Model save
Browse files- README.md +13 -13
- all_results.json +4 -4
- train_results.json +4 -4
- trainer_state.json +477 -477
README.md
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
license: apache-2.0
|
| 4 |
-
base_model: mistralai/Mistral-7B-v0.
|
| 5 |
tags:
|
| 6 |
- trl
|
| 7 |
- sft
|
|
@@ -18,9 +18,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 18 |
|
| 19 |
# zephyr-7b-sft-full
|
| 20 |
|
| 21 |
-
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
-
- Loss: 0.
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
@@ -57,16 +57,16 @@ The following hyperparameters were used during training:
|
|
| 57 |
|
| 58 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
-
| 1.
|
| 61 |
-
| 1.
|
| 62 |
-
| 1.
|
| 63 |
-
| 1.
|
| 64 |
-
| 1.
|
| 65 |
-
| 0.
|
| 66 |
-
| 0.
|
| 67 |
-
| 0.
|
| 68 |
-
| 0.
|
| 69 |
-
| 0.
|
| 70 |
|
| 71 |
|
| 72 |
### Framework versions
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
license: apache-2.0
|
| 4 |
+
base_model: mistralai/Mistral-7B-v0.3
|
| 5 |
tags:
|
| 6 |
- trl
|
| 7 |
- sft
|
|
|
|
| 18 |
|
| 19 |
# zephyr-7b-sft-full
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.3](https://huggingface.co/mistralai/Mistral-7B-v0.3) on the generator dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.9448
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 57 |
|
| 58 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
|:-------------:|:------:|:----:|:---------------:|
|
| 60 |
+
| 1.025 | 0.0923 | 100 | 1.0240 |
|
| 61 |
+
| 1.033 | 0.1846 | 200 | 1.0464 |
|
| 62 |
+
| 1.037 | 0.2769 | 300 | 1.0424 |
|
| 63 |
+
| 1.0136 | 0.3692 | 400 | 1.0295 |
|
| 64 |
+
| 1.0229 | 0.4615 | 500 | 1.0151 |
|
| 65 |
+
| 0.9745 | 0.5538 | 600 | 0.9945 |
|
| 66 |
+
| 0.9441 | 0.6461 | 700 | 0.9769 |
|
| 67 |
+
| 0.9277 | 0.7383 | 800 | 0.9613 |
|
| 68 |
+
| 0.9384 | 0.8306 | 900 | 0.9501 |
|
| 69 |
+
| 0.9216 | 0.9229 | 1000 | 0.9448 |
|
| 70 |
|
| 71 |
|
| 72 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 207864,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
+
"train_loss": 0.986508995762382,
|
| 5 |
+
"train_runtime": 33955.1767,
|
| 6 |
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 4.084,
|
| 8 |
+
"train_steps_per_second": 0.032
|
| 9 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 207864,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 0.9995385325334564,
|
| 3 |
"total_flos": 453306954547200.0,
|
| 4 |
+
"train_loss": 0.986508995762382,
|
| 5 |
+
"train_runtime": 33955.1767,
|
| 6 |
"train_samples": 207864,
|
| 7 |
+
"train_samples_per_second": 4.084,
|
| 8 |
+
"train_steps_per_second": 0.032
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -10,1611 +10,1611 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0009229349330872173,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 1.8348623853211012e-07,
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.0046146746654360865,
|
| 20 |
-
"grad_norm":
|
| 21 |
"learning_rate": 9.174311926605506e-07,
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009229349330872173,
|
| 27 |
-
"grad_norm":
|
| 28 |
"learning_rate": 1.8348623853211011e-06,
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01384402399630826,
|
| 34 |
-
"grad_norm": 2.
|
| 35 |
"learning_rate": 2.7522935779816517e-06,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.018458698661744346,
|
| 41 |
-
"grad_norm": 2.
|
| 42 |
"learning_rate": 3.6697247706422022e-06,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.023073373327180433,
|
| 48 |
-
"grad_norm":
|
| 49 |
"learning_rate": 4.587155963302753e-06,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02768804799261652,
|
| 55 |
-
"grad_norm": 2.
|
| 56 |
"learning_rate": 5.504587155963303e-06,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.032302722658052604,
|
| 62 |
-
"grad_norm": 2.
|
| 63 |
"learning_rate": 6.422018348623854e-06,
|
| 64 |
-
"loss":
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03691739732348869,
|
| 69 |
-
"grad_norm": 2.
|
| 70 |
"learning_rate": 7.3394495412844045e-06,
|
| 71 |
-
"loss": 1.
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04153207198892478,
|
| 76 |
-
"grad_norm": 2.
|
| 77 |
"learning_rate": 8.256880733944956e-06,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.046146746654360866,
|
| 83 |
-
"grad_norm":
|
| 84 |
"learning_rate": 9.174311926605506e-06,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.050761421319796954,
|
| 90 |
-
"grad_norm": 2.
|
| 91 |
"learning_rate": 1.0091743119266055e-05,
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05537609598523304,
|
| 97 |
-
"grad_norm": 2.
|
| 98 |
"learning_rate": 1.1009174311926607e-05,
|
| 99 |
-
"loss":
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05999077065066913,
|
| 104 |
-
"grad_norm": 2.
|
| 105 |
"learning_rate": 1.1926605504587156e-05,
|
| 106 |
-
"loss": 1.
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06460544531610521,
|
| 111 |
-
"grad_norm": 2.
|
| 112 |
"learning_rate": 1.2844036697247708e-05,
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0692201199815413,
|
| 118 |
-
"grad_norm": 2.
|
| 119 |
"learning_rate": 1.3761467889908258e-05,
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07383479464697738,
|
| 125 |
-
"grad_norm": 2.
|
| 126 |
"learning_rate": 1.4678899082568809e-05,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07844946931241348,
|
| 132 |
-
"grad_norm":
|
| 133 |
"learning_rate": 1.559633027522936e-05,
|
| 134 |
-
"loss": 1.
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.08306414397784956,
|
| 139 |
-
"grad_norm": 2.
|
| 140 |
"learning_rate": 1.6513761467889912e-05,
|
| 141 |
-
"loss": 1.
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08767881864328565,
|
| 146 |
-
"grad_norm": 2.
|
| 147 |
"learning_rate": 1.743119266055046e-05,
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09229349330872173,
|
| 153 |
-
"grad_norm":
|
| 154 |
"learning_rate": 1.834862385321101e-05,
|
| 155 |
-
"loss": 1.
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.09229349330872173,
|
| 160 |
-
"eval_loss": 1.
|
| 161 |
-
"eval_runtime":
|
| 162 |
-
"eval_samples_per_second":
|
| 163 |
-
"eval_steps_per_second": 0.
|
| 164 |
"step": 100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09690816797415783,
|
| 168 |
-
"grad_norm": 2.
|
| 169 |
"learning_rate": 1.9266055045871563e-05,
|
| 170 |
-
"loss": 1.
|
| 171 |
"step": 105
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.10152284263959391,
|
| 175 |
-
"grad_norm":
|
| 176 |
"learning_rate": 1.9999947982262415e-05,
|
| 177 |
-
"loss": 1.
|
| 178 |
"step": 110
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.10613751730503,
|
| 182 |
-
"grad_norm": 2.
|
| 183 |
"learning_rate": 1.9998127418269004e-05,
|
| 184 |
-
"loss": 1.
|
| 185 |
"step": 115
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.11075219197046608,
|
| 189 |
-
"grad_norm":
|
| 190 |
"learning_rate": 1.9993706508539968e-05,
|
| 191 |
-
"loss": 1.
|
| 192 |
"step": 120
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.11536686663590216,
|
| 196 |
-
"grad_norm":
|
| 197 |
"learning_rate": 1.998668640288e-05,
|
| 198 |
-
"loss": 1.
|
| 199 |
"step": 125
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11998154130133826,
|
| 203 |
-
"grad_norm":
|
| 204 |
"learning_rate": 1.997706892710117e-05,
|
| 205 |
-
"loss": 1.
|
| 206 |
"step": 130
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.12459621596677434,
|
| 210 |
-
"grad_norm": 2.
|
| 211 |
"learning_rate": 1.9964856582548094e-05,
|
| 212 |
-
"loss": 1.
|
| 213 |
"step": 135
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.12921089063221042,
|
| 217 |
-
"grad_norm": 1.
|
| 218 |
"learning_rate": 1.9950052545447354e-05,
|
| 219 |
-
"loss": 1.
|
| 220 |
"step": 140
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.13382556529764653,
|
| 224 |
-
"grad_norm": 1.
|
| 225 |
"learning_rate": 1.993266066608142e-05,
|
| 226 |
-
"loss": 1.
|
| 227 |
"step": 145
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.1384402399630826,
|
| 231 |
-
"grad_norm": 1.
|
| 232 |
"learning_rate": 1.991268546778726e-05,
|
| 233 |
-
"loss": 1.
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1430549146285187,
|
| 238 |
-
"grad_norm":
|
| 239 |
"learning_rate": 1.9890132145779885e-05,
|
| 240 |
-
"loss": 1.
|
| 241 |
"step": 155
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.14766958929395477,
|
| 245 |
-
"grad_norm": 1.
|
| 246 |
"learning_rate": 1.986500656580118e-05,
|
| 247 |
-
"loss": 1.
|
| 248 |
"step": 160
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.15228426395939088,
|
| 252 |
-
"grad_norm":
|
| 253 |
"learning_rate": 1.9837315262594307e-05,
|
| 254 |
-
"loss": 1.
|
| 255 |
"step": 165
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.15689893862482696,
|
| 259 |
-
"grad_norm": 2.
|
| 260 |
"learning_rate": 1.980706543820412e-05,
|
| 261 |
"loss": 1.0367,
|
| 262 |
"step": 170
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.16151361329026304,
|
| 266 |
-
"grad_norm":
|
| 267 |
"learning_rate": 1.9774264960104056e-05,
|
| 268 |
-
"loss": 1.
|
| 269 |
"step": 175
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.16612828795569912,
|
| 273 |
-
"grad_norm":
|
| 274 |
"learning_rate": 1.9738922359149927e-05,
|
| 275 |
-
"loss": 1.
|
| 276 |
"step": 180
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.1707429626211352,
|
| 280 |
-
"grad_norm": 1.
|
| 281 |
"learning_rate": 1.9701046827361175e-05,
|
| 282 |
-
"loss": 1.
|
| 283 |
"step": 185
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.1753576372865713,
|
| 287 |
-
"grad_norm": 1.
|
| 288 |
"learning_rate": 1.9660648215530207e-05,
|
| 289 |
-
"loss": 1.
|
| 290 |
"step": 190
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.17997231195200739,
|
| 294 |
-
"grad_norm": 1.
|
| 295 |
"learning_rate": 1.9617737030660338e-05,
|
| 296 |
-
"loss": 1.
|
| 297 |
"step": 195
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.18458698661744347,
|
| 301 |
-
"grad_norm": 1.
|
| 302 |
"learning_rate": 1.9572324433233122e-05,
|
| 303 |
-
"loss": 1.
|
| 304 |
"step": 200
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.18458698661744347,
|
| 308 |
-
"eval_loss": 1.
|
| 309 |
-
"eval_runtime":
|
| 310 |
-
"eval_samples_per_second": 23.
|
| 311 |
-
"eval_steps_per_second": 0.
|
| 312 |
"step": 200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 0.18920166128287955,
|
| 316 |
-
"grad_norm": 1.
|
| 317 |
"learning_rate": 1.9524422234305677e-05,
|
| 318 |
-
"loss": 1.
|
| 319 |
"step": 205
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.19381633594831565,
|
| 323 |
-
"grad_norm": 1.
|
| 324 |
"learning_rate": 1.9474042892438848e-05,
|
| 325 |
-
"loss": 1.
|
| 326 |
"step": 210
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.19843101061375173,
|
| 330 |
-
"grad_norm":
|
| 331 |
"learning_rate": 1.942119951045692e-05,
|
| 332 |
-
"loss": 1.
|
| 333 |
"step": 215
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.20304568527918782,
|
| 337 |
-
"grad_norm": 1.
|
| 338 |
"learning_rate": 1.9365905832039814e-05,
|
| 339 |
-
"loss": 1.
|
| 340 |
"step": 220
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 0.2076603599446239,
|
| 344 |
-
"grad_norm": 1.
|
| 345 |
"learning_rate": 1.9308176238148565e-05,
|
| 346 |
-
"loss": 1.
|
| 347 |
"step": 225
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 0.21227503461006,
|
| 351 |
-
"grad_norm": 1.
|
| 352 |
"learning_rate": 1.924802574328509e-05,
|
| 353 |
-
"loss": 1.
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.21688970927549608,
|
| 358 |
-
"grad_norm": 1.
|
| 359 |
"learning_rate": 1.9185469991587166e-05,
|
| 360 |
-
"loss": 1.
|
| 361 |
"step": 235
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.22150438394093216,
|
| 365 |
-
"grad_norm": 1.
|
| 366 |
"learning_rate": 1.912052525275965e-05,
|
| 367 |
-
"loss": 1.
|
| 368 |
"step": 240
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.22611905860636825,
|
| 372 |
-
"grad_norm": 1.
|
| 373 |
"learning_rate": 1.905320841784298e-05,
|
| 374 |
-
"loss": 1.
|
| 375 |
"step": 245
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.23073373327180433,
|
| 379 |
-
"grad_norm": 1.
|
| 380 |
"learning_rate": 1.898353699482014e-05,
|
| 381 |
-
"loss": 1.
|
| 382 |
"step": 250
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.23534840793724043,
|
| 386 |
-
"grad_norm": 1.
|
| 387 |
"learning_rate": 1.8911529104063093e-05,
|
| 388 |
-
"loss": 1.
|
| 389 |
"step": 255
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.23996308260267651,
|
| 393 |
-
"grad_norm": 1.
|
| 394 |
"learning_rate": 1.8837203473619978e-05,
|
| 395 |
-
"loss": 1.
|
| 396 |
"step": 260
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.2445777572681126,
|
| 400 |
-
"grad_norm": 1.
|
| 401 |
"learning_rate": 1.8760579434344283e-05,
|
| 402 |
-
"loss": 1.
|
| 403 |
"step": 265
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.24919243193354867,
|
| 407 |
-
"grad_norm": 1.
|
| 408 |
"learning_rate": 1.8681676914867176e-05,
|
| 409 |
-
"loss": 1.
|
| 410 |
"step": 270
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.25380710659898476,
|
| 414 |
-
"grad_norm": 1.
|
| 415 |
"learning_rate": 1.860051643641443e-05,
|
| 416 |
-
"loss": 1.
|
| 417 |
"step": 275
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.25842178126442084,
|
| 421 |
-
"grad_norm": 1.
|
| 422 |
"learning_rate": 1.8517119107469194e-05,
|
| 423 |
-
"loss": 1.
|
| 424 |
"step": 280
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.26303645592985697,
|
| 428 |
-
"grad_norm": 1.
|
| 429 |
"learning_rate": 1.8431506618282e-05,
|
| 430 |
-
"loss": 1.
|
| 431 |
"step": 285
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.26765113059529305,
|
| 435 |
-
"grad_norm": 1.
|
| 436 |
"learning_rate": 1.834370123522954e-05,
|
| 437 |
-
"loss": 1.
|
| 438 |
"step": 290
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.27226580526072913,
|
| 442 |
-
"grad_norm": 1.
|
| 443 |
"learning_rate": 1.8253725795023504e-05,
|
| 444 |
-
"loss": 1.
|
| 445 |
"step": 295
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.2768804799261652,
|
| 449 |
-
"grad_norm": 1.
|
| 450 |
"learning_rate": 1.816160369877117e-05,
|
| 451 |
-
"loss": 1.
|
| 452 |
"step": 300
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.2768804799261652,
|
| 456 |
-
"eval_loss": 1.
|
| 457 |
-
"eval_runtime":
|
| 458 |
-
"eval_samples_per_second":
|
| 459 |
-
"eval_steps_per_second": 0.
|
| 460 |
"step": 300
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 0.2814951545916013,
|
| 464 |
-
"grad_norm": 1.
|
| 465 |
"learning_rate": 1.8067358905889148e-05,
|
| 466 |
-
"loss": 1.
|
| 467 |
"step": 305
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 0.2861098292570374,
|
| 471 |
-
"grad_norm": 1.
|
| 472 |
"learning_rate": 1.797101592787194e-05,
|
| 473 |
-
"loss": 1.
|
| 474 |
"step": 310
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.29072450392247345,
|
| 478 |
-
"grad_norm": 1.
|
| 479 |
"learning_rate": 1.7872599821916922e-05,
|
| 480 |
-
"loss": 1.
|
| 481 |
"step": 315
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 0.29533917858790953,
|
| 485 |
-
"grad_norm": 1.
|
| 486 |
"learning_rate": 1.7772136184407367e-05,
|
| 487 |
-
"loss": 1.
|
| 488 |
"step": 320
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.2999538532533456,
|
| 492 |
-
"grad_norm": 1.
|
| 493 |
"learning_rate": 1.7669651144255265e-05,
|
| 494 |
-
"loss": 1.
|
| 495 |
"step": 325
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.30456852791878175,
|
| 499 |
-
"grad_norm": 1.
|
| 500 |
"learning_rate": 1.7565171356105627e-05,
|
| 501 |
-
"loss": 1.
|
| 502 |
"step": 330
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 0.30918320258421783,
|
| 506 |
-
"grad_norm":
|
| 507 |
"learning_rate": 1.7458723993404065e-05,
|
| 508 |
-
"loss": 1.
|
| 509 |
"step": 335
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 0.3137978772496539,
|
| 513 |
-
"grad_norm": 1.
|
| 514 |
"learning_rate": 1.7350336741329413e-05,
|
| 515 |
-
"loss": 1.
|
| 516 |
"step": 340
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.31841255191509,
|
| 520 |
-
"grad_norm": 1.
|
| 521 |
"learning_rate": 1.7240037789593307e-05,
|
| 522 |
-
"loss": 1.
|
| 523 |
"step": 345
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 0.3230272265805261,
|
| 527 |
-
"grad_norm": 1.
|
| 528 |
"learning_rate": 1.712785582510848e-05,
|
| 529 |
-
"loss": 1.
|
| 530 |
"step": 350
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.32764190124596215,
|
| 534 |
-
"grad_norm": 1.
|
| 535 |
"learning_rate": 1.70138200245278e-05,
|
| 536 |
-
"loss": 1.
|
| 537 |
"step": 355
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.33225657591139823,
|
| 541 |
-
"grad_norm": 2.
|
| 542 |
"learning_rate": 1.6897960046655886e-05,
|
| 543 |
-
"loss": 1.
|
| 544 |
"step": 360
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 0.3368712505768343,
|
| 548 |
-
"grad_norm": 1.
|
| 549 |
"learning_rate": 1.6780306024735384e-05,
|
| 550 |
-
"loss": 1.
|
| 551 |
"step": 365
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 0.3414859252422704,
|
| 555 |
-
"grad_norm":
|
| 556 |
"learning_rate": 1.6660888558609774e-05,
|
| 557 |
-
"loss": 1.
|
| 558 |
"step": 370
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.34610059990770653,
|
| 562 |
-
"grad_norm": 1.
|
| 563 |
"learning_rate": 1.6539738706764895e-05,
|
| 564 |
-
"loss": 1.
|
| 565 |
"step": 375
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 0.3507152745731426,
|
| 569 |
-
"grad_norm": 1.
|
| 570 |
"learning_rate": 1.6416887978251134e-05,
|
| 571 |
-
"loss": 1.
|
| 572 |
"step": 380
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.3553299492385787,
|
| 576 |
-
"grad_norm": 1.
|
| 577 |
"learning_rate": 1.6292368324488462e-05,
|
| 578 |
-
"loss":
|
| 579 |
"step": 385
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.35994462390401477,
|
| 583 |
-
"grad_norm": 1.
|
| 584 |
"learning_rate": 1.6166212130956383e-05,
|
| 585 |
-
"loss": 1.
|
| 586 |
"step": 390
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 0.36455929856945085,
|
| 590 |
-
"grad_norm": 1.
|
| 591 |
"learning_rate": 1.6038452208771037e-05,
|
| 592 |
-
"loss": 1.
|
| 593 |
"step": 395
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 0.36917397323488693,
|
| 597 |
-
"grad_norm": 1.
|
| 598 |
"learning_rate": 1.590912178615157e-05,
|
| 599 |
-
"loss": 1.
|
| 600 |
"step": 400
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 0.36917397323488693,
|
| 604 |
-
"eval_loss": 1.
|
| 605 |
-
"eval_runtime":
|
| 606 |
-
"eval_samples_per_second":
|
| 607 |
-
"eval_steps_per_second": 0.
|
| 608 |
"step": 400
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.373788647900323,
|
| 612 |
-
"grad_norm": 1.
|
| 613 |
"learning_rate": 1.5778254499778006e-05,
|
| 614 |
-
"loss": 1.
|
| 615 |
"step": 405
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 0.3784033225657591,
|
| 619 |
-
"grad_norm": 1.
|
| 620 |
"learning_rate": 1.564588438604296e-05,
|
| 621 |
-
"loss": 1.
|
| 622 |
"step": 410
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 0.3830179972311952,
|
| 626 |
-
"grad_norm": 1.
|
| 627 |
"learning_rate": 1.551204587219928e-05,
|
| 628 |
-
"loss": 0.
|
| 629 |
"step": 415
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 0.3876326718966313,
|
| 633 |
-
"grad_norm": 1.
|
| 634 |
"learning_rate": 1.5376773767406142e-05,
|
| 635 |
-
"loss": 1.
|
| 636 |
"step": 420
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 0.3922473465620674,
|
| 640 |
-
"grad_norm": 1.
|
| 641 |
"learning_rate": 1.5240103253675756e-05,
|
| 642 |
-
"loss": 1.
|
| 643 |
"step": 425
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.39686202122750347,
|
| 647 |
-
"grad_norm": 1.
|
| 648 |
"learning_rate": 1.51020698767231e-05,
|
| 649 |
-
"loss": 1.
|
| 650 |
"step": 430
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.40147669589293955,
|
| 654 |
-
"grad_norm": 1.
|
| 655 |
"learning_rate": 1.4962709536721087e-05,
|
| 656 |
-
"loss": 0.
|
| 657 |
"step": 435
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.40609137055837563,
|
| 661 |
-
"grad_norm": 1.
|
| 662 |
"learning_rate": 1.4822058478963532e-05,
|
| 663 |
-
"loss": 1.
|
| 664 |
"step": 440
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 0.4107060452238117,
|
| 668 |
-
"grad_norm": 1.
|
| 669 |
"learning_rate": 1.4680153284438345e-05,
|
| 670 |
-
"loss": 1.
|
| 671 |
"step": 445
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 0.4153207198892478,
|
| 675 |
-
"grad_norm": 1.
|
| 676 |
"learning_rate": 1.4537030860313443e-05,
|
| 677 |
-
"loss": 1.
|
| 678 |
"step": 450
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 0.41993539455468387,
|
| 682 |
-
"grad_norm": 1.
|
| 683 |
"learning_rate": 1.4392728430337801e-05,
|
| 684 |
-
"loss": 0.
|
| 685 |
"step": 455
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 0.42455006922012,
|
| 689 |
-
"grad_norm": 1.
|
| 690 |
"learning_rate": 1.4247283525160178e-05,
|
| 691 |
-
"loss": 0.
|
| 692 |
"step": 460
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.4291647438855561,
|
| 696 |
-
"grad_norm": 1.
|
| 697 |
"learning_rate": 1.4100733972568038e-05,
|
| 698 |
-
"loss": 1.
|
| 699 |
"step": 465
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.43377941855099217,
|
| 703 |
-
"grad_norm": 1.
|
| 704 |
"learning_rate": 1.3953117887649153e-05,
|
| 705 |
-
"loss": 1.
|
| 706 |
"step": 470
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.43839409321642825,
|
| 710 |
-
"grad_norm": 1.
|
| 711 |
"learning_rate": 1.3804473662878519e-05,
|
| 712 |
-
"loss": 1.
|
| 713 |
"step": 475
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.44300876788186433,
|
| 717 |
-
"grad_norm": 1.
|
| 718 |
"learning_rate": 1.3654839958133118e-05,
|
| 719 |
-
"loss": 1.
|
| 720 |
"step": 480
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.4476234425473004,
|
| 724 |
-
"grad_norm": 1.
|
| 725 |
"learning_rate": 1.3504255690637122e-05,
|
| 726 |
-
"loss": 1.
|
| 727 |
"step": 485
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 0.4522381172127365,
|
| 731 |
-
"grad_norm": 1.
|
| 732 |
"learning_rate": 1.3352760024840174e-05,
|
| 733 |
-
"loss": 0.
|
| 734 |
"step": 490
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.45685279187817257,
|
| 738 |
-
"grad_norm": 1.
|
| 739 |
"learning_rate": 1.3200392362231385e-05,
|
| 740 |
-
"loss": 0.
|
| 741 |
"step": 495
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 0.46146746654360865,
|
| 745 |
-
"grad_norm": 1.
|
| 746 |
"learning_rate": 1.3047192331091636e-05,
|
| 747 |
-
"loss": 1.
|
| 748 |
"step": 500
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 0.46146746654360865,
|
| 752 |
-
"eval_loss": 1.
|
| 753 |
-
"eval_runtime":
|
| 754 |
-
"eval_samples_per_second": 24.
|
| 755 |
-
"eval_steps_per_second": 0.
|
| 756 |
"step": 500
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.4660821412090448,
|
| 760 |
-
"grad_norm": 1.
|
| 761 |
"learning_rate": 1.2893199776186957e-05,
|
| 762 |
-
"loss": 0.
|
| 763 |
"step": 505
|
| 764 |
},
|
| 765 |
{
|
| 766 |
"epoch": 0.47069681587448087,
|
| 767 |
-
"grad_norm": 1.
|
| 768 |
"learning_rate": 1.2738454748405552e-05,
|
| 769 |
-
"loss": 1.
|
| 770 |
"step": 510
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.47531149053991695,
|
| 774 |
-
"grad_norm": 1.
|
| 775 |
"learning_rate": 1.258299749434123e-05,
|
| 776 |
-
"loss": 1.
|
| 777 |
"step": 515
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.47992616520535303,
|
| 781 |
-
"grad_norm": 1.
|
| 782 |
"learning_rate": 1.2426868445825955e-05,
|
| 783 |
-
"loss":
|
| 784 |
"step": 520
|
| 785 |
},
|
| 786 |
{
|
| 787 |
"epoch": 0.4845408398707891,
|
| 788 |
-
"grad_norm": 1.
|
| 789 |
"learning_rate": 1.2270108209414186e-05,
|
| 790 |
-
"loss": 0.
|
| 791 |
"step": 525
|
| 792 |
},
|
| 793 |
{
|
| 794 |
"epoch": 0.4891555145362252,
|
| 795 |
-
"grad_norm": 1.
|
| 796 |
"learning_rate": 1.2112757555821796e-05,
|
| 797 |
-
"loss": 0.
|
| 798 |
"step": 530
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"epoch": 0.49377018920166127,
|
| 802 |
-
"grad_norm": 1.
|
| 803 |
"learning_rate": 1.1954857409322302e-05,
|
| 804 |
-
"loss": 0.
|
| 805 |
"step": 535
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 0.49838486386709735,
|
| 809 |
-
"grad_norm": 1.
|
| 810 |
"learning_rate": 1.179644883710313e-05,
|
| 811 |
-
"loss": 0.
|
| 812 |
"step": 540
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.5029995385325334,
|
| 816 |
-
"grad_norm": 1.
|
| 817 |
"learning_rate": 1.1637573038584729e-05,
|
| 818 |
-
"loss": 0.
|
| 819 |
"step": 545
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.5076142131979695,
|
| 823 |
-
"grad_norm": 1.
|
| 824 |
"learning_rate": 1.1478271334705302e-05,
|
| 825 |
-
"loss": 0.
|
| 826 |
"step": 550
|
| 827 |
},
|
| 828 |
{
|
| 829 |
"epoch": 0.5122288878634056,
|
| 830 |
-
"grad_norm": 1.
|
| 831 |
"learning_rate": 1.1318585157173913e-05,
|
| 832 |
-
"loss": 0.
|
| 833 |
"step": 555
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.5168435625288417,
|
| 837 |
-
"grad_norm": 1.
|
| 838 |
"learning_rate": 1.115855603769479e-05,
|
| 839 |
-
"loss": 0.
|
| 840 |
"step": 560
|
| 841 |
},
|
| 842 |
{
|
| 843 |
"epoch": 0.5214582371942778,
|
| 844 |
-
"grad_norm": 1.
|
| 845 |
"learning_rate": 1.0998225597165628e-05,
|
| 846 |
-
"loss": 0.
|
| 847 |
"step": 565
|
| 848 |
},
|
| 849 |
{
|
| 850 |
"epoch": 0.5260729118597139,
|
| 851 |
-
"grad_norm": 1.
|
| 852 |
"learning_rate": 1.0837635534852687e-05,
|
| 853 |
-
"loss": 0.
|
| 854 |
"step": 570
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.53068758652515,
|
| 858 |
-
"grad_norm": 1.
|
| 859 |
"learning_rate": 1.0676827617545511e-05,
|
| 860 |
-
"loss": 0.
|
| 861 |
"step": 575
|
| 862 |
},
|
| 863 |
{
|
| 864 |
"epoch": 0.5353022611905861,
|
| 865 |
-
"grad_norm": 1.
|
| 866 |
"learning_rate": 1.0515843668694087e-05,
|
| 867 |
-
"loss": 0.
|
| 868 |
"step": 580
|
| 869 |
},
|
| 870 |
{
|
| 871 |
"epoch": 0.5399169358560222,
|
| 872 |
-
"grad_norm": 1.
|
| 873 |
"learning_rate": 1.0354725557531258e-05,
|
| 874 |
-
"loss": 0.
|
| 875 |
"step": 585
|
| 876 |
},
|
| 877 |
{
|
| 878 |
"epoch": 0.5445316105214583,
|
| 879 |
-
"grad_norm": 1.
|
| 880 |
"learning_rate": 1.0193515188183246e-05,
|
| 881 |
-
"loss": 0.
|
| 882 |
"step": 590
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.5491462851868943,
|
| 886 |
-
"grad_norm": 1.
|
| 887 |
"learning_rate": 1.003225448877108e-05,
|
| 888 |
-
"loss": 0.
|
| 889 |
"step": 595
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.5537609598523304,
|
| 893 |
-
"grad_norm": 1.
|
| 894 |
"learning_rate": 9.870985400505805e-06,
|
| 895 |
-
"loss": 0.
|
| 896 |
"step": 600
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.5537609598523304,
|
| 900 |
-
"eval_loss": 0.
|
| 901 |
-
"eval_runtime":
|
| 902 |
-
"eval_samples_per_second": 23.
|
| 903 |
-
"eval_steps_per_second": 0.
|
| 904 |
"step": 600
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 0.5583756345177665,
|
| 908 |
-
"grad_norm": 1.
|
| 909 |
"learning_rate": 9.709749866780248e-06,
|
| 910 |
-
"loss": 0.
|
| 911 |
"step": 605
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 0.5629903091832026,
|
| 915 |
-
"grad_norm": 1.
|
| 916 |
"learning_rate": 9.548589822260281e-06,
|
| 917 |
-
"loss": 0.
|
| 918 |
"step": 610
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.5676049838486387,
|
| 922 |
-
"grad_norm": 1.
|
| 923 |
"learning_rate": 9.387547181978291e-06,
|
| 924 |
-
"loss": 0.
|
| 925 |
"step": 615
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 0.5722196585140747,
|
| 929 |
-
"grad_norm": 1.
|
| 930 |
"learning_rate": 9.226663830431777e-06,
|
| 931 |
-
"loss": 0.
|
| 932 |
"step": 620
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.5768343331795108,
|
| 936 |
-
"grad_norm": 1.
|
| 937 |
"learning_rate": 9.065981610689915e-06,
|
| 938 |
-
"loss": 0.
|
| 939 |
"step": 625
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.5814490078449469,
|
| 943 |
-
"grad_norm": 1.
|
| 944 |
"learning_rate": 8.905542313510846e-06,
|
| 945 |
-
"loss": 0.
|
| 946 |
"step": 630
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 0.586063682510383,
|
| 950 |
-
"grad_norm": 1.
|
| 951 |
"learning_rate": 8.745387666472639e-06,
|
| 952 |
-
"loss": 0.
|
| 953 |
"step": 635
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.5906783571758191,
|
| 957 |
-
"grad_norm": 1.
|
| 958 |
"learning_rate": 8.58555932312059e-06,
|
| 959 |
-
"loss": 0.
|
| 960 |
"step": 640
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.5952930318412551,
|
| 964 |
-
"grad_norm": 1.
|
| 965 |
"learning_rate": 8.426098852133892e-06,
|
| 966 |
-
"loss": 0.
|
| 967 |
"step": 645
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 0.5999077065066912,
|
| 971 |
-
"grad_norm": 1.
|
| 972 |
"learning_rate": 8.267047726514278e-06,
|
| 973 |
-
"loss": 0.
|
| 974 |
"step": 650
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 0.6045223811721273,
|
| 978 |
-
"grad_norm": 1.
|
| 979 |
"learning_rate": 8.108447312799588e-06,
|
| 980 |
-
"loss": 0.
|
| 981 |
"step": 655
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.6091370558375635,
|
| 985 |
-
"grad_norm": 1.
|
| 986 |
"learning_rate": 7.950338860305049e-06,
|
| 987 |
-
"loss": 0.
|
| 988 |
"step": 660
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 0.6137517305029996,
|
| 992 |
-
"grad_norm": 1.
|
| 993 |
"learning_rate": 7.792763490394983e-06,
|
| 994 |
-
"loss": 0.
|
| 995 |
"step": 665
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 0.6183664051684357,
|
| 999 |
-
"grad_norm": 1.
|
| 1000 |
"learning_rate": 7.635762185787868e-06,
|
| 1001 |
-
"loss": 0.
|
| 1002 |
"step": 670
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 0.6229810798338717,
|
| 1006 |
-
"grad_norm": 1.
|
| 1007 |
"learning_rate": 7.479375779897379e-06,
|
| 1008 |
-
"loss": 0.
|
| 1009 |
"step": 675
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 0.6275957544993078,
|
| 1013 |
-
"grad_norm": 1.
|
| 1014 |
"learning_rate": 7.3236449462123315e-06,
|
| 1015 |
-
"loss": 0.
|
| 1016 |
"step": 680
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 0.6322104291647439,
|
| 1020 |
-
"grad_norm": 1.
|
| 1021 |
"learning_rate": 7.168610187718164e-06,
|
| 1022 |
-
"loss": 0.
|
| 1023 |
"step": 685
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.63682510383018,
|
| 1027 |
-
"grad_norm": 1.
|
| 1028 |
"learning_rate": 7.014311826362804e-06,
|
| 1029 |
-
"loss": 0.
|
| 1030 |
"step": 690
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 0.6414397784956161,
|
| 1034 |
-
"grad_norm": 1.
|
| 1035 |
"learning_rate": 6.860789992569601e-06,
|
| 1036 |
-
"loss": 0.
|
| 1037 |
"step": 695
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 0.6460544531610521,
|
| 1041 |
-
"grad_norm": 1.
|
| 1042 |
"learning_rate": 6.708084614800065e-06,
|
| 1043 |
-
"loss": 0.
|
| 1044 |
"step": 700
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.6460544531610521,
|
| 1048 |
-
"eval_loss": 0.
|
| 1049 |
-
"eval_runtime":
|
| 1050 |
-
"eval_samples_per_second": 24.
|
| 1051 |
-
"eval_steps_per_second": 0.
|
| 1052 |
"step": 700
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6506691278264882,
|
| 1056 |
-
"grad_norm": 1.
|
| 1057 |
"learning_rate": 6.556235409169154e-06,
|
| 1058 |
-
"loss": 0.
|
| 1059 |
"step": 705
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.6552838024919243,
|
| 1063 |
-
"grad_norm": 1.
|
| 1064 |
"learning_rate": 6.405281869115768e-06,
|
| 1065 |
-
"loss": 0.
|
| 1066 |
"step": 710
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6598984771573604,
|
| 1070 |
-
"grad_norm": 1.
|
| 1071 |
"learning_rate": 6.255263255131172e-06,
|
| 1072 |
-
"loss": 0.
|
| 1073 |
"step": 715
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6645131518227965,
|
| 1077 |
-
"grad_norm": 1.
|
| 1078 |
"learning_rate": 6.106218584547992e-06,
|
| 1079 |
-
"loss": 0.
|
| 1080 |
"step": 720
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6691278264882325,
|
| 1084 |
-
"grad_norm": 1.
|
| 1085 |
"learning_rate": 5.9581866213924656e-06,
|
| 1086 |
-
"loss": 0.
|
| 1087 |
"step": 725
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6737425011536686,
|
| 1091 |
-
"grad_norm": 1.
|
| 1092 |
"learning_rate": 5.811205866302571e-06,
|
| 1093 |
-
"loss": 0.
|
| 1094 |
"step": 730
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6783571758191047,
|
| 1098 |
-
"grad_norm": 1.
|
| 1099 |
"learning_rate": 5.665314546514633e-06,
|
| 1100 |
-
"loss": 0.
|
| 1101 |
"step": 735
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6829718504845408,
|
| 1105 |
-
"grad_norm": 1.
|
| 1106 |
"learning_rate": 5.520550605921091e-06,
|
| 1107 |
-
"loss": 0.
|
| 1108 |
"step": 740
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.687586525149977,
|
| 1112 |
-
"grad_norm": 1.
|
| 1113 |
"learning_rate": 5.376951695201894e-06,
|
| 1114 |
-
"loss": 0.
|
| 1115 |
"step": 745
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6922011998154131,
|
| 1119 |
-
"grad_norm": 1.
|
| 1120 |
"learning_rate": 5.234555162032221e-06,
|
| 1121 |
-
"loss": 0.
|
| 1122 |
"step": 750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6968158744808491,
|
| 1126 |
-
"grad_norm": 1.
|
| 1127 |
"learning_rate": 5.093398041368942e-06,
|
| 1128 |
-
"loss": 0.
|
| 1129 |
"step": 755
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.7014305491462852,
|
| 1133 |
-
"grad_norm": 1.
|
| 1134 |
"learning_rate": 4.9535170458184735e-06,
|
| 1135 |
-
"loss": 0.
|
| 1136 |
"step": 760
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.7060452238117213,
|
| 1140 |
-
"grad_norm": 1.
|
| 1141 |
"learning_rate": 4.81494855608843e-06,
|
| 1142 |
-
"loss": 0.
|
| 1143 |
"step": 765
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.7106598984771574,
|
| 1147 |
-
"grad_norm": 1.
|
| 1148 |
"learning_rate": 4.677728611525605e-06,
|
| 1149 |
-
"loss": 0.
|
| 1150 |
"step": 770
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.7152745731425935,
|
| 1154 |
-
"grad_norm": 1.
|
| 1155 |
"learning_rate": 4.541892900742757e-06,
|
| 1156 |
-
"loss": 0.
|
| 1157 |
"step": 775
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7198892478080295,
|
| 1161 |
-
"grad_norm": 1.
|
| 1162 |
"learning_rate": 4.407476752336575e-06,
|
| 1163 |
-
"loss": 0.
|
| 1164 |
"step": 780
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7245039224734656,
|
| 1168 |
-
"grad_norm": 1.
|
| 1169 |
"learning_rate": 4.2745151256993325e-06,
|
| 1170 |
-
"loss": 0.
|
| 1171 |
"step": 785
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7291185971389017,
|
| 1175 |
-
"grad_norm": 1.
|
| 1176 |
"learning_rate": 4.143042601926492e-06,
|
| 1177 |
-
"loss": 0.
|
| 1178 |
"step": 790
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.7337332718043378,
|
| 1182 |
-
"grad_norm": 1.
|
| 1183 |
"learning_rate": 4.013093374822789e-06,
|
| 1184 |
-
"loss": 0.
|
| 1185 |
"step": 795
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7383479464697739,
|
| 1189 |
-
"grad_norm": 1.
|
| 1190 |
"learning_rate": 3.884701242008949e-06,
|
| 1191 |
-
"loss": 0.
|
| 1192 |
"step": 800
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7383479464697739,
|
| 1196 |
-
"eval_loss": 0.
|
| 1197 |
-
"eval_runtime":
|
| 1198 |
-
"eval_samples_per_second":
|
| 1199 |
-
"eval_steps_per_second": 0.
|
| 1200 |
"step": 800
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.7429626211352099,
|
| 1204 |
-
"grad_norm": 1.
|
| 1205 |
"learning_rate": 3.757899596131529e-06,
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 805
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.747577295800646,
|
| 1211 |
-
"grad_norm": 1.
|
| 1212 |
"learning_rate": 3.6327214161780287e-06,
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 810
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.7521919704660821,
|
| 1218 |
-
"grad_norm": 1.
|
| 1219 |
"learning_rate": 3.5091992588996026e-06,
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 815
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.7568066451315182,
|
| 1225 |
-
"grad_norm": 1.
|
| 1226 |
"learning_rate": 3.387365250343615e-06,
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 820
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.7614213197969543,
|
| 1232 |
-
"grad_norm": 1.
|
| 1233 |
"learning_rate": 3.2672510774981692e-06,
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 825
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7660359944623903,
|
| 1239 |
-
"grad_norm": 1.
|
| 1240 |
"learning_rate": 3.148887980050872e-06,
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 830
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7706506691278265,
|
| 1246 |
-
"grad_norm": 1.
|
| 1247 |
"learning_rate": 3.032306742263891e-06,
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 835
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7752653437932626,
|
| 1253 |
-
"grad_norm": 1.
|
| 1254 |
"learning_rate": 2.9175376849675076e-06,
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 840
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7798800184586987,
|
| 1260 |
-
"grad_norm": 1.
|
| 1261 |
"learning_rate": 2.8046106576741605e-06,
|
| 1262 |
-
"loss": 0.
|
| 1263 |
"step": 845
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7844946931241348,
|
| 1267 |
-
"grad_norm": 1.
|
| 1268 |
"learning_rate": 2.693555030815085e-06,
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 850
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7891093677895709,
|
| 1274 |
-
"grad_norm": 1.
|
| 1275 |
"learning_rate": 2.5843996881015676e-06,
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 855
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7937240424550069,
|
| 1281 |
-
"grad_norm": 1.
|
| 1282 |
"learning_rate": 2.4771730190127616e-06,
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 860
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.798338717120443,
|
| 1288 |
-
"grad_norm": 1.
|
| 1289 |
"learning_rate": 2.3719029114120716e-06,
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 865
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.8029533917858791,
|
| 1295 |
-
"grad_norm": 1.
|
| 1296 |
"learning_rate": 2.2686167442939733e-06,
|
| 1297 |
-
"loss": 0.
|
| 1298 |
"step": 870
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.8075680664513152,
|
| 1302 |
-
"grad_norm": 1.
|
| 1303 |
"learning_rate": 2.1673413806632104e-06,
|
| 1304 |
-
"loss": 0.
|
| 1305 |
"step": 875
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.8121827411167513,
|
| 1309 |
-
"grad_norm": 1.
|
| 1310 |
"learning_rate": 2.0681031605481563e-06,
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 880
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.8167974157821873,
|
| 1316 |
-
"grad_norm": 1.
|
| 1317 |
"learning_rate": 1.9709278941502363e-06,
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 885
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.8214120904476234,
|
| 1323 |
-
"grad_norm": 1.
|
| 1324 |
"learning_rate": 1.8758408551311048e-06,
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 890
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.8260267651130595,
|
| 1330 |
-
"grad_norm": 1.
|
| 1331 |
"learning_rate": 1.7828667740394045e-06,
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 895
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.8306414397784956,
|
| 1337 |
-
"grad_norm": 1.
|
| 1338 |
"learning_rate": 1.6920298318787532e-06,
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.8306414397784956,
|
| 1344 |
-
"eval_loss": 0.
|
| 1345 |
-
"eval_runtime":
|
| 1346 |
-
"eval_samples_per_second":
|
| 1347 |
-
"eval_steps_per_second": 0.
|
| 1348 |
"step": 900
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 0.8352561144439317,
|
| 1352 |
-
"grad_norm": 1.
|
| 1353 |
"learning_rate": 1.6033536538186778e-06,
|
| 1354 |
-
"loss": 0.
|
| 1355 |
"step": 905
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.8398707891093677,
|
| 1359 |
-
"grad_norm": 1.
|
| 1360 |
"learning_rate": 1.5168613030500922e-06,
|
| 1361 |
-
"loss": 0.
|
| 1362 |
"step": 910
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 0.8444854637748038,
|
| 1366 |
-
"grad_norm": 1.
|
| 1367 |
"learning_rate": 1.4325752747869626e-06,
|
| 1368 |
-
"loss": 0.
|
| 1369 |
"step": 915
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.84910013844024,
|
| 1373 |
-
"grad_norm": 1.
|
| 1374 |
"learning_rate": 1.3505174904156593e-06,
|
| 1375 |
-
"loss": 0.
|
| 1376 |
"step": 920
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 0.8537148131056761,
|
| 1380 |
-
"grad_norm": 1.
|
| 1381 |
"learning_rate": 1.2707092917935914e-06,
|
| 1382 |
-
"loss": 0.
|
| 1383 |
"step": 925
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.8583294877711122,
|
| 1387 |
-
"grad_norm": 1.
|
| 1388 |
"learning_rate": 1.1931714356985257e-06,
|
| 1389 |
-
"loss": 0.
|
| 1390 |
"step": 930
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 0.8629441624365483,
|
| 1394 |
-
"grad_norm": 1.
|
| 1395 |
"learning_rate": 1.1179240884301158e-06,
|
| 1396 |
-
"loss": 0.
|
| 1397 |
"step": 935
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.8675588371019843,
|
| 1401 |
-
"grad_norm": 1.
|
| 1402 |
"learning_rate": 1.0449868205649648e-06,
|
| 1403 |
-
"loss": 0.
|
| 1404 |
"step": 940
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.8721735117674204,
|
| 1408 |
-
"grad_norm": 1.
|
| 1409 |
"learning_rate": 9.74378601866669e-07,
|
| 1410 |
-
"loss": 0.
|
| 1411 |
"step": 945
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.8767881864328565,
|
| 1415 |
-
"grad_norm": 1.
|
| 1416 |
"learning_rate": 9.061177963520751e-07,
|
| 1417 |
-
"loss": 0.
|
| 1418 |
"step": 950
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 0.8814028610982926,
|
| 1422 |
-
"grad_norm": 1.
|
| 1423 |
"learning_rate": 8.402221575151238e-07,
|
| 1424 |
-
"loss": 0.
|
| 1425 |
"step": 955
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.8860175357637287,
|
| 1429 |
-
"grad_norm": 1.
|
| 1430 |
"learning_rate": 7.767088237094578e-07,
|
| 1431 |
-
"loss": 0.
|
| 1432 |
"step": 960
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.8906322104291647,
|
| 1436 |
-
"grad_norm": 1.
|
| 1437 |
"learning_rate": 7.155943136910193e-07,
|
| 1438 |
-
"loss": 0.
|
| 1439 |
"step": 965
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.8952468850946008,
|
| 1443 |
-
"grad_norm": 1.
|
| 1444 |
"learning_rate": 6.568945223218048e-07,
|
| 1445 |
-
"loss": 0.
|
| 1446 |
"step": 970
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.8998615597600369,
|
| 1450 |
-
"grad_norm": 1.
|
| 1451 |
"learning_rate": 6.00624716435868e-07,
|
| 1452 |
-
"loss": 0.
|
| 1453 |
"step": 975
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.904476234425473,
|
| 1457 |
-
"grad_norm": 1.
|
| 1458 |
"learning_rate": 5.467995308686813e-07,
|
| 1459 |
-
"loss": 0.
|
| 1460 |
"step": 980
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.9090909090909091,
|
| 1464 |
-
"grad_norm": 1.
|
| 1465 |
"learning_rate": 4.954329646508505e-07,
|
| 1466 |
-
"loss": 0.
|
| 1467 |
"step": 985
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.9137055837563451,
|
| 1471 |
-
"grad_norm": 1.
|
| 1472 |
"learning_rate": 4.4653837736721273e-07,
|
| 1473 |
-
"loss": 0.
|
| 1474 |
"step": 990
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.9183202584217812,
|
| 1478 |
-
"grad_norm": 1.
|
| 1479 |
"learning_rate": 4.001284856822174e-07,
|
| 1480 |
-
"loss": 0.
|
| 1481 |
"step": 995
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.9229349330872173,
|
| 1485 |
-
"grad_norm": 1.
|
| 1486 |
"learning_rate": 3.562153600325491e-07,
|
| 1487 |
-
"loss": 0.
|
| 1488 |
"step": 1000
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.9229349330872173,
|
| 1492 |
-
"eval_loss": 0.
|
| 1493 |
-
"eval_runtime":
|
| 1494 |
-
"eval_samples_per_second": 24.
|
| 1495 |
-
"eval_steps_per_second": 0.
|
| 1496 |
"step": 1000
|
| 1497 |
},
|
| 1498 |
{
|
| 1499 |
"epoch": 0.9275496077526535,
|
| 1500 |
-
"grad_norm": 1.
|
| 1501 |
"learning_rate": 3.1481042148779674e-07,
|
| 1502 |
-
"loss": 0.
|
| 1503 |
"step": 1005
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.9321642824180896,
|
| 1507 |
-
"grad_norm": 1.
|
| 1508 |
"learning_rate": 2.7592443878003196e-07,
|
| 1509 |
-
"loss": 0.
|
| 1510 |
"step": 1010
|
| 1511 |
},
|
| 1512 |
{
|
| 1513 |
"epoch": 0.9367789570835257,
|
| 1514 |
-
"grad_norm": 1.
|
| 1515 |
"learning_rate": 2.395675255030383e-07,
|
| 1516 |
-
"loss": 0.
|
| 1517 |
"step": 1015
|
| 1518 |
},
|
| 1519 |
{
|
| 1520 |
"epoch": 0.9413936317489617,
|
| 1521 |
-
"grad_norm": 1.
|
| 1522 |
"learning_rate": 2.057491374819365e-07,
|
| 1523 |
-
"loss": 0.
|
| 1524 |
"step": 1020
|
| 1525 |
},
|
| 1526 |
{
|
| 1527 |
"epoch": 0.9460083064143978,
|
| 1528 |
-
"grad_norm": 1.
|
| 1529 |
"learning_rate": 1.7447807031388264e-07,
|
| 1530 |
-
"loss": 0.
|
| 1531 |
"step": 1025
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
"epoch": 0.9506229810798339,
|
| 1535 |
-
"grad_norm": 1.
|
| 1536 |
"learning_rate": 1.457624570804772e-07,
|
| 1537 |
-
"loss": 0.
|
| 1538 |
"step": 1030
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
"epoch": 0.95523765574527,
|
| 1542 |
-
"grad_norm": 1.
|
| 1543 |
"learning_rate": 1.196097662324902e-07,
|
| 1544 |
-
"loss": 0.
|
| 1545 |
"step": 1035
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 0.9598523304107061,
|
| 1549 |
-
"grad_norm": 1.
|
| 1550 |
"learning_rate": 9.602679964744288e-08,
|
| 1551 |
-
"loss": 0.
|
| 1552 |
"step": 1040
|
| 1553 |
},
|
| 1554 |
{
|
| 1555 |
"epoch": 0.9644670050761421,
|
| 1556 |
-
"grad_norm": 1.
|
| 1557 |
"learning_rate": 7.501969086054717e-08,
|
| 1558 |
-
"loss": 0.
|
| 1559 |
"step": 1045
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
"epoch": 0.9690816797415782,
|
| 1563 |
-
"grad_norm": 1.
|
| 1564 |
"learning_rate": 5.659390346948179e-08,
|
| 1565 |
-
"loss": 0.
|
| 1566 |
"step": 1050
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.9736963544070143,
|
| 1570 |
-
"grad_norm": 1.
|
| 1571 |
"learning_rate": 4.075422971340115e-08,
|
| 1572 |
-
"loss": 0.
|
| 1573 |
"step": 1055
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
"epoch": 0.9783110290724504,
|
| 1577 |
-
"grad_norm": 1.
|
| 1578 |
"learning_rate": 2.7504789226548977e-08,
|
| 1579 |
-
"loss": 0.
|
| 1580 |
"step": 1060
|
| 1581 |
},
|
| 1582 |
{
|
| 1583 |
"epoch": 0.9829257037378865,
|
| 1584 |
-
"grad_norm": 1.
|
| 1585 |
"learning_rate": 1.6849027966816535e-08,
|
| 1586 |
-
"loss": 0.
|
| 1587 |
"step": 1065
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
"epoch": 0.9875403784033225,
|
| 1591 |
-
"grad_norm": 1.
|
| 1592 |
"learning_rate": 8.789717319505065e-09,
|
| 1593 |
-
"loss": 0.
|
| 1594 |
"step": 1070
|
| 1595 |
},
|
| 1596 |
{
|
| 1597 |
"epoch": 0.9921550530687586,
|
| 1598 |
-
"grad_norm": 1.
|
| 1599 |
"learning_rate": 3.328953376530164e-09,
|
| 1600 |
-
"loss": 0.
|
| 1601 |
"step": 1075
|
| 1602 |
},
|
| 1603 |
{
|
| 1604 |
"epoch": 0.9967697277341947,
|
| 1605 |
-
"grad_norm": 1.
|
| 1606 |
"learning_rate": 4.681563912700693e-10,
|
| 1607 |
-
"loss": 0.
|
| 1608 |
"step": 1080
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
"epoch": 0.9995385325334564,
|
| 1612 |
"step": 1083,
|
| 1613 |
"total_flos": 453306954547200.0,
|
| 1614 |
-
"train_loss": 0.
|
| 1615 |
-
"train_runtime":
|
| 1616 |
-
"train_samples_per_second":
|
| 1617 |
-
"train_steps_per_second": 0.
|
| 1618 |
}
|
| 1619 |
],
|
| 1620 |
"logging_steps": 5,
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0009229349330872173,
|
| 13 |
+
"grad_norm": 10.726049490177841,
|
| 14 |
"learning_rate": 1.8348623853211012e-07,
|
| 15 |
+
"loss": 1.1497,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.0046146746654360865,
|
| 20 |
+
"grad_norm": 8.900657171326609,
|
| 21 |
"learning_rate": 9.174311926605506e-07,
|
| 22 |
+
"loss": 1.1397,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.009229349330872173,
|
| 27 |
+
"grad_norm": 3.735942046557525,
|
| 28 |
"learning_rate": 1.8348623853211011e-06,
|
| 29 |
+
"loss": 1.0694,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.01384402399630826,
|
| 34 |
+
"grad_norm": 2.8318178389546484,
|
| 35 |
"learning_rate": 2.7522935779816517e-06,
|
| 36 |
+
"loss": 1.0196,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.018458698661744346,
|
| 41 |
+
"grad_norm": 2.114174455764573,
|
| 42 |
"learning_rate": 3.6697247706422022e-06,
|
| 43 |
+
"loss": 1.0136,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.023073373327180433,
|
| 48 |
+
"grad_norm": 2.310487403214644,
|
| 49 |
"learning_rate": 4.587155963302753e-06,
|
| 50 |
+
"loss": 1.0042,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.02768804799261652,
|
| 55 |
+
"grad_norm": 2.4801571550073933,
|
| 56 |
"learning_rate": 5.504587155963303e-06,
|
| 57 |
+
"loss": 0.9742,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.032302722658052604,
|
| 62 |
+
"grad_norm": 2.902534165090561,
|
| 63 |
"learning_rate": 6.422018348623854e-06,
|
| 64 |
+
"loss": 1.0012,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.03691739732348869,
|
| 69 |
+
"grad_norm": 2.2959784184113547,
|
| 70 |
"learning_rate": 7.3394495412844045e-06,
|
| 71 |
+
"loss": 1.0086,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.04153207198892478,
|
| 76 |
+
"grad_norm": 2.3869379848167416,
|
| 77 |
"learning_rate": 8.256880733944956e-06,
|
| 78 |
+
"loss": 0.987,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.046146746654360866,
|
| 83 |
+
"grad_norm": 2.0557468408917527,
|
| 84 |
"learning_rate": 9.174311926605506e-06,
|
| 85 |
+
"loss": 0.9824,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.050761421319796954,
|
| 90 |
+
"grad_norm": 2.2160008605472874,
|
| 91 |
"learning_rate": 1.0091743119266055e-05,
|
| 92 |
+
"loss": 0.986,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.05537609598523304,
|
| 97 |
+
"grad_norm": 2.1320443631302006,
|
| 98 |
"learning_rate": 1.1009174311926607e-05,
|
| 99 |
+
"loss": 1.0019,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.05999077065066913,
|
| 104 |
+
"grad_norm": 2.4823142489717016,
|
| 105 |
"learning_rate": 1.1926605504587156e-05,
|
| 106 |
+
"loss": 1.0048,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.06460544531610521,
|
| 111 |
+
"grad_norm": 2.5334243917693295,
|
| 112 |
"learning_rate": 1.2844036697247708e-05,
|
| 113 |
+
"loss": 0.9828,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.0692201199815413,
|
| 118 |
+
"grad_norm": 2.436109706504398,
|
| 119 |
"learning_rate": 1.3761467889908258e-05,
|
| 120 |
+
"loss": 0.9931,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.07383479464697738,
|
| 125 |
+
"grad_norm": 2.1346723037619695,
|
| 126 |
"learning_rate": 1.4678899082568809e-05,
|
| 127 |
+
"loss": 0.986,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.07844946931241348,
|
| 132 |
+
"grad_norm": 1.9789737582877578,
|
| 133 |
"learning_rate": 1.559633027522936e-05,
|
| 134 |
+
"loss": 1.0089,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.08306414397784956,
|
| 139 |
+
"grad_norm": 2.033387276422637,
|
| 140 |
"learning_rate": 1.6513761467889912e-05,
|
| 141 |
+
"loss": 1.0083,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.08767881864328565,
|
| 146 |
+
"grad_norm": 2.243076459165097,
|
| 147 |
"learning_rate": 1.743119266055046e-05,
|
| 148 |
+
"loss": 1.0175,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.09229349330872173,
|
| 153 |
+
"grad_norm": 1.953405447633714,
|
| 154 |
"learning_rate": 1.834862385321101e-05,
|
| 155 |
+
"loss": 1.025,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.09229349330872173,
|
| 160 |
+
"eval_loss": 1.0240174531936646,
|
| 161 |
+
"eval_runtime": 714.5925,
|
| 162 |
+
"eval_samples_per_second": 21.481,
|
| 163 |
+
"eval_steps_per_second": 0.168,
|
| 164 |
"step": 100
|
| 165 |
},
|
| 166 |
{
|
| 167 |
"epoch": 0.09690816797415783,
|
| 168 |
+
"grad_norm": 2.244939327354232,
|
| 169 |
"learning_rate": 1.9266055045871563e-05,
|
| 170 |
+
"loss": 1.0494,
|
| 171 |
"step": 105
|
| 172 |
},
|
| 173 |
{
|
| 174 |
"epoch": 0.10152284263959391,
|
| 175 |
+
"grad_norm": 2.0565138965118406,
|
| 176 |
"learning_rate": 1.9999947982262415e-05,
|
| 177 |
+
"loss": 1.0345,
|
| 178 |
"step": 110
|
| 179 |
},
|
| 180 |
{
|
| 181 |
"epoch": 0.10613751730503,
|
| 182 |
+
"grad_norm": 2.0332956026689386,
|
| 183 |
"learning_rate": 1.9998127418269004e-05,
|
| 184 |
+
"loss": 1.0454,
|
| 185 |
"step": 115
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 0.11075219197046608,
|
| 189 |
+
"grad_norm": 2.1387249854354566,
|
| 190 |
"learning_rate": 1.9993706508539968e-05,
|
| 191 |
+
"loss": 1.0255,
|
| 192 |
"step": 120
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"epoch": 0.11536686663590216,
|
| 196 |
+
"grad_norm": 2.458364994524856,
|
| 197 |
"learning_rate": 1.998668640288e-05,
|
| 198 |
+
"loss": 1.0518,
|
| 199 |
"step": 125
|
| 200 |
},
|
| 201 |
{
|
| 202 |
"epoch": 0.11998154130133826,
|
| 203 |
+
"grad_norm": 1.8099694655556404,
|
| 204 |
"learning_rate": 1.997706892710117e-05,
|
| 205 |
+
"loss": 1.0224,
|
| 206 |
"step": 130
|
| 207 |
},
|
| 208 |
{
|
| 209 |
"epoch": 0.12459621596677434,
|
| 210 |
+
"grad_norm": 2.202808282950665,
|
| 211 |
"learning_rate": 1.9964856582548094e-05,
|
| 212 |
+
"loss": 1.0552,
|
| 213 |
"step": 135
|
| 214 |
},
|
| 215 |
{
|
| 216 |
"epoch": 0.12921089063221042,
|
| 217 |
+
"grad_norm": 1.7439069720540679,
|
| 218 |
"learning_rate": 1.9950052545447354e-05,
|
| 219 |
+
"loss": 1.0509,
|
| 220 |
"step": 140
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 0.13382556529764653,
|
| 224 |
+
"grad_norm": 1.8051398779741403,
|
| 225 |
"learning_rate": 1.993266066608142e-05,
|
| 226 |
+
"loss": 1.0298,
|
| 227 |
"step": 145
|
| 228 |
},
|
| 229 |
{
|
| 230 |
"epoch": 0.1384402399630826,
|
| 231 |
+
"grad_norm": 1.949872455091144,
|
| 232 |
"learning_rate": 1.991268546778726e-05,
|
| 233 |
+
"loss": 1.0152,
|
| 234 |
"step": 150
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 0.1430549146285187,
|
| 238 |
+
"grad_norm": 2.004864424785268,
|
| 239 |
"learning_rate": 1.9890132145779885e-05,
|
| 240 |
+
"loss": 1.0682,
|
| 241 |
"step": 155
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 0.14766958929395477,
|
| 245 |
+
"grad_norm": 1.9881342422206065,
|
| 246 |
"learning_rate": 1.986500656580118e-05,
|
| 247 |
+
"loss": 1.0275,
|
| 248 |
"step": 160
|
| 249 |
},
|
| 250 |
{
|
| 251 |
"epoch": 0.15228426395939088,
|
| 252 |
+
"grad_norm": 1.895358789872697,
|
| 253 |
"learning_rate": 1.9837315262594307e-05,
|
| 254 |
+
"loss": 1.0341,
|
| 255 |
"step": 165
|
| 256 |
},
|
| 257 |
{
|
| 258 |
"epoch": 0.15689893862482696,
|
| 259 |
+
"grad_norm": 2.089231797232654,
|
| 260 |
"learning_rate": 1.980706543820412e-05,
|
| 261 |
"loss": 1.0367,
|
| 262 |
"step": 170
|
| 263 |
},
|
| 264 |
{
|
| 265 |
"epoch": 0.16151361329026304,
|
| 266 |
+
"grad_norm": 1.7637365585658213,
|
| 267 |
"learning_rate": 1.9774264960104056e-05,
|
| 268 |
+
"loss": 1.0223,
|
| 269 |
"step": 175
|
| 270 |
},
|
| 271 |
{
|
| 272 |
"epoch": 0.16612828795569912,
|
| 273 |
+
"grad_norm": 1.7824319082067301,
|
| 274 |
"learning_rate": 1.9738922359149927e-05,
|
| 275 |
+
"loss": 1.0352,
|
| 276 |
"step": 180
|
| 277 |
},
|
| 278 |
{
|
| 279 |
"epoch": 0.1707429626211352,
|
| 280 |
+
"grad_norm": 1.982313038360383,
|
| 281 |
"learning_rate": 1.9701046827361175e-05,
|
| 282 |
+
"loss": 1.0386,
|
| 283 |
"step": 185
|
| 284 |
},
|
| 285 |
{
|
| 286 |
"epoch": 0.1753576372865713,
|
| 287 |
+
"grad_norm": 1.6545978849182734,
|
| 288 |
"learning_rate": 1.9660648215530207e-05,
|
| 289 |
+
"loss": 1.0247,
|
| 290 |
"step": 190
|
| 291 |
},
|
| 292 |
{
|
| 293 |
"epoch": 0.17997231195200739,
|
| 294 |
+
"grad_norm": 1.782636456859102,
|
| 295 |
"learning_rate": 1.9617737030660338e-05,
|
| 296 |
+
"loss": 1.0305,
|
| 297 |
"step": 195
|
| 298 |
},
|
| 299 |
{
|
| 300 |
"epoch": 0.18458698661744347,
|
| 301 |
+
"grad_norm": 1.8378678572944849,
|
| 302 |
"learning_rate": 1.9572324433233122e-05,
|
| 303 |
+
"loss": 1.033,
|
| 304 |
"step": 200
|
| 305 |
},
|
| 306 |
{
|
| 307 |
"epoch": 0.18458698661744347,
|
| 308 |
+
"eval_loss": 1.0464073419570923,
|
| 309 |
+
"eval_runtime": 647.0101,
|
| 310 |
+
"eval_samples_per_second": 23.725,
|
| 311 |
+
"eval_steps_per_second": 0.185,
|
| 312 |
"step": 200
|
| 313 |
},
|
| 314 |
{
|
| 315 |
"epoch": 0.18920166128287955,
|
| 316 |
+
"grad_norm": 1.6648522003545267,
|
| 317 |
"learning_rate": 1.9524422234305677e-05,
|
| 318 |
+
"loss": 1.0268,
|
| 319 |
"step": 205
|
| 320 |
},
|
| 321 |
{
|
| 322 |
"epoch": 0.19381633594831565,
|
| 323 |
+
"grad_norm": 1.7206083361405007,
|
| 324 |
"learning_rate": 1.9474042892438848e-05,
|
| 325 |
+
"loss": 1.0104,
|
| 326 |
"step": 210
|
| 327 |
},
|
| 328 |
{
|
| 329 |
"epoch": 0.19843101061375173,
|
| 330 |
+
"grad_norm": 2.7587017854007194,
|
| 331 |
"learning_rate": 1.942119951045692e-05,
|
| 332 |
+
"loss": 1.0338,
|
| 333 |
"step": 215
|
| 334 |
},
|
| 335 |
{
|
| 336 |
"epoch": 0.20304568527918782,
|
| 337 |
+
"grad_norm": 1.7930536986404009,
|
| 338 |
"learning_rate": 1.9365905832039814e-05,
|
| 339 |
+
"loss": 1.0614,
|
| 340 |
"step": 220
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 0.2076603599446239,
|
| 344 |
+
"grad_norm": 1.7259048611678551,
|
| 345 |
"learning_rate": 1.9308176238148565e-05,
|
| 346 |
+
"loss": 1.051,
|
| 347 |
"step": 225
|
| 348 |
},
|
| 349 |
{
|
| 350 |
"epoch": 0.21227503461006,
|
| 351 |
+
"grad_norm": 1.9820066538391885,
|
| 352 |
"learning_rate": 1.924802574328509e-05,
|
| 353 |
+
"loss": 1.0259,
|
| 354 |
"step": 230
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 0.21688970927549608,
|
| 358 |
+
"grad_norm": 1.9752532611378077,
|
| 359 |
"learning_rate": 1.9185469991587166e-05,
|
| 360 |
+
"loss": 1.045,
|
| 361 |
"step": 235
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 0.22150438394093216,
|
| 365 |
+
"grad_norm": 1.6132930572572803,
|
| 366 |
"learning_rate": 1.912052525275965e-05,
|
| 367 |
+
"loss": 1.0343,
|
| 368 |
"step": 240
|
| 369 |
},
|
| 370 |
{
|
| 371 |
"epoch": 0.22611905860636825,
|
| 372 |
+
"grad_norm": 1.7584131810475476,
|
| 373 |
"learning_rate": 1.905320841784298e-05,
|
| 374 |
+
"loss": 1.0341,
|
| 375 |
"step": 245
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 0.23073373327180433,
|
| 379 |
+
"grad_norm": 1.70806063018753,
|
| 380 |
"learning_rate": 1.898353699482014e-05,
|
| 381 |
+
"loss": 1.0335,
|
| 382 |
"step": 250
|
| 383 |
},
|
| 384 |
{
|
| 385 |
"epoch": 0.23534840793724043,
|
| 386 |
+
"grad_norm": 1.6516040151613625,
|
| 387 |
"learning_rate": 1.8911529104063093e-05,
|
| 388 |
+
"loss": 1.0388,
|
| 389 |
"step": 255
|
| 390 |
},
|
| 391 |
{
|
| 392 |
"epoch": 0.23996308260267651,
|
| 393 |
+
"grad_norm": 1.6090722840992302,
|
| 394 |
"learning_rate": 1.8837203473619978e-05,
|
| 395 |
+
"loss": 1.0423,
|
| 396 |
"step": 260
|
| 397 |
},
|
| 398 |
{
|
| 399 |
"epoch": 0.2445777572681126,
|
| 400 |
+
"grad_norm": 1.7713270990326546,
|
| 401 |
"learning_rate": 1.8760579434344283e-05,
|
| 402 |
+
"loss": 1.0475,
|
| 403 |
"step": 265
|
| 404 |
},
|
| 405 |
{
|
| 406 |
"epoch": 0.24919243193354867,
|
| 407 |
+
"grad_norm": 1.658315636443391,
|
| 408 |
"learning_rate": 1.8681676914867176e-05,
|
| 409 |
+
"loss": 1.0484,
|
| 410 |
"step": 270
|
| 411 |
},
|
| 412 |
{
|
| 413 |
"epoch": 0.25380710659898476,
|
| 414 |
+
"grad_norm": 1.7916964421225479,
|
| 415 |
"learning_rate": 1.860051643641443e-05,
|
| 416 |
+
"loss": 1.0312,
|
| 417 |
"step": 275
|
| 418 |
},
|
| 419 |
{
|
| 420 |
"epoch": 0.25842178126442084,
|
| 421 |
+
"grad_norm": 1.6115818403548052,
|
| 422 |
"learning_rate": 1.8517119107469194e-05,
|
| 423 |
+
"loss": 1.0406,
|
| 424 |
"step": 280
|
| 425 |
},
|
| 426 |
{
|
| 427 |
"epoch": 0.26303645592985697,
|
| 428 |
+
"grad_norm": 1.5352737986672527,
|
| 429 |
"learning_rate": 1.8431506618282e-05,
|
| 430 |
+
"loss": 1.0308,
|
| 431 |
"step": 285
|
| 432 |
},
|
| 433 |
{
|
| 434 |
"epoch": 0.26765113059529305,
|
| 435 |
+
"grad_norm": 1.6588566683503214,
|
| 436 |
"learning_rate": 1.834370123522954e-05,
|
| 437 |
+
"loss": 1.0427,
|
| 438 |
"step": 290
|
| 439 |
},
|
| 440 |
{
|
| 441 |
"epoch": 0.27226580526072913,
|
| 442 |
+
"grad_norm": 1.5272038076819447,
|
| 443 |
"learning_rate": 1.8253725795023504e-05,
|
| 444 |
+
"loss": 1.0309,
|
| 445 |
"step": 295
|
| 446 |
},
|
| 447 |
{
|
| 448 |
"epoch": 0.2768804799261652,
|
| 449 |
+
"grad_norm": 1.5715430487703328,
|
| 450 |
"learning_rate": 1.816160369877117e-05,
|
| 451 |
+
"loss": 1.037,
|
| 452 |
"step": 300
|
| 453 |
},
|
| 454 |
{
|
| 455 |
"epoch": 0.2768804799261652,
|
| 456 |
+
"eval_loss": 1.0424165725708008,
|
| 457 |
+
"eval_runtime": 645.7635,
|
| 458 |
+
"eval_samples_per_second": 23.77,
|
| 459 |
+
"eval_steps_per_second": 0.186,
|
| 460 |
"step": 300
|
| 461 |
},
|
| 462 |
{
|
| 463 |
"epoch": 0.2814951545916013,
|
| 464 |
+
"grad_norm": 1.7174412533746373,
|
| 465 |
"learning_rate": 1.8067358905889148e-05,
|
| 466 |
+
"loss": 1.0107,
|
| 467 |
"step": 305
|
| 468 |
},
|
| 469 |
{
|
| 470 |
"epoch": 0.2861098292570374,
|
| 471 |
+
"grad_norm": 1.6294029724129888,
|
| 472 |
"learning_rate": 1.797101592787194e-05,
|
| 473 |
+
"loss": 1.0333,
|
| 474 |
"step": 310
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 0.29072450392247345,
|
| 478 |
+
"grad_norm": 1.5569394040938476,
|
| 479 |
"learning_rate": 1.7872599821916922e-05,
|
| 480 |
+
"loss": 1.0253,
|
| 481 |
"step": 315
|
| 482 |
},
|
| 483 |
{
|
| 484 |
"epoch": 0.29533917858790953,
|
| 485 |
+
"grad_norm": 1.5576353282044078,
|
| 486 |
"learning_rate": 1.7772136184407367e-05,
|
| 487 |
+
"loss": 1.0258,
|
| 488 |
"step": 320
|
| 489 |
},
|
| 490 |
{
|
| 491 |
"epoch": 0.2999538532533456,
|
| 492 |
+
"grad_norm": 1.6327760193096779,
|
| 493 |
"learning_rate": 1.7669651144255265e-05,
|
| 494 |
+
"loss": 1.0354,
|
| 495 |
"step": 325
|
| 496 |
},
|
| 497 |
{
|
| 498 |
"epoch": 0.30456852791878175,
|
| 499 |
+
"grad_norm": 1.5902185774151916,
|
| 500 |
"learning_rate": 1.7565171356105627e-05,
|
| 501 |
+
"loss": 1.0473,
|
| 502 |
"step": 330
|
| 503 |
},
|
| 504 |
{
|
| 505 |
"epoch": 0.30918320258421783,
|
| 506 |
+
"grad_norm": 1.6853713064054245,
|
| 507 |
"learning_rate": 1.7458723993404065e-05,
|
| 508 |
+
"loss": 1.0423,
|
| 509 |
"step": 335
|
| 510 |
},
|
| 511 |
{
|
| 512 |
"epoch": 0.3137978772496539,
|
| 513 |
+
"grad_norm": 1.6156935982005598,
|
| 514 |
"learning_rate": 1.7350336741329413e-05,
|
| 515 |
+
"loss": 1.032,
|
| 516 |
"step": 340
|
| 517 |
},
|
| 518 |
{
|
| 519 |
"epoch": 0.31841255191509,
|
| 520 |
+
"grad_norm": 1.6752591408238855,
|
| 521 |
"learning_rate": 1.7240037789593307e-05,
|
| 522 |
+
"loss": 1.0409,
|
| 523 |
"step": 345
|
| 524 |
},
|
| 525 |
{
|
| 526 |
"epoch": 0.3230272265805261,
|
| 527 |
+
"grad_norm": 1.6216674961461026,
|
| 528 |
"learning_rate": 1.712785582510848e-05,
|
| 529 |
+
"loss": 1.0146,
|
| 530 |
"step": 350
|
| 531 |
},
|
| 532 |
{
|
| 533 |
"epoch": 0.32764190124596215,
|
| 534 |
+
"grad_norm": 1.5384588472944032,
|
| 535 |
"learning_rate": 1.70138200245278e-05,
|
| 536 |
+
"loss": 1.0257,
|
| 537 |
"step": 355
|
| 538 |
},
|
| 539 |
{
|
| 540 |
"epoch": 0.33225657591139823,
|
| 541 |
+
"grad_norm": 2.288338756325292,
|
| 542 |
"learning_rate": 1.6897960046655886e-05,
|
| 543 |
+
"loss": 1.033,
|
| 544 |
"step": 360
|
| 545 |
},
|
| 546 |
{
|
| 547 |
"epoch": 0.3368712505768343,
|
| 548 |
+
"grad_norm": 1.7385161337142583,
|
| 549 |
"learning_rate": 1.6780306024735384e-05,
|
| 550 |
+
"loss": 1.0213,
|
| 551 |
"step": 365
|
| 552 |
},
|
| 553 |
{
|
| 554 |
"epoch": 0.3414859252422704,
|
| 555 |
+
"grad_norm": 2.2001962662713885,
|
| 556 |
"learning_rate": 1.6660888558609774e-05,
|
| 557 |
+
"loss": 1.0451,
|
| 558 |
"step": 370
|
| 559 |
},
|
| 560 |
{
|
| 561 |
"epoch": 0.34610059990770653,
|
| 562 |
+
"grad_norm": 1.5300219579349443,
|
| 563 |
"learning_rate": 1.6539738706764895e-05,
|
| 564 |
+
"loss": 1.0282,
|
| 565 |
"step": 375
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 0.3507152745731426,
|
| 569 |
+
"grad_norm": 1.4444114105238783,
|
| 570 |
"learning_rate": 1.6416887978251134e-05,
|
| 571 |
+
"loss": 1.0203,
|
| 572 |
"step": 380
|
| 573 |
},
|
| 574 |
{
|
| 575 |
"epoch": 0.3553299492385787,
|
| 576 |
+
"grad_norm": 1.5804824265139066,
|
| 577 |
"learning_rate": 1.6292368324488462e-05,
|
| 578 |
+
"loss": 1.0012,
|
| 579 |
"step": 385
|
| 580 |
},
|
| 581 |
{
|
| 582 |
"epoch": 0.35994462390401477,
|
| 583 |
+
"grad_norm": 1.4610394973013912,
|
| 584 |
"learning_rate": 1.6166212130956383e-05,
|
| 585 |
+
"loss": 1.0116,
|
| 586 |
"step": 390
|
| 587 |
},
|
| 588 |
{
|
| 589 |
"epoch": 0.36455929856945085,
|
| 590 |
+
"grad_norm": 1.5120140583271204,
|
| 591 |
"learning_rate": 1.6038452208771037e-05,
|
| 592 |
+
"loss": 1.0319,
|
| 593 |
"step": 395
|
| 594 |
},
|
| 595 |
{
|
| 596 |
"epoch": 0.36917397323488693,
|
| 597 |
+
"grad_norm": 1.516281881967303,
|
| 598 |
"learning_rate": 1.590912178615157e-05,
|
| 599 |
+
"loss": 1.0136,
|
| 600 |
"step": 400
|
| 601 |
},
|
| 602 |
{
|
| 603 |
"epoch": 0.36917397323488693,
|
| 604 |
+
"eval_loss": 1.0295383930206299,
|
| 605 |
+
"eval_runtime": 651.8755,
|
| 606 |
+
"eval_samples_per_second": 23.547,
|
| 607 |
+
"eval_steps_per_second": 0.184,
|
| 608 |
"step": 400
|
| 609 |
},
|
| 610 |
{
|
| 611 |
"epoch": 0.373788647900323,
|
| 612 |
+
"grad_norm": 1.7544165660209403,
|
| 613 |
"learning_rate": 1.5778254499778006e-05,
|
| 614 |
+
"loss": 1.0196,
|
| 615 |
"step": 405
|
| 616 |
},
|
| 617 |
{
|
| 618 |
"epoch": 0.3784033225657591,
|
| 619 |
+
"grad_norm": 1.5692875836635374,
|
| 620 |
"learning_rate": 1.564588438604296e-05,
|
| 621 |
+
"loss": 1.0094,
|
| 622 |
"step": 410
|
| 623 |
},
|
| 624 |
{
|
| 625 |
"epoch": 0.3830179972311952,
|
| 626 |
+
"grad_norm": 1.5588154190139185,
|
| 627 |
"learning_rate": 1.551204587219928e-05,
|
| 628 |
+
"loss": 0.9973,
|
| 629 |
"step": 415
|
| 630 |
},
|
| 631 |
{
|
| 632 |
"epoch": 0.3876326718966313,
|
| 633 |
+
"grad_norm": 1.773673892309899,
|
| 634 |
"learning_rate": 1.5376773767406142e-05,
|
| 635 |
+
"loss": 1.0388,
|
| 636 |
"step": 420
|
| 637 |
},
|
| 638 |
{
|
| 639 |
"epoch": 0.3922473465620674,
|
| 640 |
+
"grad_norm": 1.5489285794659653,
|
| 641 |
"learning_rate": 1.5240103253675756e-05,
|
| 642 |
+
"loss": 1.0087,
|
| 643 |
"step": 425
|
| 644 |
},
|
| 645 |
{
|
| 646 |
"epoch": 0.39686202122750347,
|
| 647 |
+
"grad_norm": 1.6551129777825688,
|
| 648 |
"learning_rate": 1.51020698767231e-05,
|
| 649 |
+
"loss": 1.0164,
|
| 650 |
"step": 430
|
| 651 |
},
|
| 652 |
{
|
| 653 |
"epoch": 0.40147669589293955,
|
| 654 |
+
"grad_norm": 1.4090672242535114,
|
| 655 |
"learning_rate": 1.4962709536721087e-05,
|
| 656 |
+
"loss": 0.997,
|
| 657 |
"step": 435
|
| 658 |
},
|
| 659 |
{
|
| 660 |
"epoch": 0.40609137055837563,
|
| 661 |
+
"grad_norm": 1.5171777358410203,
|
| 662 |
"learning_rate": 1.4822058478963532e-05,
|
| 663 |
+
"loss": 1.0132,
|
| 664 |
"step": 440
|
| 665 |
},
|
| 666 |
{
|
| 667 |
"epoch": 0.4107060452238117,
|
| 668 |
+
"grad_norm": 1.4969201768034885,
|
| 669 |
"learning_rate": 1.4680153284438345e-05,
|
| 670 |
+
"loss": 1.0119,
|
| 671 |
"step": 445
|
| 672 |
},
|
| 673 |
{
|
| 674 |
"epoch": 0.4153207198892478,
|
| 675 |
+
"grad_norm": 1.5363202791746906,
|
| 676 |
"learning_rate": 1.4537030860313443e-05,
|
| 677 |
+
"loss": 1.0188,
|
| 678 |
"step": 450
|
| 679 |
},
|
| 680 |
{
|
| 681 |
"epoch": 0.41993539455468387,
|
| 682 |
+
"grad_norm": 1.4963468746071473,
|
| 683 |
"learning_rate": 1.4392728430337801e-05,
|
| 684 |
+
"loss": 0.9952,
|
| 685 |
"step": 455
|
| 686 |
},
|
| 687 |
{
|
| 688 |
"epoch": 0.42455006922012,
|
| 689 |
+
"grad_norm": 1.500810806740765,
|
| 690 |
"learning_rate": 1.4247283525160178e-05,
|
| 691 |
+
"loss": 0.9973,
|
| 692 |
"step": 460
|
| 693 |
},
|
| 694 |
{
|
| 695 |
"epoch": 0.4291647438855561,
|
| 696 |
+
"grad_norm": 1.5508662694848825,
|
| 697 |
"learning_rate": 1.4100733972568038e-05,
|
| 698 |
+
"loss": 1.0085,
|
| 699 |
"step": 465
|
| 700 |
},
|
| 701 |
{
|
| 702 |
"epoch": 0.43377941855099217,
|
| 703 |
+
"grad_norm": 1.5955354844051932,
|
| 704 |
"learning_rate": 1.3953117887649153e-05,
|
| 705 |
+
"loss": 1.0215,
|
| 706 |
"step": 470
|
| 707 |
},
|
| 708 |
{
|
| 709 |
"epoch": 0.43839409321642825,
|
| 710 |
+
"grad_norm": 1.4682043182906732,
|
| 711 |
"learning_rate": 1.3804473662878519e-05,
|
| 712 |
+
"loss": 1.0143,
|
| 713 |
"step": 475
|
| 714 |
},
|
| 715 |
{
|
| 716 |
"epoch": 0.44300876788186433,
|
| 717 |
+
"grad_norm": 1.4980307846950924,
|
| 718 |
"learning_rate": 1.3654839958133118e-05,
|
| 719 |
+
"loss": 1.0026,
|
| 720 |
"step": 480
|
| 721 |
},
|
| 722 |
{
|
| 723 |
"epoch": 0.4476234425473004,
|
| 724 |
+
"grad_norm": 1.5233835818444807,
|
| 725 |
"learning_rate": 1.3504255690637122e-05,
|
| 726 |
+
"loss": 1.0205,
|
| 727 |
"step": 485
|
| 728 |
},
|
| 729 |
{
|
| 730 |
"epoch": 0.4522381172127365,
|
| 731 |
+
"grad_norm": 1.4814525071349245,
|
| 732 |
"learning_rate": 1.3352760024840174e-05,
|
| 733 |
+
"loss": 0.9967,
|
| 734 |
"step": 490
|
| 735 |
},
|
| 736 |
{
|
| 737 |
"epoch": 0.45685279187817257,
|
| 738 |
+
"grad_norm": 1.4499082430240968,
|
| 739 |
"learning_rate": 1.3200392362231385e-05,
|
| 740 |
+
"loss": 0.9842,
|
| 741 |
"step": 495
|
| 742 |
},
|
| 743 |
{
|
| 744 |
"epoch": 0.46146746654360865,
|
| 745 |
+
"grad_norm": 1.5561735389313882,
|
| 746 |
"learning_rate": 1.3047192331091636e-05,
|
| 747 |
+
"loss": 1.0229,
|
| 748 |
"step": 500
|
| 749 |
},
|
| 750 |
{
|
| 751 |
"epoch": 0.46146746654360865,
|
| 752 |
+
"eval_loss": 1.015141248703003,
|
| 753 |
+
"eval_runtime": 633.8229,
|
| 754 |
+
"eval_samples_per_second": 24.218,
|
| 755 |
+
"eval_steps_per_second": 0.189,
|
| 756 |
"step": 500
|
| 757 |
},
|
| 758 |
{
|
| 759 |
"epoch": 0.4660821412090448,
|
| 760 |
+
"grad_norm": 1.4233232765996602,
|
| 761 |
"learning_rate": 1.2893199776186957e-05,
|
| 762 |
+
"loss": 0.9936,
|
| 763 |
"step": 505
|
| 764 |
},
|
| 765 |
{
|
| 766 |
"epoch": 0.47069681587448087,
|
| 767 |
+
"grad_norm": 1.54900822797248,
|
| 768 |
"learning_rate": 1.2738454748405552e-05,
|
| 769 |
+
"loss": 1.0102,
|
| 770 |
"step": 510
|
| 771 |
},
|
| 772 |
{
|
| 773 |
"epoch": 0.47531149053991695,
|
| 774 |
+
"grad_norm": 1.4249354749013639,
|
| 775 |
"learning_rate": 1.258299749434123e-05,
|
| 776 |
+
"loss": 1.013,
|
| 777 |
"step": 515
|
| 778 |
},
|
| 779 |
{
|
| 780 |
"epoch": 0.47992616520535303,
|
| 781 |
+
"grad_norm": 1.468956411146474,
|
| 782 |
"learning_rate": 1.2426868445825955e-05,
|
| 783 |
+
"loss": 1.0027,
|
| 784 |
"step": 520
|
| 785 |
},
|
| 786 |
{
|
| 787 |
"epoch": 0.4845408398707891,
|
| 788 |
+
"grad_norm": 1.515134495058657,
|
| 789 |
"learning_rate": 1.2270108209414186e-05,
|
| 790 |
+
"loss": 0.9825,
|
| 791 |
"step": 525
|
| 792 |
},
|
| 793 |
{
|
| 794 |
"epoch": 0.4891555145362252,
|
| 795 |
+
"grad_norm": 1.49493206284371,
|
| 796 |
"learning_rate": 1.2112757555821796e-05,
|
| 797 |
+
"loss": 0.9968,
|
| 798 |
"step": 530
|
| 799 |
},
|
| 800 |
{
|
| 801 |
"epoch": 0.49377018920166127,
|
| 802 |
+
"grad_norm": 1.494232964423619,
|
| 803 |
"learning_rate": 1.1954857409322302e-05,
|
| 804 |
+
"loss": 0.9808,
|
| 805 |
"step": 535
|
| 806 |
},
|
| 807 |
{
|
| 808 |
"epoch": 0.49838486386709735,
|
| 809 |
+
"grad_norm": 1.5895499778471747,
|
| 810 |
"learning_rate": 1.179644883710313e-05,
|
| 811 |
+
"loss": 0.996,
|
| 812 |
"step": 540
|
| 813 |
},
|
| 814 |
{
|
| 815 |
"epoch": 0.5029995385325334,
|
| 816 |
+
"grad_norm": 1.575516689496947,
|
| 817 |
"learning_rate": 1.1637573038584729e-05,
|
| 818 |
+
"loss": 0.9843,
|
| 819 |
"step": 545
|
| 820 |
},
|
| 821 |
{
|
| 822 |
"epoch": 0.5076142131979695,
|
| 823 |
+
"grad_norm": 1.5289310135121519,
|
| 824 |
"learning_rate": 1.1478271334705302e-05,
|
| 825 |
+
"loss": 0.9897,
|
| 826 |
"step": 550
|
| 827 |
},
|
| 828 |
{
|
| 829 |
"epoch": 0.5122288878634056,
|
| 830 |
+
"grad_norm": 1.487892885517731,
|
| 831 |
"learning_rate": 1.1318585157173913e-05,
|
| 832 |
+
"loss": 0.9965,
|
| 833 |
"step": 555
|
| 834 |
},
|
| 835 |
{
|
| 836 |
"epoch": 0.5168435625288417,
|
| 837 |
+
"grad_norm": 1.504695649448808,
|
| 838 |
"learning_rate": 1.115855603769479e-05,
|
| 839 |
+
"loss": 0.9864,
|
| 840 |
"step": 560
|
| 841 |
},
|
| 842 |
{
|
| 843 |
"epoch": 0.5214582371942778,
|
| 844 |
+
"grad_norm": 1.444258657078223,
|
| 845 |
"learning_rate": 1.0998225597165628e-05,
|
| 846 |
+
"loss": 0.9824,
|
| 847 |
"step": 565
|
| 848 |
},
|
| 849 |
{
|
| 850 |
"epoch": 0.5260729118597139,
|
| 851 |
+
"grad_norm": 1.452291205660523,
|
| 852 |
"learning_rate": 1.0837635534852687e-05,
|
| 853 |
+
"loss": 0.9806,
|
| 854 |
"step": 570
|
| 855 |
},
|
| 856 |
{
|
| 857 |
"epoch": 0.53068758652515,
|
| 858 |
+
"grad_norm": 1.4809970617721466,
|
| 859 |
"learning_rate": 1.0676827617545511e-05,
|
| 860 |
+
"loss": 0.98,
|
| 861 |
"step": 575
|
| 862 |
},
|
| 863 |
{
|
| 864 |
"epoch": 0.5353022611905861,
|
| 865 |
+
"grad_norm": 1.4688234901022226,
|
| 866 |
"learning_rate": 1.0515843668694087e-05,
|
| 867 |
+
"loss": 0.9785,
|
| 868 |
"step": 580
|
| 869 |
},
|
| 870 |
{
|
| 871 |
"epoch": 0.5399169358560222,
|
| 872 |
+
"grad_norm": 1.4825659064745627,
|
| 873 |
"learning_rate": 1.0354725557531258e-05,
|
| 874 |
+
"loss": 0.9776,
|
| 875 |
"step": 585
|
| 876 |
},
|
| 877 |
{
|
| 878 |
"epoch": 0.5445316105214583,
|
| 879 |
+
"grad_norm": 1.3801777122885093,
|
| 880 |
"learning_rate": 1.0193515188183246e-05,
|
| 881 |
+
"loss": 0.9687,
|
| 882 |
"step": 590
|
| 883 |
},
|
| 884 |
{
|
| 885 |
"epoch": 0.5491462851868943,
|
| 886 |
+
"grad_norm": 1.421023225061784,
|
| 887 |
"learning_rate": 1.003225448877108e-05,
|
| 888 |
+
"loss": 0.9964,
|
| 889 |
"step": 595
|
| 890 |
},
|
| 891 |
{
|
| 892 |
"epoch": 0.5537609598523304,
|
| 893 |
+
"grad_norm": 1.3889284539657671,
|
| 894 |
"learning_rate": 9.870985400505805e-06,
|
| 895 |
+
"loss": 0.9745,
|
| 896 |
"step": 600
|
| 897 |
},
|
| 898 |
{
|
| 899 |
"epoch": 0.5537609598523304,
|
| 900 |
+
"eval_loss": 0.9945215582847595,
|
| 901 |
+
"eval_runtime": 661.3891,
|
| 902 |
+
"eval_samples_per_second": 23.209,
|
| 903 |
+
"eval_steps_per_second": 0.181,
|
| 904 |
"step": 600
|
| 905 |
},
|
| 906 |
{
|
| 907 |
"epoch": 0.5583756345177665,
|
| 908 |
+
"grad_norm": 1.4429569586116144,
|
| 909 |
"learning_rate": 9.709749866780248e-06,
|
| 910 |
+
"loss": 0.9805,
|
| 911 |
"step": 605
|
| 912 |
},
|
| 913 |
{
|
| 914 |
"epoch": 0.5629903091832026,
|
| 915 |
+
"grad_norm": 1.4656449742761994,
|
| 916 |
"learning_rate": 9.548589822260281e-06,
|
| 917 |
+
"loss": 0.9895,
|
| 918 |
"step": 610
|
| 919 |
},
|
| 920 |
{
|
| 921 |
"epoch": 0.5676049838486387,
|
| 922 |
+
"grad_norm": 1.3965932035586004,
|
| 923 |
"learning_rate": 9.387547181978291e-06,
|
| 924 |
+
"loss": 0.9744,
|
| 925 |
"step": 615
|
| 926 |
},
|
| 927 |
{
|
| 928 |
"epoch": 0.5722196585140747,
|
| 929 |
+
"grad_norm": 1.359374657149616,
|
| 930 |
"learning_rate": 9.226663830431777e-06,
|
| 931 |
+
"loss": 0.9824,
|
| 932 |
"step": 620
|
| 933 |
},
|
| 934 |
{
|
| 935 |
"epoch": 0.5768343331795108,
|
| 936 |
+
"grad_norm": 1.3668229629199753,
|
| 937 |
"learning_rate": 9.065981610689915e-06,
|
| 938 |
+
"loss": 0.9706,
|
| 939 |
"step": 625
|
| 940 |
},
|
| 941 |
{
|
| 942 |
"epoch": 0.5814490078449469,
|
| 943 |
+
"grad_norm": 1.373500531171451,
|
| 944 |
"learning_rate": 8.905542313510846e-06,
|
| 945 |
+
"loss": 0.9796,
|
| 946 |
"step": 630
|
| 947 |
},
|
| 948 |
{
|
| 949 |
"epoch": 0.586063682510383,
|
| 950 |
+
"grad_norm": 1.4067124446675243,
|
| 951 |
"learning_rate": 8.745387666472639e-06,
|
| 952 |
+
"loss": 0.9879,
|
| 953 |
"step": 635
|
| 954 |
},
|
| 955 |
{
|
| 956 |
"epoch": 0.5906783571758191,
|
| 957 |
+
"grad_norm": 1.436245514521079,
|
| 958 |
"learning_rate": 8.58555932312059e-06,
|
| 959 |
+
"loss": 0.9894,
|
| 960 |
"step": 640
|
| 961 |
},
|
| 962 |
{
|
| 963 |
"epoch": 0.5952930318412551,
|
| 964 |
+
"grad_norm": 1.429504715827128,
|
| 965 |
"learning_rate": 8.426098852133892e-06,
|
| 966 |
+
"loss": 0.9643,
|
| 967 |
"step": 645
|
| 968 |
},
|
| 969 |
{
|
| 970 |
"epoch": 0.5999077065066912,
|
| 971 |
+
"grad_norm": 1.3728127558164411,
|
| 972 |
"learning_rate": 8.267047726514278e-06,
|
| 973 |
+
"loss": 0.9813,
|
| 974 |
"step": 650
|
| 975 |
},
|
| 976 |
{
|
| 977 |
"epoch": 0.6045223811721273,
|
| 978 |
+
"grad_norm": 1.3422366968784711,
|
| 979 |
"learning_rate": 8.108447312799588e-06,
|
| 980 |
+
"loss": 0.972,
|
| 981 |
"step": 655
|
| 982 |
},
|
| 983 |
{
|
| 984 |
"epoch": 0.6091370558375635,
|
| 985 |
+
"grad_norm": 1.4348417465202754,
|
| 986 |
"learning_rate": 7.950338860305049e-06,
|
| 987 |
+
"loss": 0.9638,
|
| 988 |
"step": 660
|
| 989 |
},
|
| 990 |
{
|
| 991 |
"epoch": 0.6137517305029996,
|
| 992 |
+
"grad_norm": 1.3342023162033965,
|
| 993 |
"learning_rate": 7.792763490394983e-06,
|
| 994 |
+
"loss": 0.9733,
|
| 995 |
"step": 665
|
| 996 |
},
|
| 997 |
{
|
| 998 |
"epoch": 0.6183664051684357,
|
| 999 |
+
"grad_norm": 1.361475388045652,
|
| 1000 |
"learning_rate": 7.635762185787868e-06,
|
| 1001 |
+
"loss": 0.9773,
|
| 1002 |
"step": 670
|
| 1003 |
},
|
| 1004 |
{
|
| 1005 |
"epoch": 0.6229810798338717,
|
| 1006 |
+
"grad_norm": 1.3634924688905254,
|
| 1007 |
"learning_rate": 7.479375779897379e-06,
|
| 1008 |
+
"loss": 0.9747,
|
| 1009 |
"step": 675
|
| 1010 |
},
|
| 1011 |
{
|
| 1012 |
"epoch": 0.6275957544993078,
|
| 1013 |
+
"grad_norm": 1.3732265984949414,
|
| 1014 |
"learning_rate": 7.3236449462123315e-06,
|
| 1015 |
+
"loss": 0.9678,
|
| 1016 |
"step": 680
|
| 1017 |
},
|
| 1018 |
{
|
| 1019 |
"epoch": 0.6322104291647439,
|
| 1020 |
+
"grad_norm": 1.4464461120602612,
|
| 1021 |
"learning_rate": 7.168610187718164e-06,
|
| 1022 |
+
"loss": 0.9662,
|
| 1023 |
"step": 685
|
| 1024 |
},
|
| 1025 |
{
|
| 1026 |
"epoch": 0.63682510383018,
|
| 1027 |
+
"grad_norm": 1.3931117990795983,
|
| 1028 |
"learning_rate": 7.014311826362804e-06,
|
| 1029 |
+
"loss": 0.9641,
|
| 1030 |
"step": 690
|
| 1031 |
},
|
| 1032 |
{
|
| 1033 |
"epoch": 0.6414397784956161,
|
| 1034 |
+
"grad_norm": 1.366546097704984,
|
| 1035 |
"learning_rate": 6.860789992569601e-06,
|
| 1036 |
+
"loss": 0.9787,
|
| 1037 |
"step": 695
|
| 1038 |
},
|
| 1039 |
{
|
| 1040 |
"epoch": 0.6460544531610521,
|
| 1041 |
+
"grad_norm": 1.3945778923545584,
|
| 1042 |
"learning_rate": 6.708084614800065e-06,
|
| 1043 |
+
"loss": 0.9441,
|
| 1044 |
"step": 700
|
| 1045 |
},
|
| 1046 |
{
|
| 1047 |
"epoch": 0.6460544531610521,
|
| 1048 |
+
"eval_loss": 0.9769104719161987,
|
| 1049 |
+
"eval_runtime": 633.9092,
|
| 1050 |
+
"eval_samples_per_second": 24.215,
|
| 1051 |
+
"eval_steps_per_second": 0.189,
|
| 1052 |
"step": 700
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 0.6506691278264882,
|
| 1056 |
+
"grad_norm": 1.364918482537208,
|
| 1057 |
"learning_rate": 6.556235409169154e-06,
|
| 1058 |
+
"loss": 0.9437,
|
| 1059 |
"step": 705
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
"epoch": 0.6552838024919243,
|
| 1063 |
+
"grad_norm": 1.3330832614943129,
|
| 1064 |
"learning_rate": 6.405281869115768e-06,
|
| 1065 |
+
"loss": 0.9482,
|
| 1066 |
"step": 710
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
"epoch": 0.6598984771573604,
|
| 1070 |
+
"grad_norm": 1.3291401175998692,
|
| 1071 |
"learning_rate": 6.255263255131172e-06,
|
| 1072 |
+
"loss": 0.9646,
|
| 1073 |
"step": 715
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
"epoch": 0.6645131518227965,
|
| 1077 |
+
"grad_norm": 1.3661394031338707,
|
| 1078 |
"learning_rate": 6.106218584547992e-06,
|
| 1079 |
+
"loss": 0.9649,
|
| 1080 |
"step": 720
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
"epoch": 0.6691278264882325,
|
| 1084 |
+
"grad_norm": 1.3117340443959773,
|
| 1085 |
"learning_rate": 5.9581866213924656e-06,
|
| 1086 |
+
"loss": 0.9525,
|
| 1087 |
"step": 725
|
| 1088 |
},
|
| 1089 |
{
|
| 1090 |
"epoch": 0.6737425011536686,
|
| 1091 |
+
"grad_norm": 1.364658394013176,
|
| 1092 |
"learning_rate": 5.811205866302571e-06,
|
| 1093 |
+
"loss": 0.9516,
|
| 1094 |
"step": 730
|
| 1095 |
},
|
| 1096 |
{
|
| 1097 |
"epoch": 0.6783571758191047,
|
| 1098 |
+
"grad_norm": 1.331994492768848,
|
| 1099 |
"learning_rate": 5.665314546514633e-06,
|
| 1100 |
+
"loss": 0.954,
|
| 1101 |
"step": 735
|
| 1102 |
},
|
| 1103 |
{
|
| 1104 |
"epoch": 0.6829718504845408,
|
| 1105 |
+
"grad_norm": 1.3743467262940992,
|
| 1106 |
"learning_rate": 5.520550605921091e-06,
|
| 1107 |
+
"loss": 0.9554,
|
| 1108 |
"step": 740
|
| 1109 |
},
|
| 1110 |
{
|
| 1111 |
"epoch": 0.687586525149977,
|
| 1112 |
+
"grad_norm": 1.3312291076208118,
|
| 1113 |
"learning_rate": 5.376951695201894e-06,
|
| 1114 |
+
"loss": 0.9565,
|
| 1115 |
"step": 745
|
| 1116 |
},
|
| 1117 |
{
|
| 1118 |
"epoch": 0.6922011998154131,
|
| 1119 |
+
"grad_norm": 1.3832998972367352,
|
| 1120 |
"learning_rate": 5.234555162032221e-06,
|
| 1121 |
+
"loss": 0.9475,
|
| 1122 |
"step": 750
|
| 1123 |
},
|
| 1124 |
{
|
| 1125 |
"epoch": 0.6968158744808491,
|
| 1126 |
+
"grad_norm": 1.389562727942595,
|
| 1127 |
"learning_rate": 5.093398041368942e-06,
|
| 1128 |
+
"loss": 0.9574,
|
| 1129 |
"step": 755
|
| 1130 |
},
|
| 1131 |
{
|
| 1132 |
"epoch": 0.7014305491462852,
|
| 1133 |
+
"grad_norm": 1.3714452844531986,
|
| 1134 |
"learning_rate": 4.9535170458184735e-06,
|
| 1135 |
+
"loss": 0.9581,
|
| 1136 |
"step": 760
|
| 1137 |
},
|
| 1138 |
{
|
| 1139 |
"epoch": 0.7060452238117213,
|
| 1140 |
+
"grad_norm": 1.3477889477630838,
|
| 1141 |
"learning_rate": 4.81494855608843e-06,
|
| 1142 |
+
"loss": 0.9561,
|
| 1143 |
"step": 765
|
| 1144 |
},
|
| 1145 |
{
|
| 1146 |
"epoch": 0.7106598984771574,
|
| 1147 |
+
"grad_norm": 1.4491931180376743,
|
| 1148 |
"learning_rate": 4.677728611525605e-06,
|
| 1149 |
+
"loss": 0.9512,
|
| 1150 |
"step": 770
|
| 1151 |
},
|
| 1152 |
{
|
| 1153 |
"epoch": 0.7152745731425935,
|
| 1154 |
+
"grad_norm": 1.3241497550464327,
|
| 1155 |
"learning_rate": 4.541892900742757e-06,
|
| 1156 |
+
"loss": 0.9422,
|
| 1157 |
"step": 775
|
| 1158 |
},
|
| 1159 |
{
|
| 1160 |
"epoch": 0.7198892478080295,
|
| 1161 |
+
"grad_norm": 1.314421280157553,
|
| 1162 |
"learning_rate": 4.407476752336575e-06,
|
| 1163 |
+
"loss": 0.943,
|
| 1164 |
"step": 780
|
| 1165 |
},
|
| 1166 |
{
|
| 1167 |
"epoch": 0.7245039224734656,
|
| 1168 |
+
"grad_norm": 1.2755970945876594,
|
| 1169 |
"learning_rate": 4.2745151256993325e-06,
|
| 1170 |
+
"loss": 0.9426,
|
| 1171 |
"step": 785
|
| 1172 |
},
|
| 1173 |
{
|
| 1174 |
"epoch": 0.7291185971389017,
|
| 1175 |
+
"grad_norm": 1.332124542587031,
|
| 1176 |
"learning_rate": 4.143042601926492e-06,
|
| 1177 |
+
"loss": 0.9533,
|
| 1178 |
"step": 790
|
| 1179 |
},
|
| 1180 |
{
|
| 1181 |
"epoch": 0.7337332718043378,
|
| 1182 |
+
"grad_norm": 1.3708413423330084,
|
| 1183 |
"learning_rate": 4.013093374822789e-06,
|
| 1184 |
+
"loss": 0.9374,
|
| 1185 |
"step": 795
|
| 1186 |
},
|
| 1187 |
{
|
| 1188 |
"epoch": 0.7383479464697739,
|
| 1189 |
+
"grad_norm": 1.27203160584856,
|
| 1190 |
"learning_rate": 3.884701242008949e-06,
|
| 1191 |
+
"loss": 0.9277,
|
| 1192 |
"step": 800
|
| 1193 |
},
|
| 1194 |
{
|
| 1195 |
"epoch": 0.7383479464697739,
|
| 1196 |
+
"eval_loss": 0.9612703323364258,
|
| 1197 |
+
"eval_runtime": 651.9955,
|
| 1198 |
+
"eval_samples_per_second": 23.543,
|
| 1199 |
+
"eval_steps_per_second": 0.184,
|
| 1200 |
"step": 800
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 0.7429626211352099,
|
| 1204 |
+
"grad_norm": 1.31377359076841,
|
| 1205 |
"learning_rate": 3.757899596131529e-06,
|
| 1206 |
+
"loss": 0.9611,
|
| 1207 |
"step": 805
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 0.747577295800646,
|
| 1211 |
+
"grad_norm": 1.2975998004112579,
|
| 1212 |
"learning_rate": 3.6327214161780287e-06,
|
| 1213 |
+
"loss": 0.9798,
|
| 1214 |
"step": 810
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 0.7521919704660821,
|
| 1218 |
+
"grad_norm": 1.3894464184722144,
|
| 1219 |
"learning_rate": 3.5091992588996026e-06,
|
| 1220 |
+
"loss": 0.9567,
|
| 1221 |
"step": 815
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 0.7568066451315182,
|
| 1225 |
+
"grad_norm": 1.282473554343769,
|
| 1226 |
"learning_rate": 3.387365250343615e-06,
|
| 1227 |
+
"loss": 0.954,
|
| 1228 |
"step": 820
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 0.7614213197969543,
|
| 1232 |
+
"grad_norm": 1.3647805901178591,
|
| 1233 |
"learning_rate": 3.2672510774981692e-06,
|
| 1234 |
+
"loss": 0.9361,
|
| 1235 |
"step": 825
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 0.7660359944623903,
|
| 1239 |
+
"grad_norm": 1.347888877348781,
|
| 1240 |
"learning_rate": 3.148887980050872e-06,
|
| 1241 |
+
"loss": 0.9432,
|
| 1242 |
"step": 830
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 0.7706506691278265,
|
| 1246 |
+
"grad_norm": 1.377090549096584,
|
| 1247 |
"learning_rate": 3.032306742263891e-06,
|
| 1248 |
+
"loss": 0.9519,
|
| 1249 |
"step": 835
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 0.7752653437932626,
|
| 1253 |
+
"grad_norm": 1.287471382339416,
|
| 1254 |
"learning_rate": 2.9175376849675076e-06,
|
| 1255 |
+
"loss": 0.9607,
|
| 1256 |
"step": 840
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 0.7798800184586987,
|
| 1260 |
+
"grad_norm": 1.4070979616878627,
|
| 1261 |
"learning_rate": 2.8046106576741605e-06,
|
| 1262 |
+
"loss": 0.929,
|
| 1263 |
"step": 845
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 0.7844946931241348,
|
| 1267 |
+
"grad_norm": 1.296812382608952,
|
| 1268 |
"learning_rate": 2.693555030815085e-06,
|
| 1269 |
+
"loss": 0.9383,
|
| 1270 |
"step": 850
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 0.7891093677895709,
|
| 1274 |
+
"grad_norm": 1.345957217872087,
|
| 1275 |
"learning_rate": 2.5843996881015676e-06,
|
| 1276 |
+
"loss": 0.9378,
|
| 1277 |
"step": 855
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 0.7937240424550069,
|
| 1281 |
+
"grad_norm": 1.3218142306826084,
|
| 1282 |
"learning_rate": 2.4771730190127616e-06,
|
| 1283 |
+
"loss": 0.9353,
|
| 1284 |
"step": 860
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 0.798338717120443,
|
| 1288 |
+
"grad_norm": 1.2932161650654428,
|
| 1289 |
"learning_rate": 2.3719029114120716e-06,
|
| 1290 |
+
"loss": 0.9333,
|
| 1291 |
"step": 865
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 0.8029533917858791,
|
| 1295 |
+
"grad_norm": 1.36407173107211,
|
| 1296 |
"learning_rate": 2.2686167442939733e-06,
|
| 1297 |
+
"loss": 0.9401,
|
| 1298 |
"step": 870
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 0.8075680664513152,
|
| 1302 |
+
"grad_norm": 1.302064737279862,
|
| 1303 |
"learning_rate": 2.1673413806632104e-06,
|
| 1304 |
+
"loss": 0.939,
|
| 1305 |
"step": 875
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 0.8121827411167513,
|
| 1309 |
+
"grad_norm": 1.3229653318762729,
|
| 1310 |
"learning_rate": 2.0681031605481563e-06,
|
| 1311 |
+
"loss": 0.9355,
|
| 1312 |
"step": 880
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 0.8167974157821873,
|
| 1316 |
+
"grad_norm": 1.3252223231469167,
|
| 1317 |
"learning_rate": 1.9709278941502363e-06,
|
| 1318 |
+
"loss": 0.9344,
|
| 1319 |
"step": 885
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 0.8214120904476234,
|
| 1323 |
+
"grad_norm": 1.3270338774644677,
|
| 1324 |
"learning_rate": 1.8758408551311048e-06,
|
| 1325 |
+
"loss": 0.9321,
|
| 1326 |
"step": 890
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 0.8260267651130595,
|
| 1330 |
+
"grad_norm": 1.3116552675081934,
|
| 1331 |
"learning_rate": 1.7828667740394045e-06,
|
| 1332 |
+
"loss": 0.9513,
|
| 1333 |
"step": 895
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 0.8306414397784956,
|
| 1337 |
+
"grad_norm": 1.2617601634166857,
|
| 1338 |
"learning_rate": 1.6920298318787532e-06,
|
| 1339 |
+
"loss": 0.9384,
|
| 1340 |
"step": 900
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 0.8306414397784956,
|
| 1344 |
+
"eval_loss": 0.9501336812973022,
|
| 1345 |
+
"eval_runtime": 633.7004,
|
| 1346 |
+
"eval_samples_per_second": 24.223,
|
| 1347 |
+
"eval_steps_per_second": 0.189,
|
| 1348 |
"step": 900
|
| 1349 |
},
|
| 1350 |
{
|
| 1351 |
"epoch": 0.8352561144439317,
|
| 1352 |
+
"grad_norm": 1.3036810903046379,
|
| 1353 |
"learning_rate": 1.6033536538186778e-06,
|
| 1354 |
+
"loss": 0.9363,
|
| 1355 |
"step": 905
|
| 1356 |
},
|
| 1357 |
{
|
| 1358 |
"epoch": 0.8398707891093677,
|
| 1359 |
+
"grad_norm": 1.3184782817538288,
|
| 1360 |
"learning_rate": 1.5168613030500922e-06,
|
| 1361 |
+
"loss": 0.9254,
|
| 1362 |
"step": 910
|
| 1363 |
},
|
| 1364 |
{
|
| 1365 |
"epoch": 0.8444854637748038,
|
| 1366 |
+
"grad_norm": 1.3311174144545366,
|
| 1367 |
"learning_rate": 1.4325752747869626e-06,
|
| 1368 |
+
"loss": 0.9401,
|
| 1369 |
"step": 915
|
| 1370 |
},
|
| 1371 |
{
|
| 1372 |
"epoch": 0.84910013844024,
|
| 1373 |
+
"grad_norm": 1.2485403971692124,
|
| 1374 |
"learning_rate": 1.3505174904156593e-06,
|
| 1375 |
+
"loss": 0.9305,
|
| 1376 |
"step": 920
|
| 1377 |
},
|
| 1378 |
{
|
| 1379 |
"epoch": 0.8537148131056761,
|
| 1380 |
+
"grad_norm": 1.302635001110673,
|
| 1381 |
"learning_rate": 1.2707092917935914e-06,
|
| 1382 |
+
"loss": 0.9393,
|
| 1383 |
"step": 925
|
| 1384 |
},
|
| 1385 |
{
|
| 1386 |
"epoch": 0.8583294877711122,
|
| 1387 |
+
"grad_norm": 1.3633354249041523,
|
| 1388 |
"learning_rate": 1.1931714356985257e-06,
|
| 1389 |
+
"loss": 0.9312,
|
| 1390 |
"step": 930
|
| 1391 |
},
|
| 1392 |
{
|
| 1393 |
"epoch": 0.8629441624365483,
|
| 1394 |
+
"grad_norm": 1.2815769915508204,
|
| 1395 |
"learning_rate": 1.1179240884301158e-06,
|
| 1396 |
+
"loss": 0.9217,
|
| 1397 |
"step": 935
|
| 1398 |
},
|
| 1399 |
{
|
| 1400 |
"epoch": 0.8675588371019843,
|
| 1401 |
+
"grad_norm": 1.3419818473322924,
|
| 1402 |
"learning_rate": 1.0449868205649648e-06,
|
| 1403 |
+
"loss": 0.9168,
|
| 1404 |
"step": 940
|
| 1405 |
},
|
| 1406 |
{
|
| 1407 |
"epoch": 0.8721735117674204,
|
| 1408 |
+
"grad_norm": 1.3006753146842553,
|
| 1409 |
"learning_rate": 9.74378601866669e-07,
|
| 1410 |
+
"loss": 0.9413,
|
| 1411 |
"step": 945
|
| 1412 |
},
|
| 1413 |
{
|
| 1414 |
"epoch": 0.8767881864328565,
|
| 1415 |
+
"grad_norm": 1.275285958470618,
|
| 1416 |
"learning_rate": 9.061177963520751e-07,
|
| 1417 |
+
"loss": 0.9446,
|
| 1418 |
"step": 950
|
| 1419 |
},
|
| 1420 |
{
|
| 1421 |
"epoch": 0.8814028610982926,
|
| 1422 |
+
"grad_norm": 1.2680303522716787,
|
| 1423 |
"learning_rate": 8.402221575151238e-07,
|
| 1424 |
+
"loss": 0.9161,
|
| 1425 |
"step": 955
|
| 1426 |
},
|
| 1427 |
{
|
| 1428 |
"epoch": 0.8860175357637287,
|
| 1429 |
+
"grad_norm": 1.2495043326934117,
|
| 1430 |
"learning_rate": 7.767088237094578e-07,
|
| 1431 |
+
"loss": 0.9333,
|
| 1432 |
"step": 960
|
| 1433 |
},
|
| 1434 |
{
|
| 1435 |
"epoch": 0.8906322104291647,
|
| 1436 |
+
"grad_norm": 1.3345940160548069,
|
| 1437 |
"learning_rate": 7.155943136910193e-07,
|
| 1438 |
+
"loss": 0.9353,
|
| 1439 |
"step": 965
|
| 1440 |
},
|
| 1441 |
{
|
| 1442 |
"epoch": 0.8952468850946008,
|
| 1443 |
+
"grad_norm": 1.3252439901574087,
|
| 1444 |
"learning_rate": 6.568945223218048e-07,
|
| 1445 |
+
"loss": 0.9381,
|
| 1446 |
"step": 970
|
| 1447 |
},
|
| 1448 |
{
|
| 1449 |
"epoch": 0.8998615597600369,
|
| 1450 |
+
"grad_norm": 1.2795828639710098,
|
| 1451 |
"learning_rate": 6.00624716435868e-07,
|
| 1452 |
+
"loss": 0.9199,
|
| 1453 |
"step": 975
|
| 1454 |
},
|
| 1455 |
{
|
| 1456 |
"epoch": 0.904476234425473,
|
| 1457 |
+
"grad_norm": 1.3413550832303935,
|
| 1458 |
"learning_rate": 5.467995308686813e-07,
|
| 1459 |
+
"loss": 0.94,
|
| 1460 |
"step": 980
|
| 1461 |
},
|
| 1462 |
{
|
| 1463 |
"epoch": 0.9090909090909091,
|
| 1464 |
+
"grad_norm": 1.3436022079621268,
|
| 1465 |
"learning_rate": 4.954329646508505e-07,
|
| 1466 |
+
"loss": 0.9313,
|
| 1467 |
"step": 985
|
| 1468 |
},
|
| 1469 |
{
|
| 1470 |
"epoch": 0.9137055837563451,
|
| 1471 |
+
"grad_norm": 1.3709638383836422,
|
| 1472 |
"learning_rate": 4.4653837736721273e-07,
|
| 1473 |
+
"loss": 0.9346,
|
| 1474 |
"step": 990
|
| 1475 |
},
|
| 1476 |
{
|
| 1477 |
"epoch": 0.9183202584217812,
|
| 1478 |
+
"grad_norm": 1.2668090185362362,
|
| 1479 |
"learning_rate": 4.001284856822174e-07,
|
| 1480 |
+
"loss": 0.9408,
|
| 1481 |
"step": 995
|
| 1482 |
},
|
| 1483 |
{
|
| 1484 |
"epoch": 0.9229349330872173,
|
| 1485 |
+
"grad_norm": 1.2697339769613498,
|
| 1486 |
"learning_rate": 3.562153600325491e-07,
|
| 1487 |
+
"loss": 0.9216,
|
| 1488 |
"step": 1000
|
| 1489 |
},
|
| 1490 |
{
|
| 1491 |
"epoch": 0.9229349330872173,
|
| 1492 |
+
"eval_loss": 0.9448357224464417,
|
| 1493 |
+
"eval_runtime": 634.4549,
|
| 1494 |
+
"eval_samples_per_second": 24.194,
|
| 1495 |
+
"eval_steps_per_second": 0.189,
|
| 1496 |
"step": 1000
|
| 1497 |
},
|
| 1498 |
{
|
| 1499 |
"epoch": 0.9275496077526535,
|
| 1500 |
+
"grad_norm": 1.281655332093928,
|
| 1501 |
"learning_rate": 3.1481042148779674e-07,
|
| 1502 |
+
"loss": 0.9399,
|
| 1503 |
"step": 1005
|
| 1504 |
},
|
| 1505 |
{
|
| 1506 |
"epoch": 0.9321642824180896,
|
| 1507 |
+
"grad_norm": 1.3600956214897377,
|
| 1508 |
"learning_rate": 2.7592443878003196e-07,
|
| 1509 |
+
"loss": 0.9408,
|
| 1510 |
"step": 1010
|
| 1511 |
},
|
| 1512 |
{
|
| 1513 |
"epoch": 0.9367789570835257,
|
| 1514 |
+
"grad_norm": 1.3158687412306498,
|
| 1515 |
"learning_rate": 2.395675255030383e-07,
|
| 1516 |
+
"loss": 0.9227,
|
| 1517 |
"step": 1015
|
| 1518 |
},
|
| 1519 |
{
|
| 1520 |
"epoch": 0.9413936317489617,
|
| 1521 |
+
"grad_norm": 1.321042213237677,
|
| 1522 |
"learning_rate": 2.057491374819365e-07,
|
| 1523 |
+
"loss": 0.9251,
|
| 1524 |
"step": 1020
|
| 1525 |
},
|
| 1526 |
{
|
| 1527 |
"epoch": 0.9460083064143978,
|
| 1528 |
+
"grad_norm": 1.2706676798657595,
|
| 1529 |
"learning_rate": 1.7447807031388264e-07,
|
| 1530 |
+
"loss": 0.9332,
|
| 1531 |
"step": 1025
|
| 1532 |
},
|
| 1533 |
{
|
| 1534 |
"epoch": 0.9506229810798339,
|
| 1535 |
+
"grad_norm": 1.3025080688365438,
|
| 1536 |
"learning_rate": 1.457624570804772e-07,
|
| 1537 |
+
"loss": 0.9171,
|
| 1538 |
"step": 1030
|
| 1539 |
},
|
| 1540 |
{
|
| 1541 |
"epoch": 0.95523765574527,
|
| 1542 |
+
"grad_norm": 1.3183940856356338,
|
| 1543 |
"learning_rate": 1.196097662324902e-07,
|
| 1544 |
+
"loss": 0.9394,
|
| 1545 |
"step": 1035
|
| 1546 |
},
|
| 1547 |
{
|
| 1548 |
"epoch": 0.9598523304107061,
|
| 1549 |
+
"grad_norm": 1.2959502551091049,
|
| 1550 |
"learning_rate": 9.602679964744288e-08,
|
| 1551 |
+
"loss": 0.9171,
|
| 1552 |
"step": 1040
|
| 1553 |
},
|
| 1554 |
{
|
| 1555 |
"epoch": 0.9644670050761421,
|
| 1556 |
+
"grad_norm": 1.3183746243018815,
|
| 1557 |
"learning_rate": 7.501969086054717e-08,
|
| 1558 |
+
"loss": 0.9328,
|
| 1559 |
"step": 1045
|
| 1560 |
},
|
| 1561 |
{
|
| 1562 |
"epoch": 0.9690816797415782,
|
| 1563 |
+
"grad_norm": 1.3307943534831945,
|
| 1564 |
"learning_rate": 5.659390346948179e-08,
|
| 1565 |
+
"loss": 0.9424,
|
| 1566 |
"step": 1050
|
| 1567 |
},
|
| 1568 |
{
|
| 1569 |
"epoch": 0.9736963544070143,
|
| 1570 |
+
"grad_norm": 1.2730525780758684,
|
| 1571 |
"learning_rate": 4.075422971340115e-08,
|
| 1572 |
+
"loss": 0.9402,
|
| 1573 |
"step": 1055
|
| 1574 |
},
|
| 1575 |
{
|
| 1576 |
"epoch": 0.9783110290724504,
|
| 1577 |
+
"grad_norm": 1.3093625889696177,
|
| 1578 |
"learning_rate": 2.7504789226548977e-08,
|
| 1579 |
+
"loss": 0.9275,
|
| 1580 |
"step": 1060
|
| 1581 |
},
|
| 1582 |
{
|
| 1583 |
"epoch": 0.9829257037378865,
|
| 1584 |
+
"grad_norm": 1.2514296608879,
|
| 1585 |
"learning_rate": 1.6849027966816535e-08,
|
| 1586 |
+
"loss": 0.9269,
|
| 1587 |
"step": 1065
|
| 1588 |
},
|
| 1589 |
{
|
| 1590 |
"epoch": 0.9875403784033225,
|
| 1591 |
+
"grad_norm": 1.273379135333167,
|
| 1592 |
"learning_rate": 8.789717319505065e-09,
|
| 1593 |
+
"loss": 0.9362,
|
| 1594 |
"step": 1070
|
| 1595 |
},
|
| 1596 |
{
|
| 1597 |
"epoch": 0.9921550530687586,
|
| 1598 |
+
"grad_norm": 1.2811332538414983,
|
| 1599 |
"learning_rate": 3.328953376530164e-09,
|
| 1600 |
+
"loss": 0.9313,
|
| 1601 |
"step": 1075
|
| 1602 |
},
|
| 1603 |
{
|
| 1604 |
"epoch": 0.9967697277341947,
|
| 1605 |
+
"grad_norm": 1.3117234277097758,
|
| 1606 |
"learning_rate": 4.681563912700693e-10,
|
| 1607 |
+
"loss": 0.9204,
|
| 1608 |
"step": 1080
|
| 1609 |
},
|
| 1610 |
{
|
| 1611 |
"epoch": 0.9995385325334564,
|
| 1612 |
"step": 1083,
|
| 1613 |
"total_flos": 453306954547200.0,
|
| 1614 |
+
"train_loss": 0.986508995762382,
|
| 1615 |
+
"train_runtime": 33955.1767,
|
| 1616 |
+
"train_samples_per_second": 4.084,
|
| 1617 |
+
"train_steps_per_second": 0.032
|
| 1618 |
}
|
| 1619 |
],
|
| 1620 |
"logging_steps": 5,
|