |
-
"grad_norm": 0.
|
| 1589 |
"learning_rate": 2.765301803645426e-08,
|
| 1590 |
-
"loss": 0.
|
| 1591 |
"step": 226
|
| 1592 |
},
|
| 1593 |
{
|
| 1594 |
"epoch": 0.9742489270386266,
|
| 1595 |
-
"grad_norm": 0.
|
| 1596 |
"learning_rate": 2.0321477449619098e-08,
|
| 1597 |
-
"loss": 0.
|
| 1598 |
"step": 227
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
"epoch": 0.9785407725321889,
|
| 1602 |
-
"grad_norm": 0.
|
| 1603 |
"learning_rate": 1.411506063912882e-08,
|
| 1604 |
-
"loss": 0.
|
| 1605 |
"step": 228
|
| 1606 |
},
|
| 1607 |
{
|
| 1608 |
"epoch": 0.9828326180257511,
|
| 1609 |
-
"grad_norm": 0.
|
| 1610 |
"learning_rate": 9.035169901754902e-09,
|
| 1611 |
-
"loss": 0.
|
| 1612 |
"step": 229
|
| 1613 |
},
|
| 1614 |
{
|
| 1615 |
"epoch": 0.9871244635193133,
|
| 1616 |
-
"grad_norm": 0.
|
| 1617 |
"learning_rate": 5.082953003528457e-09,
|
| 1618 |
"loss": 0.2759,
|
| 1619 |
"step": 230
|
| 1620 |
},
|
| 1621 |
{
|
| 1622 |
"epoch": 0.9914163090128756,
|
| 1623 |
-
"grad_norm": 0.
|
| 1624 |
"learning_rate": 2.2593029204076578e-09,
|
| 1625 |
-
"loss": 0.
|
| 1626 |
"step": 231
|
| 1627 |
},
|
| 1628 |
{
|
| 1629 |
"epoch": 0.9957081545064378,
|
| 1630 |
-
"grad_norm": 0.
|
| 1631 |
"learning_rate": 5.648576365169245e-10,
|
| 1632 |
-
"loss": 0.
|
| 1633 |
"step": 232
|
| 1634 |
},
|
| 1635 |
{
|
| 1636 |
"epoch": 1.0,
|
| 1637 |
-
"grad_norm": 0.
|
| 1638 |
"learning_rate": 0.0,
|
| 1639 |
-
"loss": 0.
|
| 1640 |
"step": 233
|
| 1641 |
},
|
| 1642 |
{
|
| 1643 |
"epoch": 1.0,
|
| 1644 |
-
"eval_loss": 0.
|
| 1645 |
-
"eval_runtime": 1.
|
| 1646 |
-
"eval_samples_per_second":
|
| 1647 |
-
"eval_steps_per_second": 6.
|
| 1648 |
"step": 233
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.0,
|
| 1652 |
"step": 233,
|
| 1653 |
"total_flos": 1.933383782157517e+16,
|
| 1654 |
-
"train_loss": 0.
|
| 1655 |
-
"train_runtime": 303.
|
| 1656 |
-
"train_samples_per_second": 36.
|
| 1657 |
-
"train_steps_per_second": 0.
|
| 1658 |
}
|
| 1659 |
],
|
| 1660 |
"logging_steps": 1,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.004291845493562232,
|
| 13 |
+
"grad_norm": 6.858954429626465,
|
| 14 |
"learning_rate": 4.1666666666666667e-07,
|
| 15 |
"loss": 2.3401,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.008583690987124463,
|
| 20 |
+
"grad_norm": 7.1901326179504395,
|
| 21 |
"learning_rate": 8.333333333333333e-07,
|
| 22 |
"loss": 2.3774,
|
| 23 |
"step": 2
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.012875536480686695,
|
| 27 |
+
"grad_norm": 6.998518466949463,
|
| 28 |
"learning_rate": 1.25e-06,
|
| 29 |
+
"loss": 2.3589,
|
| 30 |
"step": 3
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.017167381974248927,
|
| 34 |
+
"grad_norm": 6.999710559844971,
|
| 35 |
"learning_rate": 1.6666666666666667e-06,
|
| 36 |
+
"loss": 2.3641,
|
| 37 |
"step": 4
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.02145922746781116,
|
| 41 |
+
"grad_norm": 6.936703681945801,
|
| 42 |
"learning_rate": 2.0833333333333334e-06,
|
| 43 |
+
"loss": 2.3491,
|
| 44 |
"step": 5
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.02575107296137339,
|
| 48 |
+
"grad_norm": 6.817142963409424,
|
| 49 |
"learning_rate": 2.5e-06,
|
| 50 |
+
"loss": 2.28,
|
| 51 |
"step": 6
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.030042918454935622,
|
| 55 |
+
"grad_norm": 6.60942268371582,
|
| 56 |
"learning_rate": 2.916666666666667e-06,
|
| 57 |
+
"loss": 2.2736,
|
| 58 |
"step": 7
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.034334763948497854,
|
| 62 |
+
"grad_norm": 5.893615245819092,
|
| 63 |
"learning_rate": 3.3333333333333333e-06,
|
| 64 |
+
"loss": 2.1437,
|
| 65 |
"step": 8
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03862660944206009,
|
| 69 |
+
"grad_norm": 5.829619407653809,
|
| 70 |
"learning_rate": 3.7500000000000005e-06,
|
| 71 |
+
"loss": 2.1175,
|
| 72 |
"step": 9
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04291845493562232,
|
| 76 |
+
"grad_norm": 5.597039699554443,
|
| 77 |
"learning_rate": 4.166666666666667e-06,
|
| 78 |
+
"loss": 2.0907,
|
| 79 |
"step": 10
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.04721030042918455,
|
| 83 |
+
"grad_norm": 4.393044948577881,
|
| 84 |
"learning_rate": 4.583333333333333e-06,
|
| 85 |
+
"loss": 1.7745,
|
| 86 |
"step": 11
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.05150214592274678,
|
| 90 |
+
"grad_norm": 4.335526466369629,
|
| 91 |
"learning_rate": 5e-06,
|
| 92 |
+
"loss": 1.6767,
|
| 93 |
"step": 12
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.055793991416309016,
|
| 97 |
+
"grad_norm": 4.237323760986328,
|
| 98 |
"learning_rate": 5.416666666666667e-06,
|
| 99 |
+
"loss": 1.6224,
|
| 100 |
"step": 13
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.060085836909871244,
|
| 104 |
+
"grad_norm": 3.4697413444519043,
|
| 105 |
"learning_rate": 5.833333333333334e-06,
|
| 106 |
+
"loss": 1.1534,
|
| 107 |
"step": 14
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06437768240343347,
|
| 111 |
+
"grad_norm": 3.5880987644195557,
|
| 112 |
"learning_rate": 6.25e-06,
|
| 113 |
+
"loss": 1.0938,
|
| 114 |
"step": 15
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.06866952789699571,
|
| 118 |
+
"grad_norm": 3.3989222049713135,
|
| 119 |
"learning_rate": 6.666666666666667e-06,
|
| 120 |
+
"loss": 1.0218,
|
| 121 |
"step": 16
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07296137339055794,
|
| 125 |
+
"grad_norm": 2.876070976257324,
|
| 126 |
"learning_rate": 7.083333333333335e-06,
|
| 127 |
+
"loss": 0.9028,
|
| 128 |
"step": 17
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07725321888412018,
|
| 132 |
+
"grad_norm": 2.5071043968200684,
|
| 133 |
"learning_rate": 7.500000000000001e-06,
|
| 134 |
+
"loss": 0.8486,
|
| 135 |
"step": 18
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.0815450643776824,
|
| 139 |
+
"grad_norm": 2.0654499530792236,
|
| 140 |
"learning_rate": 7.916666666666667e-06,
|
| 141 |
+
"loss": 0.5864,
|
| 142 |
"step": 19
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08583690987124463,
|
| 146 |
+
"grad_norm": 1.3941829204559326,
|
| 147 |
"learning_rate": 8.333333333333334e-06,
|
| 148 |
+
"loss": 0.4791,
|
| 149 |
"step": 20
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09012875536480687,
|
| 153 |
+
"grad_norm": 1.0642451047897339,
|
| 154 |
"learning_rate": 8.750000000000001e-06,
|
| 155 |
+
"loss": 0.4628,
|
| 156 |
"step": 21
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.0944206008583691,
|
| 160 |
+
"grad_norm": 0.8186908960342407,
|
| 161 |
"learning_rate": 9.166666666666666e-06,
|
| 162 |
+
"loss": 0.423,
|
| 163 |
"step": 22
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.09871244635193133,
|
| 167 |
+
"grad_norm": 0.714146614074707,
|
| 168 |
"learning_rate": 9.583333333333335e-06,
|
| 169 |
+
"loss": 0.4155,
|
| 170 |
"step": 23
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.10300429184549356,
|
| 174 |
+
"grad_norm": 0.6719127297401428,
|
| 175 |
"learning_rate": 1e-05,
|
| 176 |
+
"loss": 0.4172,
|
| 177 |
"step": 24
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.1072961373390558,
|
| 181 |
+
"grad_norm": 0.6532909870147705,
|
| 182 |
"learning_rate": 9.999435142363484e-06,
|
| 183 |
+
"loss": 0.3825,
|
| 184 |
"step": 25
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.11158798283261803,
|
| 188 |
+
"grad_norm": 0.7833576202392578,
|
| 189 |
"learning_rate": 9.997740697079595e-06,
|
| 190 |
+
"loss": 0.4211,
|
| 191 |
"step": 26
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.11587982832618025,
|
| 195 |
+
"grad_norm": 0.6481274366378784,
|
| 196 |
"learning_rate": 9.994917046996472e-06,
|
| 197 |
+
"loss": 0.3679,
|
| 198 |
"step": 27
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.12017167381974249,
|
| 202 |
+
"grad_norm": 0.6363394856452942,
|
| 203 |
"learning_rate": 9.990964830098246e-06,
|
| 204 |
+
"loss": 0.3789,
|
| 205 |
"step": 28
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.12446351931330472,
|
| 209 |
+
"grad_norm": 0.6472516655921936,
|
| 210 |
"learning_rate": 9.985884939360873e-06,
|
| 211 |
+
"loss": 0.4076,
|
| 212 |
"step": 29
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.12875536480686695,
|
| 216 |
+
"grad_norm": 0.5628584623336792,
|
| 217 |
"learning_rate": 9.979678522550382e-06,
|
| 218 |
+
"loss": 0.3832,
|
| 219 |
"step": 30
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.13304721030042918,
|
| 223 |
+
"grad_norm": 0.5358518362045288,
|
| 224 |
"learning_rate": 9.972346981963546e-06,
|
| 225 |
+
"loss": 0.3658,
|
| 226 |
"step": 31
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.13733905579399142,
|
| 230 |
+
"grad_norm": 0.5220690965652466,
|
| 231 |
"learning_rate": 9.963891974111042e-06,
|
| 232 |
+
"loss": 0.3743,
|
| 233 |
"step": 32
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.14163090128755365,
|
| 237 |
+
"grad_norm": 0.544533908367157,
|
| 238 |
"learning_rate": 9.95431540934317e-06,
|
| 239 |
+
"loss": 0.3732,
|
| 240 |
"step": 33
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.1459227467811159,
|
| 244 |
+
"grad_norm": 0.5741094946861267,
|
| 245 |
"learning_rate": 9.943619451418225e-06,
|
| 246 |
+
"loss": 0.3981,
|
| 247 |
"step": 34
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.15021459227467812,
|
| 251 |
+
"grad_norm": 0.5270912647247314,
|
| 252 |
"learning_rate": 9.931806517013612e-06,
|
| 253 |
+
"loss": 0.3583,
|
| 254 |
"step": 35
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.15450643776824036,
|
| 258 |
+
"grad_norm": 0.49920132756233215,
|
| 259 |
"learning_rate": 9.918879275179819e-06,
|
| 260 |
+
"loss": 0.3414,
|
| 261 |
"step": 36
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.15879828326180256,
|
| 265 |
+
"grad_norm": 0.5417159199714661,
|
| 266 |
"learning_rate": 9.904840646737346e-06,
|
| 267 |
+
"loss": 0.3734,
|
| 268 |
"step": 37
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.1630901287553648,
|
| 272 |
+
"grad_norm": 0.4884810149669647,
|
| 273 |
"learning_rate": 9.889693803616793e-06,
|
| 274 |
+
"loss": 0.3682,
|
| 275 |
"step": 38
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.16738197424892703,
|
| 279 |
+
"grad_norm": 0.48894229531288147,
|
| 280 |
"learning_rate": 9.873442168142158e-06,
|
| 281 |
+
"loss": 0.3607,
|
| 282 |
"step": 39
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.17167381974248927,
|
| 286 |
+
"grad_norm": 0.4725055396556854,
|
| 287 |
"learning_rate": 9.856089412257605e-06,
|
| 288 |
+
"loss": 0.3348,
|
| 289 |
"step": 40
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.1759656652360515,
|
| 293 |
+
"grad_norm": 0.499160498380661,
|
| 294 |
"learning_rate": 9.837639456697802e-06,
|
| 295 |
+
"loss": 0.382,
|
| 296 |
"step": 41
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.18025751072961374,
|
| 300 |
+
"grad_norm": 0.4489183723926544,
|
| 301 |
"learning_rate": 9.818096470102067e-06,
|
| 302 |
+
"loss": 0.3341,
|
| 303 |
"step": 42
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.18454935622317598,
|
| 307 |
+
"grad_norm": 0.4789492189884186,
|
| 308 |
"learning_rate": 9.797464868072489e-06,
|
| 309 |
+
"loss": 0.3521,
|
| 310 |
"step": 43
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.1888412017167382,
|
| 314 |
+
"grad_norm": 0.47436925768852234,
|
| 315 |
"learning_rate": 9.775749312176249e-06,
|
| 316 |
+
"loss": 0.3302,
|
| 317 |
"step": 44
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.19313304721030042,
|
| 321 |
+
"grad_norm": 0.4419718086719513,
|
| 322 |
"learning_rate": 9.752954708892379e-06,
|
| 323 |
+
"loss": 0.3392,
|
| 324 |
"step": 45
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.19742489270386265,
|
| 328 |
+
"grad_norm": 0.4800013601779938,
|
| 329 |
"learning_rate": 9.729086208503174e-06,
|
| 330 |
+
"loss": 0.3416,
|
| 331 |
"step": 46
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.2017167381974249,
|
| 335 |
+
"grad_norm": 0.48074233531951904,
|
| 336 |
"learning_rate": 9.704149203930522e-06,
|
| 337 |
+
"loss": 0.3632,
|
| 338 |
"step": 47
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.20600858369098712,
|
| 342 |
+
"grad_norm": 0.48519065976142883,
|
| 343 |
"learning_rate": 9.67814932951741e-06,
|
| 344 |
+
"loss": 0.3706,
|
| 345 |
"step": 48
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.21030042918454936,
|
| 349 |
+
"grad_norm": 0.46741950511932373,
|
| 350 |
"learning_rate": 9.651092459754879e-06,
|
| 351 |
+
"loss": 0.3728,
|
| 352 |
"step": 49
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.2145922746781116,
|
| 356 |
+
"grad_norm": 0.46153953671455383,
|
| 357 |
"learning_rate": 9.622984707954732e-06,
|
| 358 |
+
"loss": 0.348,
|
| 359 |
"step": 50
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.21888412017167383,
|
| 363 |
+
"grad_norm": 0.42454302310943604,
|
| 364 |
"learning_rate": 9.593832424868271e-06,
|
| 365 |
+
"loss": 0.3319,
|
| 366 |
"step": 51
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.22317596566523606,
|
| 370 |
+
"grad_norm": 0.47633853554725647,
|
| 371 |
"learning_rate": 9.563642197251382e-06,
|
| 372 |
+
"loss": 0.3699,
|
| 373 |
"step": 52
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.22746781115879827,
|
| 377 |
+
"grad_norm": 0.46166110038757324,
|
| 378 |
"learning_rate": 9.532420846376316e-06,
|
| 379 |
+
"loss": 0.3278,
|
| 380 |
"step": 53
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.2317596566523605,
|
| 384 |
+
"grad_norm": 0.4451698362827301,
|
| 385 |
"learning_rate": 9.500175426490455e-06,
|
| 386 |
+
"loss": 0.3165,
|
| 387 |
"step": 54
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.23605150214592274,
|
| 391 |
+
"grad_norm": 0.4674447774887085,
|
| 392 |
"learning_rate": 9.466913223222467e-06,
|
| 393 |
+
"loss": 0.3487,
|
| 394 |
"step": 55
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.24034334763948498,
|
| 398 |
+
"grad_norm": 0.44866660237312317,
|
| 399 |
"learning_rate": 9.432641751936162e-06,
|
| 400 |
+
"loss": 0.3286,
|
| 401 |
"step": 56
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.2446351931330472,
|
| 405 |
+
"grad_norm": 0.4655809700489044,
|
| 406 |
"learning_rate": 9.397368756032445e-06,
|
| 407 |
"loss": 0.3435,
|
| 408 |
"step": 57
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.24892703862660945,
|
| 412 |
+
"grad_norm": 0.4543526768684387,
|
| 413 |
"learning_rate": 9.361102205199762e-06,
|
| 414 |
+
"loss": 0.3448,
|
| 415 |
"step": 58
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.2532188841201717,
|
| 419 |
+
"grad_norm": 0.46444252133369446,
|
| 420 |
"learning_rate": 9.32385029361338e-06,
|
| 421 |
+
"loss": 0.3244,
|
| 422 |
"step": 59
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.2575107296137339,
|
| 426 |
+
"grad_norm": 0.44596582651138306,
|
| 427 |
"learning_rate": 9.285621438083997e-06,
|
| 428 |
+
"loss": 0.3183,
|
| 429 |
"step": 60
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.26180257510729615,
|
| 433 |
+
"grad_norm": 0.4672187864780426,
|
| 434 |
"learning_rate": 9.246424276156008e-06,
|
| 435 |
+
"loss": 0.3288,
|
| 436 |
"step": 61
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.26609442060085836,
|
| 440 |
+
"grad_norm": 0.451576292514801,
|
| 441 |
"learning_rate": 9.206267664155906e-06,
|
| 442 |
+
"loss": 0.3268,
|
| 443 |
"step": 62
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.2703862660944206,
|
| 447 |
+
"grad_norm": 0.4450378119945526,
|
| 448 |
"learning_rate": 9.165160675191272e-06,
|
| 449 |
+
"loss": 0.3318,
|
| 450 |
"step": 63
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.27467811158798283,
|
| 454 |
+
"grad_norm": 0.4558464586734772,
|
| 455 |
"learning_rate": 9.123112597100759e-06,
|
| 456 |
+
"loss": 0.3366,
|
| 457 |
"step": 64
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.27896995708154504,
|
| 461 |
+
"grad_norm": 0.5316265821456909,
|
| 462 |
"learning_rate": 9.080132930355567e-06,
|
| 463 |
+
"loss": 0.3323,
|
| 464 |
"step": 65
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.2832618025751073,
|
| 468 |
+
"grad_norm": 0.42524275183677673,
|
| 469 |
"learning_rate": 9.03623138591289e-06,
|
| 470 |
+
"loss": 0.3076,
|
| 471 |
"step": 66
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.2875536480686695,
|
| 475 |
+
"grad_norm": 0.4055148661136627,
|
| 476 |
"learning_rate": 8.99141788302178e-06,
|
| 477 |
+
"loss": 0.3071,
|
| 478 |
"step": 67
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.2918454935622318,
|
| 482 |
+
"grad_norm": 0.4067463278770447,
|
| 483 |
"learning_rate": 8.94570254698197e-06,
|
| 484 |
+
"loss": 0.2992,
|
| 485 |
"step": 68
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.296137339055794,
|
| 489 |
+
"grad_norm": 0.4295370280742645,
|
| 490 |
"learning_rate": 8.899095706856122e-06,
|
| 491 |
+
"loss": 0.321,
|
| 492 |
"step": 69
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.30042918454935624,
|
| 496 |
+
"grad_norm": 0.44975805282592773,
|
| 497 |
"learning_rate": 8.851607893136065e-06,
|
| 498 |
+
"loss": 0.3521,
|
| 499 |
"step": 70
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.30472103004291845,
|
| 503 |
+
"grad_norm": 0.47437217831611633,
|
| 504 |
"learning_rate": 8.803249835363486e-06,
|
| 505 |
+
"loss": 0.2967,
|
| 506 |
"step": 71
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.3090128755364807,
|
| 510 |
+
"grad_norm": 0.4566970765590668,
|
| 511 |
"learning_rate": 8.754032459705672e-06,
|
| 512 |
+
"loss": 0.3449,
|
| 513 |
"step": 72
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.3133047210300429,
|
| 517 |
+
"grad_norm": 0.4559485614299774,
|
| 518 |
"learning_rate": 8.703966886486819e-06,
|
| 519 |
+
"loss": 0.3286,
|
| 520 |
"step": 73
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.31759656652360513,
|
| 524 |
+
"grad_norm": 0.42962053418159485,
|
| 525 |
"learning_rate": 8.65306442767547e-06,
|
| 526 |
+
"loss": 0.3195,
|
| 527 |
"step": 74
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.3218884120171674,
|
| 531 |
+
"grad_norm": 0.4547812342643738,
|
| 532 |
"learning_rate": 8.601336584328659e-06,
|
| 533 |
+
"loss": 0.3347,
|
| 534 |
"step": 75
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.3261802575107296,
|
| 538 |
+
"grad_norm": 0.5012584328651428,
|
| 539 |
"learning_rate": 8.548795043993316e-06,
|
| 540 |
+
"loss": 0.2947,
|
| 541 |
"step": 76
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.33047210300429186,
|
| 545 |
+
"grad_norm": 0.4377043545246124,
|
| 546 |
"learning_rate": 8.495451678065563e-06,
|
| 547 |
+
"loss": 0.3323,
|
| 548 |
"step": 77
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.33476394849785407,
|
| 552 |
+
"grad_norm": 0.45930778980255127,
|
| 553 |
"learning_rate": 8.441318539108433e-06,
|
| 554 |
+
"loss": 0.3447,
|
| 555 |
"step": 78
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.33905579399141633,
|
| 559 |
+
"grad_norm": 0.43653759360313416,
|
| 560 |
"learning_rate": 8.386407858128707e-06,
|
| 561 |
+
"loss": 0.3362,
|
| 562 |
"step": 79
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.34334763948497854,
|
| 566 |
+
"grad_norm": 0.4360993206501007,
|
| 567 |
"learning_rate": 8.330732041813367e-06,
|
| 568 |
+
"loss": 0.313,
|
| 569 |
"step": 80
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.34763948497854075,
|
| 573 |
+
"grad_norm": 0.45763546228408813,
|
| 574 |
"learning_rate": 8.274303669726427e-06,
|
| 575 |
+
"loss": 0.3427,
|
| 576 |
"step": 81
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.351931330472103,
|
| 580 |
+
"grad_norm": 0.45791178941726685,
|
| 581 |
"learning_rate": 8.217135491466636e-06,
|
| 582 |
+
"loss": 0.3459,
|
| 583 |
"step": 82
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.3562231759656652,
|
| 587 |
+
"grad_norm": 0.4212036430835724,
|
| 588 |
"learning_rate": 8.15924042378682e-06,
|
| 589 |
+
"loss": 0.3039,
|
| 590 |
"step": 83
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.3605150214592275,
|
| 594 |
+
"grad_norm": 0.4088297486305237,
|
| 595 |
"learning_rate": 8.100631547675417e-06,
|
| 596 |
+
"loss": 0.3036,
|
| 597 |
"step": 84
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.3648068669527897,
|
| 601 |
+
"grad_norm": 0.44220373034477234,
|
| 602 |
"learning_rate": 8.041322105400923e-06,
|
| 603 |
+
"loss": 0.3334,
|
| 604 |
"step": 85
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.36909871244635195,
|
| 608 |
+
"grad_norm": 0.4390133023262024,
|
| 609 |
"learning_rate": 7.981325497519892e-06,
|
| 610 |
+
"loss": 0.3183,
|
| 611 |
"step": 86
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.37339055793991416,
|
| 615 |
+
"grad_norm": 0.46158358454704285,
|
| 616 |
"learning_rate": 7.920655279849173e-06,
|
| 617 |
+
"loss": 0.3211,
|
| 618 |
"step": 87
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.3776824034334764,
|
| 622 |
+
"grad_norm": 0.444137305021286,
|
| 623 |
"learning_rate": 7.859325160403073e-06,
|
| 624 |
+
"loss": 0.3392,
|
| 625 |
"step": 88
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.38197424892703863,
|
| 629 |
+
"grad_norm": 0.4566199779510498,
|
| 630 |
"learning_rate": 7.797348996296116e-06,
|
| 631 |
+
"loss": 0.3201,
|
| 632 |
"step": 89
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.38626609442060084,
|
| 636 |
+
"grad_norm": 0.45487260818481445,
|
| 637 |
"learning_rate": 7.734740790612137e-06,
|
| 638 |
+
"loss": 0.3399,
|
| 639 |
"step": 90
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.3905579399141631,
|
| 643 |
+
"grad_norm": 0.4396146833896637,
|
| 644 |
"learning_rate": 7.671514689240366e-06,
|
| 645 |
+
"loss": 0.3076,
|
| 646 |
"step": 91
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.3948497854077253,
|
| 650 |
+
"grad_norm": 0.45286861062049866,
|
| 651 |
"learning_rate": 7.607684977679284e-06,
|
| 652 |
+
"loss": 0.299,
|
| 653 |
"step": 92
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.39914163090128757,
|
| 657 |
+
"grad_norm": 0.4087929129600525,
|
| 658 |
"learning_rate": 7.543266077808893e-06,
|
| 659 |
+
"loss": 0.2845,
|
| 660 |
"step": 93
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.4034334763948498,
|
| 664 |
+
"grad_norm": 0.4473104774951935,
|
| 665 |
"learning_rate": 7.478272544632204e-06,
|
| 666 |
+
"loss": 0.3198,
|
| 667 |
"step": 94
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.40772532188841204,
|
| 671 |
+
"grad_norm": 0.4244577884674072,
|
| 672 |
"learning_rate": 7.412719062986632e-06,
|
| 673 |
+
"loss": 0.3367,
|
| 674 |
"step": 95
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.41201716738197425,
|
| 678 |
+
"grad_norm": 0.4132303297519684,
|
| 679 |
"learning_rate": 7.3466204442260605e-06,
|
| 680 |
"loss": 0.276,
|
| 681 |
"step": 96
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.41630901287553645,
|
| 685 |
+
"grad_norm": 0.4498296082019806,
|
| 686 |
"learning_rate": 7.279991622874319e-06,
|
| 687 |
+
"loss": 0.3412,
|
| 688 |
"step": 97
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.4206008583690987,
|
| 692 |
+
"grad_norm": 0.45308294892311096,
|
| 693 |
"learning_rate": 7.212847653250828e-06,
|
| 694 |
+
"loss": 0.3001,
|
| 695 |
"step": 98
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.4248927038626609,
|
| 699 |
+
"grad_norm": 0.4460461437702179,
|
| 700 |
"learning_rate": 7.145203706069183e-06,
|
| 701 |
+
"loss": 0.2855,
|
| 702 |
"step": 99
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.4291845493562232,
|
| 706 |
+
"grad_norm": 0.43338119983673096,
|
| 707 |
"learning_rate": 7.0770750650094335e-06,
|
| 708 |
+
"loss": 0.3366,
|
| 709 |
"step": 100
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.4334763948497854,
|
| 713 |
+
"grad_norm": 0.4553949534893036,
|
| 714 |
"learning_rate": 7.008477123264849e-06,
|
| 715 |
+
"loss": 0.3444,
|
| 716 |
"step": 101
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.43776824034334766,
|
| 720 |
+
"grad_norm": 0.42553162574768066,
|
| 721 |
"learning_rate": 6.939425380063924e-06,
|
| 722 |
+
"loss": 0.3146,
|
| 723 |
"step": 102
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.44206008583690987,
|
| 727 |
+
"grad_norm": 0.42176350951194763,
|
| 728 |
"learning_rate": 6.869935437168449e-06,
|
| 729 |
+
"loss": 0.3221,
|
| 730 |
"step": 103
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.44635193133047213,
|
| 734 |
+
"grad_norm": 0.44573134183883667,
|
| 735 |
"learning_rate": 6.800022995348381e-06,
|
| 736 |
"loss": 0.3182,
|
| 737 |
"step": 104
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.45064377682403434,
|
| 741 |
+
"grad_norm": 0.43147024512290955,
|
| 742 |
"learning_rate": 6.729703850834381e-06,
|
| 743 |
+
"loss": 0.3246,
|
| 744 |
"step": 105
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.45493562231759654,
|
| 748 |
+
"grad_norm": 0.402482807636261,
|
| 749 |
"learning_rate": 6.65899389174876e-06,
|
| 750 |
+
"loss": 0.2823,
|
| 751 |
"step": 106
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.4592274678111588,
|
| 755 |
+
"grad_norm": 0.4000365138053894,
|
| 756 |
"learning_rate": 6.587909094515663e-06,
|
| 757 |
+
"loss": 0.2955,
|
| 758 |
"step": 107
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.463519313304721,
|
| 762 |
+
"grad_norm": 0.3905521333217621,
|
| 763 |
"learning_rate": 6.5164655202513135e-06,
|
| 764 |
+
"loss": 0.2793,
|
| 765 |
"step": 108
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.4678111587982833,
|
| 769 |
+
"grad_norm": 0.4438071548938751,
|
| 770 |
"learning_rate": 6.444679311135112e-06,
|
| 771 |
+
"loss": 0.34,
|
| 772 |
"step": 109
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.4721030042918455,
|
| 776 |
+
"grad_norm": 0.45224955677986145,
|
| 777 |
"learning_rate": 6.372566686762427e-06,
|
| 778 |
+
"loss": 0.3406,
|
| 779 |
"step": 110
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.47639484978540775,
|
| 783 |
+
"grad_norm": 0.45422661304473877,
|
| 784 |
"learning_rate": 6.300143940479881e-06,
|
| 785 |
+
"loss": 0.3243,
|
| 786 |
"step": 111
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.48068669527896996,
|
| 790 |
+
"grad_norm": 0.526847779750824,
|
| 791 |
"learning_rate": 6.227427435703997e-06,
|
| 792 |
"loss": 0.2892,
|
| 793 |
"step": 112
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.48497854077253216,
|
| 797 |
+
"grad_norm": 0.4258354604244232,
|
| 798 |
"learning_rate": 6.154433602223979e-06,
|
| 799 |
+
"loss": 0.269,
|
| 800 |
"step": 113
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.4892703862660944,
|
| 804 |
+
"grad_norm": 0.40687990188598633,
|
| 805 |
"learning_rate": 6.0811789324895365e-06,
|
| 806 |
+
"loss": 0.3001,
|
| 807 |
"step": 114
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.49356223175965663,
|
| 811 |
+
"grad_norm": 0.43260207772254944,
|
| 812 |
"learning_rate": 6.0076799778845105e-06,
|
| 813 |
+
"loss": 0.306,
|
| 814 |
"step": 115
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.4978540772532189,
|
| 818 |
+
"grad_norm": 0.43823885917663574,
|
| 819 |
"learning_rate": 5.933953344987215e-06,
|
| 820 |
+
"loss": 0.3119,
|
| 821 |
"step": 116
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.5021459227467812,
|
| 825 |
+
"grad_norm": 0.4540681540966034,
|
| 826 |
"learning_rate": 5.860015691818292e-06,
|
| 827 |
+
"loss": 0.3307,
|
| 828 |
"step": 117
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.5064377682403434,
|
| 832 |
+
"grad_norm": 0.4211110770702362,
|
| 833 |
"learning_rate": 5.78588372407695e-06,
|
| 834 |
+
"loss": 0.2904,
|
| 835 |
"step": 118
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.5107296137339056,
|
| 839 |
+
"grad_norm": 0.42265021800994873,
|
| 840 |
"learning_rate": 5.711574191366427e-06,
|
| 841 |
+
"loss": 0.3069,
|
| 842 |
"step": 119
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.5150214592274678,
|
| 846 |
+
"grad_norm": 0.42844855785369873,
|
| 847 |
"learning_rate": 5.637103883409525e-06,
|
| 848 |
+
"loss": 0.3033,
|
| 849 |
"step": 120
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.51931330472103,
|
| 853 |
+
"grad_norm": 0.4592513144016266,
|
| 854 |
"learning_rate": 5.562489626255104e-06,
|
| 855 |
+
"loss": 0.3125,
|
| 856 |
"step": 121
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.5236051502145923,
|
| 860 |
+
"grad_norm": 0.4309280216693878,
|
| 861 |
"learning_rate": 5.487748278476342e-06,
|
| 862 |
+
"loss": 0.3124,
|
| 863 |
"step": 122
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.5278969957081545,
|
| 867 |
+
"grad_norm": 0.3942621946334839,
|
| 868 |
"learning_rate": 5.412896727361663e-06,
|
| 869 |
+
"loss": 0.2525,
|
| 870 |
"step": 123
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.5321888412017167,
|
| 874 |
+
"grad_norm": 0.40998613834381104,
|
| 875 |
"learning_rate": 5.337951885099167e-06,
|
| 876 |
+
"loss": 0.2895,
|
| 877 |
"step": 124
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.5364806866952789,
|
| 881 |
+
"grad_norm": 0.4335436522960663,
|
| 882 |
"learning_rate": 5.262930684955439e-06,
|
| 883 |
+
"loss": 0.2924,
|
| 884 |
"step": 125
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.5407725321888412,
|
| 888 |
+
"grad_norm": 0.4402632415294647,
|
| 889 |
"learning_rate": 5.187850077449604e-06,
|
| 890 |
"loss": 0.317,
|
| 891 |
"step": 126
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.5450643776824035,
|
| 895 |
+
"grad_norm": 0.39414867758750916,
|
| 896 |
"learning_rate": 5.112727026523461e-06,
|
| 897 |
+
"loss": 0.2753,
|
| 898 |
"step": 127
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.5493562231759657,
|
| 902 |
+
"grad_norm": 0.4759502410888672,
|
| 903 |
"learning_rate": 5.03757850570861e-06,
|
| 904 |
"loss": 0.3066,
|
| 905 |
"step": 128
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.5536480686695279,
|
| 909 |
+
"grad_norm": 0.41338634490966797,
|
| 910 |
"learning_rate": 4.9624214942913916e-06,
|
| 911 |
+
"loss": 0.3013,
|
| 912 |
"step": 129
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.5579399141630901,
|
| 916 |
+
"grad_norm": 0.4515892267227173,
|
| 917 |
"learning_rate": 4.88727297347654e-06,
|
| 918 |
+
"loss": 0.306,
|
| 919 |
"step": 130
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.5622317596566524,
|
| 923 |
+
"grad_norm": 0.3964577317237854,
|
| 924 |
"learning_rate": 4.8121499225503974e-06,
|
| 925 |
"loss": 0.2874,
|
| 926 |
"step": 131
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.5665236051502146,
|
| 930 |
+
"grad_norm": 0.41198039054870605,
|
| 931 |
"learning_rate": 4.737069315044562e-06,
|
| 932 |
+
"loss": 0.2947,
|
| 933 |
"step": 132
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.5708154506437768,
|
| 937 |
+
"grad_norm": 0.4514983892440796,
|
| 938 |
"learning_rate": 4.662048114900837e-06,
|
| 939 |
+
"loss": 0.3235,
|
| 940 |
"step": 133
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.575107296137339,
|
| 944 |
+
"grad_norm": 0.4610046446323395,
|
| 945 |
"learning_rate": 4.587103272638339e-06,
|
| 946 |
+
"loss": 0.3268,
|
| 947 |
"step": 134
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.5793991416309013,
|
| 951 |
+
"grad_norm": 0.4211605489253998,
|
| 952 |
"learning_rate": 4.512251721523659e-06,
|
| 953 |
"loss": 0.2671,
|
| 954 |
"step": 135
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.5836909871244635,
|
| 958 |
+
"grad_norm": 0.4412516951560974,
|
| 959 |
"learning_rate": 4.437510373744897e-06,
|
| 960 |
+
"loss": 0.3072,
|
| 961 |
"step": 136
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.5879828326180258,
|
| 965 |
+
"grad_norm": 0.4627492427825928,
|
| 966 |
"learning_rate": 4.362896116590475e-06,
|
| 967 |
+
"loss": 0.3031,
|
| 968 |
"step": 137
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.592274678111588,
|
| 972 |
+
"grad_norm": 0.4873397648334503,
|
| 973 |
"learning_rate": 4.2884258086335755e-06,
|
| 974 |
+
"loss": 0.3223,
|
| 975 |
"step": 138
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.5965665236051502,
|
| 979 |
+
"grad_norm": 0.41365429759025574,
|
| 980 |
"learning_rate": 4.214116275923051e-06,
|
| 981 |
+
"loss": 0.2736,
|
| 982 |
"step": 139
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.6008583690987125,
|
| 986 |
+
"grad_norm": 0.45329636335372925,
|
| 987 |
"learning_rate": 4.1399843081817085e-06,
|
| 988 |
+
"loss": 0.3119,
|
| 989 |
"step": 140
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.6051502145922747,
|
| 993 |
+
"grad_norm": 0.41913923621177673,
|
| 994 |
"learning_rate": 4.066046655012786e-06,
|
| 995 |
+
"loss": 0.3037,
|
| 996 |
"step": 141
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.6094420600858369,
|
| 1000 |
+
"grad_norm": 0.43529585003852844,
|
| 1001 |
"learning_rate": 3.992320022115492e-06,
|
| 1002 |
+
"loss": 0.3148,
|
| 1003 |
"step": 142
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.6137339055793991,
|
| 1007 |
+
"grad_norm": 0.4258601665496826,
|
| 1008 |
"learning_rate": 3.918821067510464e-06,
|
| 1009 |
+
"loss": 0.2917,
|
| 1010 |
"step": 143
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.6180257510729614,
|
| 1014 |
+
"grad_norm": 0.46047812700271606,
|
| 1015 |
"learning_rate": 3.845566397776022e-06,
|
| 1016 |
+
"loss": 0.3298,
|
| 1017 |
"step": 144
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.6223175965665236,
|
| 1021 |
+
"grad_norm": 0.42097219824790955,
|
| 1022 |
"learning_rate": 3.7725725642960047e-06,
|
| 1023 |
+
"loss": 0.3056,
|
| 1024 |
"step": 145
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.6266094420600858,
|
| 1028 |
+
"grad_norm": 0.43911123275756836,
|
| 1029 |
"learning_rate": 3.6998560595201188e-06,
|
| 1030 |
+
"loss": 0.2912,
|
| 1031 |
"step": 146
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.630901287553648,
|
| 1035 |
+
"grad_norm": 0.42664578557014465,
|
| 1036 |
"learning_rate": 3.627433313237576e-06,
|
| 1037 |
+
"loss": 0.2955,
|
| 1038 |
"step": 147
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.6351931330472103,
|
| 1042 |
+
"grad_norm": 0.42000484466552734,
|
| 1043 |
"learning_rate": 3.555320688864889e-06,
|
| 1044 |
+
"loss": 0.2903,
|
| 1045 |
"step": 148
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.6394849785407726,
|
| 1049 |
+
"grad_norm": 0.4376209080219269,
|
| 1050 |
"learning_rate": 3.483534479748688e-06,
|
| 1051 |
+
"loss": 0.291,
|
| 1052 |
"step": 149
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6437768240343348,
|
| 1056 |
+
"grad_norm": 0.4324072301387787,
|
| 1057 |
"learning_rate": 3.4120909054843375e-06,
|
| 1058 |
+
"loss": 0.2824,
|
| 1059 |
"step": 150
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.648068669527897,
|
| 1063 |
+
"grad_norm": 0.38668328523635864,
|
| 1064 |
"learning_rate": 3.3410061082512422e-06,
|
| 1065 |
+
"loss": 0.2586,
|
| 1066 |
"step": 151
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6523605150214592,
|
| 1070 |
+
"grad_norm": 0.40362098813056946,
|
| 1071 |
"learning_rate": 3.2702961491656197e-06,
|
| 1072 |
+
"loss": 0.2853,
|
| 1073 |
"step": 152
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6566523605150214,
|
| 1077 |
+
"grad_norm": 0.3973497152328491,
|
| 1078 |
"learning_rate": 3.1999770046516198e-06,
|
| 1079 |
+
"loss": 0.2826,
|
| 1080 |
"step": 153
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6609442060085837,
|
| 1084 |
+
"grad_norm": 0.4138431251049042,
|
| 1085 |
"learning_rate": 3.130064562831553e-06,
|
| 1086 |
+
"loss": 0.2757,
|
| 1087 |
"step": 154
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6652360515021459,
|
| 1091 |
+
"grad_norm": 0.4256836771965027,
|
| 1092 |
"learning_rate": 3.0605746199360755e-06,
|
| 1093 |
+
"loss": 0.2978,
|
| 1094 |
"step": 155
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6695278969957081,
|
| 1098 |
+
"grad_norm": 0.41062039136886597,
|
| 1099 |
"learning_rate": 2.991522876735154e-06,
|
| 1100 |
+
"loss": 0.2859,
|
| 1101 |
"step": 156
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6738197424892703,
|
| 1105 |
+
"grad_norm": 0.44024190306663513,
|
| 1106 |
"learning_rate": 2.9229249349905686e-06,
|
| 1107 |
+
"loss": 0.3014,
|
| 1108 |
"step": 157
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.6781115879828327,
|
| 1112 |
+
"grad_norm": 0.44576936960220337,
|
| 1113 |
"learning_rate": 2.8547962939308187e-06,
|
| 1114 |
+
"loss": 0.2828,
|
| 1115 |
"step": 158
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6824034334763949,
|
| 1119 |
+
"grad_norm": 0.4253944456577301,
|
| 1120 |
"learning_rate": 2.787152346749173e-06,
|
| 1121 |
+
"loss": 0.3,
|
| 1122 |
"step": 159
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6866952789699571,
|
| 1126 |
+
"grad_norm": 0.40661880373954773,
|
| 1127 |
"learning_rate": 2.720008377125682e-06,
|
| 1128 |
+
"loss": 0.2811,
|
| 1129 |
"step": 160
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.6909871244635193,
|
| 1133 |
+
"grad_norm": 0.45691847801208496,
|
| 1134 |
"learning_rate": 2.6533795557739407e-06,
|
| 1135 |
"loss": 0.3238,
|
| 1136 |
"step": 161
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.6952789699570815,
|
| 1140 |
+
"grad_norm": 0.42418181896209717,
|
| 1141 |
"learning_rate": 2.5872809370133704e-06,
|
| 1142 |
+
"loss": 0.2935,
|
| 1143 |
"step": 162
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.6995708154506438,
|
| 1147 |
+
"grad_norm": 0.42245230078697205,
|
| 1148 |
"learning_rate": 2.5217274553677975e-06,
|
| 1149 |
"loss": 0.2985,
|
| 1150 |
"step": 163
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.703862660944206,
|
| 1154 |
+
"grad_norm": 0.41597869992256165,
|
| 1155 |
"learning_rate": 2.4567339221911086e-06,
|
| 1156 |
+
"loss": 0.2923,
|
| 1157 |
"step": 164
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7081545064377682,
|
| 1161 |
+
"grad_norm": 0.44589149951934814,
|
| 1162 |
"learning_rate": 2.3923150223207176e-06,
|
| 1163 |
+
"loss": 0.3106,
|
| 1164 |
"step": 165
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7124463519313304,
|
| 1168 |
+
"grad_norm": 0.40725013613700867,
|
| 1169 |
"learning_rate": 2.328485310759635e-06,
|
| 1170 |
+
"loss": 0.2861,
|
| 1171 |
"step": 166
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7167381974248928,
|
| 1175 |
+
"grad_norm": 0.44540056586265564,
|
| 1176 |
"learning_rate": 2.265259209387867e-06,
|
| 1177 |
+
"loss": 0.3039,
|
| 1178 |
"step": 167
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.721030042918455,
|
| 1182 |
+
"grad_norm": 0.4068719446659088,
|
| 1183 |
"learning_rate": 2.202651003703885e-06,
|
| 1184 |
+
"loss": 0.2802,
|
| 1185 |
"step": 168
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7253218884120172,
|
| 1189 |
+
"grad_norm": 0.39111173152923584,
|
| 1190 |
"learning_rate": 2.140674839596931e-06,
|
| 1191 |
+
"loss": 0.2656,
|
| 1192 |
"step": 169
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7296137339055794,
|
| 1196 |
+
"grad_norm": 0.4605322480201721,
|
| 1197 |
"learning_rate": 2.0793447201508288e-06,
|
| 1198 |
+
"loss": 0.3151,
|
| 1199 |
"step": 170
|
| 1200 |
},
|
| 1201 |
{
|
| 1202 |
"epoch": 0.7339055793991416,
|
| 1203 |
+
"grad_norm": 0.4269479513168335,
|
| 1204 |
"learning_rate": 2.01867450248011e-06,
|
| 1205 |
+
"loss": 0.3022,
|
| 1206 |
"step": 171
|
| 1207 |
},
|
| 1208 |
{
|
| 1209 |
"epoch": 0.7381974248927039,
|
| 1210 |
+
"grad_norm": 0.3947524130344391,
|
| 1211 |
"learning_rate": 1.9586778945990785e-06,
|
| 1212 |
+
"loss": 0.2815,
|
| 1213 |
"step": 172
|
| 1214 |
},
|
| 1215 |
{
|
| 1216 |
"epoch": 0.7424892703862661,
|
| 1217 |
+
"grad_norm": 0.43399283289909363,
|
| 1218 |
"learning_rate": 1.8993684523245842e-06,
|
| 1219 |
+
"loss": 0.2899,
|
| 1220 |
"step": 173
|
| 1221 |
},
|
| 1222 |
{
|
| 1223 |
"epoch": 0.7467811158798283,
|
| 1224 |
+
"grad_norm": 0.4567514955997467,
|
| 1225 |
"learning_rate": 1.8407595762131814e-06,
|
| 1226 |
+
"loss": 0.3411,
|
| 1227 |
"step": 174
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 0.7510729613733905,
|
| 1231 |
+
"grad_norm": 0.4480878710746765,
|
| 1232 |
"learning_rate": 1.7828645085333645e-06,
|
| 1233 |
+
"loss": 0.3458,
|
| 1234 |
"step": 175
|
| 1235 |
},
|
| 1236 |
{
|
| 1237 |
"epoch": 0.7553648068669528,
|
| 1238 |
+
"grad_norm": 0.44001972675323486,
|
| 1239 |
"learning_rate": 1.7256963302735752e-06,
|
| 1240 |
+
"loss": 0.3171,
|
| 1241 |
"step": 176
|
| 1242 |
},
|
| 1243 |
{
|
| 1244 |
"epoch": 0.759656652360515,
|
| 1245 |
+
"grad_norm": 0.42058637738227844,
|
| 1246 |
"learning_rate": 1.6692679581866334e-06,
|
| 1247 |
+
"loss": 0.3039,
|
| 1248 |
"step": 177
|
| 1249 |
},
|
| 1250 |
{
|
| 1251 |
"epoch": 0.7639484978540773,
|
| 1252 |
+
"grad_norm": 0.42307794094085693,
|
| 1253 |
"learning_rate": 1.6135921418712959e-06,
|
| 1254 |
+
"loss": 0.2734,
|
| 1255 |
"step": 178
|
| 1256 |
},
|
| 1257 |
{
|
| 1258 |
"epoch": 0.7682403433476395,
|
| 1259 |
+
"grad_norm": 0.42678767442703247,
|
| 1260 |
"learning_rate": 1.5586814608915673e-06,
|
| 1261 |
+
"loss": 0.3135,
|
| 1262 |
"step": 179
|
| 1263 |
},
|
| 1264 |
{
|
| 1265 |
"epoch": 0.7725321888412017,
|
| 1266 |
+
"grad_norm": 0.4102153480052948,
|
| 1267 |
"learning_rate": 1.5045483219344387e-06,
|
| 1268 |
+
"loss": 0.2792,
|
| 1269 |
"step": 180
|
| 1270 |
},
|
| 1271 |
{
|
| 1272 |
"epoch": 0.776824034334764,
|
| 1273 |
+
"grad_norm": 0.40929415822029114,
|
| 1274 |
"learning_rate": 1.4512049560066837e-06,
|
| 1275 |
"loss": 0.2758,
|
| 1276 |
"step": 181
|
| 1277 |
},
|
| 1278 |
{
|
| 1279 |
"epoch": 0.7811158798283262,
|
| 1280 |
+
"grad_norm": 0.42089492082595825,
|
| 1281 |
"learning_rate": 1.3986634156713418e-06,
|
| 1282 |
+
"loss": 0.2835,
|
| 1283 |
"step": 182
|
| 1284 |
},
|
| 1285 |
{
|
| 1286 |
"epoch": 0.7854077253218884,
|
| 1287 |
+
"grad_norm": 0.420242041349411,
|
| 1288 |
"learning_rate": 1.3469355723245303e-06,
|
| 1289 |
+
"loss": 0.2855,
|
| 1290 |
"step": 183
|
| 1291 |
},
|
| 1292 |
{
|
| 1293 |
"epoch": 0.7896995708154506,
|
| 1294 |
+
"grad_norm": 0.45139551162719727,
|
| 1295 |
"learning_rate": 1.2960331135131826e-06,
|
| 1296 |
+
"loss": 0.3284,
|
| 1297 |
"step": 184
|
| 1298 |
},
|
| 1299 |
{
|
| 1300 |
"epoch": 0.7939914163090128,
|
| 1301 |
+
"grad_norm": 0.4033341109752655,
|
| 1302 |
"learning_rate": 1.245967540294329e-06,
|
| 1303 |
+
"loss": 0.2848,
|
| 1304 |
"step": 185
|
| 1305 |
},
|
| 1306 |
{
|
| 1307 |
"epoch": 0.7982832618025751,
|
| 1308 |
+
"grad_norm": 0.37155604362487793,
|
| 1309 |
"learning_rate": 1.1967501646365147e-06,
|
| 1310 |
+
"loss": 0.2757,
|
| 1311 |
"step": 186
|
| 1312 |
},
|
| 1313 |
{
|
| 1314 |
"epoch": 0.8025751072961373,
|
| 1315 |
+
"grad_norm": 0.4291873872280121,
|
| 1316 |
"learning_rate": 1.1483921068639353e-06,
|
| 1317 |
+
"loss": 0.2755,
|
| 1318 |
"step": 187
|
| 1319 |
},
|
| 1320 |
{
|
| 1321 |
"epoch": 0.8068669527896996,
|
| 1322 |
+
"grad_norm": 0.4211515784263611,
|
| 1323 |
"learning_rate": 1.1009042931438784e-06,
|
| 1324 |
+
"loss": 0.298,
|
| 1325 |
"step": 188
|
| 1326 |
},
|
| 1327 |
{
|
| 1328 |
"epoch": 0.8111587982832618,
|
| 1329 |
+
"grad_norm": 0.4309849441051483,
|
| 1330 |
"learning_rate": 1.0542974530180327e-06,
|
| 1331 |
+
"loss": 0.3067,
|
| 1332 |
"step": 189
|
| 1333 |
},
|
| 1334 |
{
|
| 1335 |
"epoch": 0.8154506437768241,
|
| 1336 |
+
"grad_norm": 0.3826170563697815,
|
| 1337 |
"learning_rate": 1.00858211697822e-06,
|
| 1338 |
+
"loss": 0.265,
|
| 1339 |
"step": 190
|
| 1340 |
},
|
| 1341 |
{
|
| 1342 |
"epoch": 0.8197424892703863,
|
| 1343 |
+
"grad_norm": 0.40116748213768005,
|
| 1344 |
"learning_rate": 9.637686140871121e-07,
|
| 1345 |
+
"loss": 0.2901,
|
| 1346 |
"step": 191
|
| 1347 |
},
|
| 1348 |
{
|
| 1349 |
"epoch": 0.8240343347639485,
|
| 1350 |
+
"grad_norm": 0.40444761514663696,
|
| 1351 |
"learning_rate": 9.198670696444339e-07,
|
| 1352 |
+
"loss": 0.2805,
|
| 1353 |
"step": 192
|
| 1354 |
},
|
| 1355 |
{
|
| 1356 |
"epoch": 0.8283261802575107,
|
| 1357 |
+
"grad_norm": 0.521791934967041,
|
| 1358 |
"learning_rate": 8.768874028992431e-07,
|
| 1359 |
+
"loss": 0.3184,
|
| 1360 |
"step": 193
|
| 1361 |
},
|
| 1362 |
{
|
| 1363 |
"epoch": 0.8326180257510729,
|
| 1364 |
+
"grad_norm": 0.4177843928337097,
|
| 1365 |
"learning_rate": 8.348393248087289e-07,
|
| 1366 |
+
"loss": 0.3183,
|
| 1367 |
"step": 194
|
| 1368 |
},
|
| 1369 |
{
|
| 1370 |
"epoch": 0.8369098712446352,
|
| 1371 |
+
"grad_norm": 0.44546449184417725,
|
| 1372 |
"learning_rate": 7.937323358440935e-07,
|
| 1373 |
+
"loss": 0.3407,
|
| 1374 |
"step": 195
|
| 1375 |
},
|
| 1376 |
{
|
| 1377 |
"epoch": 0.8412017167381974,
|
| 1378 |
+
"grad_norm": 0.43112218379974365,
|
| 1379 |
"learning_rate": 7.535757238439939e-07,
|
| 1380 |
+
"loss": 0.3112,
|
| 1381 |
"step": 196
|
| 1382 |
},
|
| 1383 |
{
|
| 1384 |
"epoch": 0.8454935622317596,
|
| 1385 |
+
"grad_norm": 0.38515380024909973,
|
| 1386 |
"learning_rate": 7.143785619160026e-07,
|
| 1387 |
+
"loss": 0.2781,
|
| 1388 |
"step": 197
|
| 1389 |
},
|
| 1390 |
{
|
| 1391 |
"epoch": 0.8497854077253219,
|
| 1392 |
+
"grad_norm": 0.40132516622543335,
|
| 1393 |
"learning_rate": 6.761497063866207e-07,
|
| 1394 |
+
"loss": 0.2836,
|
| 1395 |
"step": 198
|
| 1396 |
},
|
| 1397 |
{
|
| 1398 |
"epoch": 0.8540772532188842,
|
| 1399 |
+
"grad_norm": 0.43590113520622253,
|
| 1400 |
"learning_rate": 6.388977948002406e-07,
|
| 1401 |
+
"loss": 0.3019,
|
| 1402 |
"step": 199
|
| 1403 |
},
|
| 1404 |
{
|
| 1405 |
"epoch": 0.8583690987124464,
|
| 1406 |
+
"grad_norm": 0.39947134256362915,
|
| 1407 |
"learning_rate": 6.026312439675553e-07,
|
| 1408 |
+
"loss": 0.2741,
|
| 1409 |
"step": 200
|
| 1410 |
},
|
| 1411 |
{
|
| 1412 |
"epoch": 0.8626609442060086,
|
| 1413 |
+
"grad_norm": 0.3957541882991791,
|
| 1414 |
"learning_rate": 5.673582480638395e-07,
|
| 1415 |
+
"loss": 0.3023,
|
| 1416 |
"step": 201
|
| 1417 |
},
|
| 1418 |
{
|
| 1419 |
"epoch": 0.8669527896995708,
|
| 1420 |
+
"grad_norm": 0.44793784618377686,
|
| 1421 |
"learning_rate": 5.330867767775333e-07,
|
| 1422 |
+
"loss": 0.3392,
|
| 1423 |
"step": 202
|
| 1424 |
},
|
| 1425 |
{
|
| 1426 |
"epoch": 0.871244635193133,
|
| 1427 |
+
"grad_norm": 0.40923067927360535,
|
| 1428 |
"learning_rate": 4.998245735095459e-07,
|
| 1429 |
+
"loss": 0.3041,
|
| 1430 |
"step": 203
|
| 1431 |
},
|
| 1432 |
{
|
| 1433 |
"epoch": 0.8755364806866953,
|
| 1434 |
+
"grad_norm": 0.413678377866745,
|
| 1435 |
"learning_rate": 4.6757915362368567e-07,
|
| 1436 |
+
"loss": 0.2865,
|
| 1437 |
"step": 204
|
| 1438 |
},
|
| 1439 |
{
|
| 1440 |
"epoch": 0.8798283261802575,
|
| 1441 |
+
"grad_norm": 0.4134410619735718,
|
| 1442 |
"learning_rate": 4.363578027486187e-07,
|
| 1443 |
+
"loss": 0.302,
|
| 1444 |
"step": 205
|
| 1445 |
},
|
| 1446 |
{
|
| 1447 |
"epoch": 0.8841201716738197,
|
| 1448 |
+
"grad_norm": 0.41365522146224976,
|
| 1449 |
"learning_rate": 4.0616757513173123e-07,
|
| 1450 |
"loss": 0.2828,
|
| 1451 |
"step": 206
|
| 1452 |
},
|
| 1453 |
{
|
| 1454 |
"epoch": 0.8884120171673819,
|
| 1455 |
+
"grad_norm": 0.4147285521030426,
|
| 1456 |
"learning_rate": 3.7701529204526856e-07,
|
| 1457 |
+
"loss": 0.2937,
|
| 1458 |
"step": 207
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 0.8927038626609443,
|
| 1462 |
+
"grad_norm": 0.41686710715293884,
|
| 1463 |
"learning_rate": 3.4890754024512254e-07,
|
| 1464 |
+
"loss": 0.3021,
|
| 1465 |
"step": 208
|
| 1466 |
},
|
| 1467 |
{
|
| 1468 |
"epoch": 0.8969957081545065,
|
| 1469 |
+
"grad_norm": 0.4329932630062103,
|
| 1470 |
"learning_rate": 3.2185067048259245e-07,
|
| 1471 |
"loss": 0.2926,
|
| 1472 |
"step": 209
|
| 1473 |
},
|
| 1474 |
{
|
| 1475 |
"epoch": 0.9012875536480687,
|
| 1476 |
+
"grad_norm": 0.3969036936759949,
|
| 1477 |
"learning_rate": 2.9585079606947843e-07,
|
| 1478 |
+
"loss": 0.2973,
|
| 1479 |
"step": 210
|
| 1480 |
},
|
| 1481 |
{
|
| 1482 |
"epoch": 0.9055793991416309,
|
| 1483 |
+
"grad_norm": 0.42916810512542725,
|
| 1484 |
"learning_rate": 2.7091379149682683e-07,
|
| 1485 |
+
"loss": 0.2745,
|
| 1486 |
"step": 211
|
| 1487 |
},
|
| 1488 |
{
|
| 1489 |
"epoch": 0.9098712446351931,
|
| 1490 |
+
"grad_norm": 0.3912080228328705,
|
| 1491 |
"learning_rate": 2.470452911076227e-07,
|
| 1492 |
+
"loss": 0.2763,
|
| 1493 |
"step": 212
|
| 1494 |
},
|
| 1495 |
{
|
| 1496 |
"epoch": 0.9141630901287554,
|
| 1497 |
+
"grad_norm": 0.4045410454273224,
|
| 1498 |
"learning_rate": 2.242506878237538e-07,
|
| 1499 |
+
"loss": 0.2944,
|
| 1500 |
"step": 213
|
| 1501 |
},
|
| 1502 |
{
|
| 1503 |
"epoch": 0.9184549356223176,
|
| 1504 |
+
"grad_norm": 0.3928059935569763,
|
| 1505 |
"learning_rate": 2.0253513192751374e-07,
|
| 1506 |
+
"loss": 0.2821,
|
| 1507 |
"step": 214
|
| 1508 |
},
|
| 1509 |
{
|
| 1510 |
"epoch": 0.9227467811158798,
|
| 1511 |
+
"grad_norm": 0.42325636744499207,
|
| 1512 |
"learning_rate": 1.8190352989793325e-07,
|
| 1513 |
"loss": 0.2807,
|
| 1514 |
"step": 215
|
| 1515 |
},
|
| 1516 |
{
|
| 1517 |
"epoch": 0.927038626609442,
|
| 1518 |
+
"grad_norm": 0.4236631691455841,
|
| 1519 |
"learning_rate": 1.6236054330219853e-07,
|
| 1520 |
+
"loss": 0.2996,
|
| 1521 |
"step": 216
|
| 1522 |
},
|
| 1523 |
{
|
| 1524 |
"epoch": 0.9313304721030042,
|
| 1525 |
+
"grad_norm": 0.416293740272522,
|
| 1526 |
"learning_rate": 1.439105877423963e-07,
|
| 1527 |
+
"loss": 0.2958,
|
| 1528 |
"step": 217
|
| 1529 |
},
|
| 1530 |
{
|
| 1531 |
"epoch": 0.9356223175965666,
|
| 1532 |
+
"grad_norm": 0.4194984436035156,
|
| 1533 |
"learning_rate": 1.2655783185784253e-07,
|
| 1534 |
+
"loss": 0.2971,
|
| 1535 |
"step": 218
|
| 1536 |
},
|
| 1537 |
{
|
| 1538 |
"epoch": 0.9399141630901288,
|
| 1539 |
+
"grad_norm": 0.4220767617225647,
|
| 1540 |
"learning_rate": 1.1030619638320805e-07,
|
| 1541 |
+
"loss": 0.2947,
|
| 1542 |
"step": 219
|
| 1543 |
},
|
| 1544 |
{
|
| 1545 |
"epoch": 0.944206008583691,
|
| 1546 |
+
"grad_norm": 0.463470995426178,
|
| 1547 |
"learning_rate": 9.51593532626538e-08,
|
| 1548 |
+
"loss": 0.2804,
|
| 1549 |
"step": 220
|
| 1550 |
},
|
| 1551 |
{
|
| 1552 |
"epoch": 0.9484978540772532,
|
| 1553 |
+
"grad_norm": 0.44654470682144165,
|
| 1554 |
"learning_rate": 8.11207248201834e-08,
|
| 1555 |
+
"loss": 0.2907,
|
| 1556 |
"step": 221
|
| 1557 |
},
|
| 1558 |
{
|
| 1559 |
"epoch": 0.9527896995708155,
|
| 1560 |
+
"grad_norm": 0.5606074333190918,
|
| 1561 |
"learning_rate": 6.819348298638839e-08,
|
| 1562 |
+
"loss": 0.3248,
|
| 1563 |
"step": 222
|
| 1564 |
},
|
| 1565 |
{
|
| 1566 |
"epoch": 0.9570815450643777,
|
| 1567 |
+
"grad_norm": 0.41357025504112244,
|
| 1568 |
"learning_rate": 5.638054858177644e-08,
|
| 1569 |
+
"loss": 0.2782,
|
| 1570 |
"step": 223
|
| 1571 |
},
|
| 1572 |
{
|
| 1573 |
"epoch": 0.9613733905579399,
|
| 1574 |
+
"grad_norm": 0.4360516369342804,
|
| 1575 |
"learning_rate": 4.568459065683206e-08,
|
| 1576 |
+
"loss": 0.2866,
|
| 1577 |
"step": 224
|
| 1578 |
},
|
| 1579 |
{
|
| 1580 |
"epoch": 0.9656652360515021,
|
| 1581 |
+
"grad_norm": 0.43068474531173706,
|
| 1582 |
"learning_rate": 3.610802588895845e-08,
|
| 1583 |
+
"loss": 0.3177,
|
| 1584 |
"step": 225
|
| 1585 |
},
|
| 1586 |
{
|
| 1587 |
"epoch": 0.9699570815450643,
|
| 1588 |
+
"grad_norm": 0.39279019832611084,
|
| 1589 |
"learning_rate": 2.765301803645426e-08,
|
| 1590 |
+
"loss": 0.2584,
|
| 1591 |
"step": 226
|
| 1592 |
},
|
| 1593 |
{
|
| 1594 |
"epoch": 0.9742489270386266,
|
| 1595 |
+
"grad_norm": 0.42877382040023804,
|
| 1596 |
"learning_rate": 2.0321477449619098e-08,
|
| 1597 |
+
"loss": 0.3091,
|
| 1598 |
"step": 227
|
| 1599 |
},
|
| 1600 |
{
|
| 1601 |
"epoch": 0.9785407725321889,
|
| 1602 |
+
"grad_norm": 0.4641684889793396,
|
| 1603 |
"learning_rate": 1.411506063912882e-08,
|
| 1604 |
+
"loss": 0.3223,
|
| 1605 |
"step": 228
|
| 1606 |
},
|
| 1607 |
{
|
| 1608 |
"epoch": 0.9828326180257511,
|
| 1609 |
+
"grad_norm": 0.41311776638031006,
|
| 1610 |
"learning_rate": 9.035169901754902e-09,
|
| 1611 |
+
"loss": 0.2827,
|
| 1612 |
"step": 229
|
| 1613 |
},
|
| 1614 |
{
|
| 1615 |
"epoch": 0.9871244635193133,
|
| 1616 |
+
"grad_norm": 0.40819963812828064,
|
| 1617 |
"learning_rate": 5.082953003528457e-09,
|
| 1618 |
"loss": 0.2759,
|
| 1619 |
"step": 230
|
| 1620 |
},
|
| 1621 |
{
|
| 1622 |
"epoch": 0.9914163090128756,
|
| 1623 |
+
"grad_norm": 0.39194202423095703,
|
| 1624 |
"learning_rate": 2.2593029204076578e-09,
|
| 1625 |
+
"loss": 0.2787,
|
| 1626 |
"step": 231
|
| 1627 |
},
|
| 1628 |
{
|
| 1629 |
"epoch": 0.9957081545064378,
|
| 1630 |
+
"grad_norm": 0.43726563453674316,
|
| 1631 |
"learning_rate": 5.648576365169245e-10,
|
| 1632 |
+
"loss": 0.3068,
|
| 1633 |
"step": 232
|
| 1634 |
},
|
| 1635 |
{
|
| 1636 |
"epoch": 1.0,
|
| 1637 |
+
"grad_norm": 0.40751731395721436,
|
| 1638 |
"learning_rate": 0.0,
|
| 1639 |
+
"loss": 0.3021,
|
| 1640 |
"step": 233
|
| 1641 |
},
|
| 1642 |
{
|
| 1643 |
"epoch": 1.0,
|
| 1644 |
+
"eval_loss": 0.3932158648967743,
|
| 1645 |
+
"eval_runtime": 1.5342,
|
| 1646 |
+
"eval_samples_per_second": 149.92,
|
| 1647 |
+
"eval_steps_per_second": 6.518,
|
| 1648 |
"step": 233
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.0,
|
| 1652 |
"step": 233,
|
| 1653 |
"total_flos": 1.933383782157517e+16,
|
| 1654 |
+
"train_loss": 0.43314852683840904,
|
| 1655 |
+
"train_runtime": 303.7389,
|
| 1656 |
+
"train_samples_per_second": 36.676,
|
| 1657 |
+
"train_steps_per_second": 0.767
|
| 1658 |
}
|
| 1659 |
],
|
| 1660 |
"logging_steps": 1,
|