🍻 cheers
Browse files- README.md +4 -3
- all_results.json +11 -11
- eval_results.json +6 -6
- runs/Mar17_19-33-07_add18dc15495/events.out.tfevents.1710704648.add18dc15495.3839.4 +3 -0
- train_results.json +6 -6
- trainer_state.json +68 -825
README.md
CHANGED
|
@@ -2,6 +2,7 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224
|
| 4 |
tags:
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
metrics:
|
| 7 |
- accuracy
|
|
@@ -15,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# vit-lr-0.0001
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: google/vit-base-patch16-224
|
| 4 |
tags:
|
| 5 |
+
- image-classification
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- accuracy
|
|
|
|
| 16 |
|
| 17 |
# vit-lr-0.0001
|
| 18 |
|
| 19 |
+
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the skin-cancer dataset.
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
+
- Loss: 0.4974
|
| 22 |
+
- Accuracy: 0.8169
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
all_results.json
CHANGED
|
@@ -1,13 +1,13 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 40.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 8.
|
| 8 |
-
"total_flos":
|
| 9 |
-
"train_loss": 0.
|
| 10 |
-
"train_runtime":
|
| 11 |
-
"train_samples_per_second":
|
| 12 |
-
"train_steps_per_second":
|
| 13 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.18,
|
| 3 |
+
"eval_accuracy": 0.8169209431345353,
|
| 4 |
+
"eval_loss": 0.49744343757629395,
|
| 5 |
+
"eval_runtime": 40.9787,
|
| 6 |
+
"eval_samples_per_second": 70.378,
|
| 7 |
+
"eval_steps_per_second": 8.809,
|
| 8 |
+
"total_flos": 8.667092503134536e+17,
|
| 9 |
+
"train_loss": 0.4869789886474609,
|
| 10 |
+
"train_runtime": 547.2513,
|
| 11 |
+
"train_samples_per_second": 937.047,
|
| 12 |
+
"train_steps_per_second": 58.657
|
| 13 |
}
|
eval_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 40.
|
| 6 |
-
"eval_samples_per_second":
|
| 7 |
-
"eval_steps_per_second": 8.
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.18,
|
| 3 |
+
"eval_accuracy": 0.8169209431345353,
|
| 4 |
+
"eval_loss": 0.49744343757629395,
|
| 5 |
+
"eval_runtime": 40.9787,
|
| 6 |
+
"eval_samples_per_second": 70.378,
|
| 7 |
+
"eval_steps_per_second": 8.809
|
| 8 |
}
|
runs/Mar17_19-33-07_add18dc15495/events.out.tfevents.1710704648.add18dc15495.3839.4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8a580650500e0fec83150a051bcf3229ae6bbae2e90f9f1a108a0e796d5f241
|
| 3 |
+
size 734
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 2.18,
|
| 3 |
+
"total_flos": 8.667092503134536e+17,
|
| 4 |
+
"train_loss": 0.4869789886474609,
|
| 5 |
+
"train_runtime": 547.2513,
|
| 6 |
+
"train_samples_per_second": 937.047,
|
| 7 |
+
"train_steps_per_second": 58.657
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,898 +1,141 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./vit-lr-0.0001/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 100,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.03,
|
| 13 |
-
"grad_norm": 3.3859810829162598,
|
| 14 |
-
"learning_rate": 9.997196261682243e-05,
|
| 15 |
-
"loss": 1.1978,
|
| 16 |
-
"step": 10
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.06,
|
| 20 |
-
"grad_norm": 4.35668420791626,
|
| 21 |
-
"learning_rate": 9.994080996884736e-05,
|
| 22 |
-
"loss": 0.717,
|
| 23 |
-
"step": 20
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.09,
|
| 27 |
-
"grad_norm": 5.1708455085754395,
|
| 28 |
-
"learning_rate": 9.990965732087227e-05,
|
| 29 |
-
"loss": 0.8267,
|
| 30 |
-
"step": 30
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.12,
|
| 34 |
-
"grad_norm": 9.000005722045898,
|
| 35 |
-
"learning_rate": 9.98785046728972e-05,
|
| 36 |
-
"loss": 1.0157,
|
| 37 |
-
"step": 40
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.16,
|
| 41 |
-
"grad_norm": 4.164905071258545,
|
| 42 |
-
"learning_rate": 9.984735202492212e-05,
|
| 43 |
-
"loss": 0.7406,
|
| 44 |
-
"step": 50
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.19,
|
| 48 |
-
"grad_norm": 3.7239935398101807,
|
| 49 |
-
"learning_rate": 9.981619937694705e-05,
|
| 50 |
-
"loss": 0.7899,
|
| 51 |
-
"step": 60
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.22,
|
| 55 |
-
"grad_norm": 3.9820024967193604,
|
| 56 |
-
"learning_rate": 9.978504672897196e-05,
|
| 57 |
-
"loss": 0.816,
|
| 58 |
-
"step": 70
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.25,
|
| 62 |
-
"grad_norm": 3.7813899517059326,
|
| 63 |
-
"learning_rate": 9.975389408099689e-05,
|
| 64 |
-
"loss": 0.6441,
|
| 65 |
-
"step": 80
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.28,
|
| 69 |
-
"grad_norm": 6.419355869293213,
|
| 70 |
-
"learning_rate": 9.972274143302182e-05,
|
| 71 |
-
"loss": 0.6568,
|
| 72 |
-
"step": 90
|
| 73 |
-
},
|
| 74 |
{
|
| 75 |
"epoch": 0.31,
|
| 76 |
-
"grad_norm":
|
| 77 |
-
"learning_rate": 9.
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 100
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.31,
|
| 83 |
-
"eval_accuracy": 0.
|
| 84 |
-
"eval_loss": 0.
|
| 85 |
-
"eval_runtime": 39.
|
| 86 |
-
"eval_samples_per_second": 72.
|
| 87 |
-
"eval_steps_per_second": 9.
|
| 88 |
"step": 100
|
| 89 |
},
|
| 90 |
-
{
|
| 91 |
-
"epoch": 0.34,
|
| 92 |
-
"grad_norm": 4.352388858795166,
|
| 93 |
-
"learning_rate": 9.966043613707165e-05,
|
| 94 |
-
"loss": 0.6567,
|
| 95 |
-
"step": 110
|
| 96 |
-
},
|
| 97 |
-
{
|
| 98 |
-
"epoch": 0.37,
|
| 99 |
-
"grad_norm": 6.960423946380615,
|
| 100 |
-
"learning_rate": 9.962928348909658e-05,
|
| 101 |
-
"loss": 0.643,
|
| 102 |
-
"step": 120
|
| 103 |
-
},
|
| 104 |
-
{
|
| 105 |
-
"epoch": 0.4,
|
| 106 |
-
"grad_norm": 8.54637336730957,
|
| 107 |
-
"learning_rate": 9.95981308411215e-05,
|
| 108 |
-
"loss": 0.6606,
|
| 109 |
-
"step": 130
|
| 110 |
-
},
|
| 111 |
-
{
|
| 112 |
-
"epoch": 0.44,
|
| 113 |
-
"grad_norm": 1.6916728019714355,
|
| 114 |
-
"learning_rate": 9.956697819314643e-05,
|
| 115 |
-
"loss": 0.5345,
|
| 116 |
-
"step": 140
|
| 117 |
-
},
|
| 118 |
-
{
|
| 119 |
-
"epoch": 0.47,
|
| 120 |
-
"grad_norm": 5.490638732910156,
|
| 121 |
-
"learning_rate": 9.953582554517134e-05,
|
| 122 |
-
"loss": 0.6008,
|
| 123 |
-
"step": 150
|
| 124 |
-
},
|
| 125 |
-
{
|
| 126 |
-
"epoch": 0.5,
|
| 127 |
-
"grad_norm": 6.159054279327393,
|
| 128 |
-
"learning_rate": 9.950467289719627e-05,
|
| 129 |
-
"loss": 0.5466,
|
| 130 |
-
"step": 160
|
| 131 |
-
},
|
| 132 |
-
{
|
| 133 |
-
"epoch": 0.53,
|
| 134 |
-
"grad_norm": 2.6157217025756836,
|
| 135 |
-
"learning_rate": 9.947352024922119e-05,
|
| 136 |
-
"loss": 0.5968,
|
| 137 |
-
"step": 170
|
| 138 |
-
},
|
| 139 |
-
{
|
| 140 |
-
"epoch": 0.56,
|
| 141 |
-
"grad_norm": 7.270638465881348,
|
| 142 |
-
"learning_rate": 9.944236760124612e-05,
|
| 143 |
-
"loss": 0.6522,
|
| 144 |
-
"step": 180
|
| 145 |
-
},
|
| 146 |
-
{
|
| 147 |
-
"epoch": 0.59,
|
| 148 |
-
"grad_norm": 6.296653747558594,
|
| 149 |
-
"learning_rate": 9.941121495327103e-05,
|
| 150 |
-
"loss": 0.5591,
|
| 151 |
-
"step": 190
|
| 152 |
-
},
|
| 153 |
{
|
| 154 |
"epoch": 0.62,
|
| 155 |
-
"grad_norm": 6.
|
| 156 |
-
"learning_rate": 9.
|
| 157 |
-
"loss": 0.
|
| 158 |
"step": 200
|
| 159 |
},
|
| 160 |
{
|
| 161 |
"epoch": 0.62,
|
| 162 |
-
"eval_accuracy": 0.
|
| 163 |
-
"eval_loss": 0.
|
| 164 |
-
"eval_runtime": 39.
|
| 165 |
-
"eval_samples_per_second":
|
| 166 |
-
"eval_steps_per_second": 9.
|
| 167 |
"step": 200
|
| 168 |
},
|
| 169 |
-
{
|
| 170 |
-
"epoch": 0.65,
|
| 171 |
-
"grad_norm": 4.321172714233398,
|
| 172 |
-
"learning_rate": 9.934890965732088e-05,
|
| 173 |
-
"loss": 0.606,
|
| 174 |
-
"step": 210
|
| 175 |
-
},
|
| 176 |
-
{
|
| 177 |
-
"epoch": 0.69,
|
| 178 |
-
"grad_norm": 6.226727485656738,
|
| 179 |
-
"learning_rate": 9.93177570093458e-05,
|
| 180 |
-
"loss": 0.7798,
|
| 181 |
-
"step": 220
|
| 182 |
-
},
|
| 183 |
-
{
|
| 184 |
-
"epoch": 0.72,
|
| 185 |
-
"grad_norm": 3.879312515258789,
|
| 186 |
-
"learning_rate": 9.928660436137072e-05,
|
| 187 |
-
"loss": 0.4047,
|
| 188 |
-
"step": 230
|
| 189 |
-
},
|
| 190 |
-
{
|
| 191 |
-
"epoch": 0.75,
|
| 192 |
-
"grad_norm": 2.546088457107544,
|
| 193 |
-
"learning_rate": 9.925545171339564e-05,
|
| 194 |
-
"loss": 0.559,
|
| 195 |
-
"step": 240
|
| 196 |
-
},
|
| 197 |
-
{
|
| 198 |
-
"epoch": 0.78,
|
| 199 |
-
"grad_norm": 6.213643550872803,
|
| 200 |
-
"learning_rate": 9.922429906542056e-05,
|
| 201 |
-
"loss": 0.7639,
|
| 202 |
-
"step": 250
|
| 203 |
-
},
|
| 204 |
-
{
|
| 205 |
-
"epoch": 0.81,
|
| 206 |
-
"grad_norm": 3.789020299911499,
|
| 207 |
-
"learning_rate": 9.91931464174455e-05,
|
| 208 |
-
"loss": 0.5372,
|
| 209 |
-
"step": 260
|
| 210 |
-
},
|
| 211 |
-
{
|
| 212 |
-
"epoch": 0.84,
|
| 213 |
-
"grad_norm": 5.764331340789795,
|
| 214 |
-
"learning_rate": 9.916199376947041e-05,
|
| 215 |
-
"loss": 0.5031,
|
| 216 |
-
"step": 270
|
| 217 |
-
},
|
| 218 |
-
{
|
| 219 |
-
"epoch": 0.87,
|
| 220 |
-
"grad_norm": 2.2293648719787598,
|
| 221 |
-
"learning_rate": 9.913084112149534e-05,
|
| 222 |
-
"loss": 0.4379,
|
| 223 |
-
"step": 280
|
| 224 |
-
},
|
| 225 |
-
{
|
| 226 |
-
"epoch": 0.9,
|
| 227 |
-
"grad_norm": 6.797378063201904,
|
| 228 |
-
"learning_rate": 9.909968847352025e-05,
|
| 229 |
-
"loss": 0.6147,
|
| 230 |
-
"step": 290
|
| 231 |
-
},
|
| 232 |
{
|
| 233 |
"epoch": 0.93,
|
| 234 |
-
"grad_norm": 6.
|
| 235 |
-
"learning_rate": 9.
|
| 236 |
-
"loss": 0.
|
| 237 |
"step": 300
|
| 238 |
},
|
| 239 |
{
|
| 240 |
"epoch": 0.93,
|
| 241 |
-
"eval_accuracy": 0.
|
| 242 |
-
"eval_loss": 0.
|
| 243 |
-
"eval_runtime":
|
| 244 |
-
"eval_samples_per_second":
|
| 245 |
-
"eval_steps_per_second": 9.
|
| 246 |
"step": 300
|
| 247 |
},
|
| 248 |
-
{
|
| 249 |
-
"epoch": 0.97,
|
| 250 |
-
"grad_norm": 3.367398262023926,
|
| 251 |
-
"learning_rate": 9.90373831775701e-05,
|
| 252 |
-
"loss": 0.5153,
|
| 253 |
-
"step": 310
|
| 254 |
-
},
|
| 255 |
-
{
|
| 256 |
-
"epoch": 1.0,
|
| 257 |
-
"grad_norm": 5.517517566680908,
|
| 258 |
-
"learning_rate": 9.900623052959503e-05,
|
| 259 |
-
"loss": 0.5022,
|
| 260 |
-
"step": 320
|
| 261 |
-
},
|
| 262 |
-
{
|
| 263 |
-
"epoch": 1.03,
|
| 264 |
-
"grad_norm": 6.2498393058776855,
|
| 265 |
-
"learning_rate": 9.897507788161994e-05,
|
| 266 |
-
"loss": 0.4581,
|
| 267 |
-
"step": 330
|
| 268 |
-
},
|
| 269 |
-
{
|
| 270 |
-
"epoch": 1.06,
|
| 271 |
-
"grad_norm": 5.515881538391113,
|
| 272 |
-
"learning_rate": 9.894392523364486e-05,
|
| 273 |
-
"loss": 0.3453,
|
| 274 |
-
"step": 340
|
| 275 |
-
},
|
| 276 |
-
{
|
| 277 |
-
"epoch": 1.09,
|
| 278 |
-
"grad_norm": 3.255171060562134,
|
| 279 |
-
"learning_rate": 9.891277258566979e-05,
|
| 280 |
-
"loss": 0.4332,
|
| 281 |
-
"step": 350
|
| 282 |
-
},
|
| 283 |
-
{
|
| 284 |
-
"epoch": 1.12,
|
| 285 |
-
"grad_norm": 2.262460231781006,
|
| 286 |
-
"learning_rate": 9.888161993769472e-05,
|
| 287 |
-
"loss": 0.3463,
|
| 288 |
-
"step": 360
|
| 289 |
-
},
|
| 290 |
-
{
|
| 291 |
-
"epoch": 1.15,
|
| 292 |
-
"grad_norm": 5.0724711418151855,
|
| 293 |
-
"learning_rate": 9.885046728971963e-05,
|
| 294 |
-
"loss": 0.5136,
|
| 295 |
-
"step": 370
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"epoch": 1.18,
|
| 299 |
-
"grad_norm": 3.8894565105438232,
|
| 300 |
-
"learning_rate": 9.881931464174455e-05,
|
| 301 |
-
"loss": 0.3792,
|
| 302 |
-
"step": 380
|
| 303 |
-
},
|
| 304 |
-
{
|
| 305 |
-
"epoch": 1.21,
|
| 306 |
-
"grad_norm": 4.299795627593994,
|
| 307 |
-
"learning_rate": 9.878816199376948e-05,
|
| 308 |
-
"loss": 0.4804,
|
| 309 |
-
"step": 390
|
| 310 |
-
},
|
| 311 |
{
|
| 312 |
"epoch": 1.25,
|
| 313 |
-
"grad_norm":
|
| 314 |
-
"learning_rate": 9.
|
| 315 |
-
"loss": 0.
|
| 316 |
"step": 400
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 1.25,
|
| 320 |
-
"eval_accuracy": 0.
|
| 321 |
-
"eval_loss": 0.
|
| 322 |
-
"eval_runtime": 39.
|
| 323 |
-
"eval_samples_per_second":
|
| 324 |
-
"eval_steps_per_second": 9.
|
| 325 |
"step": 400
|
| 326 |
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 1.28,
|
| 329 |
-
"grad_norm": 4.152671813964844,
|
| 330 |
-
"learning_rate": 9.872585669781932e-05,
|
| 331 |
-
"loss": 0.4629,
|
| 332 |
-
"step": 410
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 1.31,
|
| 336 |
-
"grad_norm": 3.4808621406555176,
|
| 337 |
-
"learning_rate": 9.869470404984425e-05,
|
| 338 |
-
"loss": 0.3583,
|
| 339 |
-
"step": 420
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 1.34,
|
| 343 |
-
"grad_norm": 1.483371615409851,
|
| 344 |
-
"learning_rate": 9.866355140186917e-05,
|
| 345 |
-
"loss": 0.5018,
|
| 346 |
-
"step": 430
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 1.37,
|
| 350 |
-
"grad_norm": 4.952483654022217,
|
| 351 |
-
"learning_rate": 9.863239875389408e-05,
|
| 352 |
-
"loss": 0.4338,
|
| 353 |
-
"step": 440
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 1.4,
|
| 357 |
-
"grad_norm": 4.527029037475586,
|
| 358 |
-
"learning_rate": 9.860124610591901e-05,
|
| 359 |
-
"loss": 0.437,
|
| 360 |
-
"step": 450
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 1.43,
|
| 364 |
-
"grad_norm": 7.609699726104736,
|
| 365 |
-
"learning_rate": 9.857009345794394e-05,
|
| 366 |
-
"loss": 0.3796,
|
| 367 |
-
"step": 460
|
| 368 |
-
},
|
| 369 |
-
{
|
| 370 |
-
"epoch": 1.46,
|
| 371 |
-
"grad_norm": 3.572444200515747,
|
| 372 |
-
"learning_rate": 9.853894080996885e-05,
|
| 373 |
-
"loss": 0.3551,
|
| 374 |
-
"step": 470
|
| 375 |
-
},
|
| 376 |
-
{
|
| 377 |
-
"epoch": 1.5,
|
| 378 |
-
"grad_norm": 8.147821426391602,
|
| 379 |
-
"learning_rate": 9.850778816199377e-05,
|
| 380 |
-
"loss": 0.4182,
|
| 381 |
-
"step": 480
|
| 382 |
-
},
|
| 383 |
-
{
|
| 384 |
-
"epoch": 1.53,
|
| 385 |
-
"grad_norm": 6.7267279624938965,
|
| 386 |
-
"learning_rate": 9.84766355140187e-05,
|
| 387 |
-
"loss": 0.3584,
|
| 388 |
-
"step": 490
|
| 389 |
-
},
|
| 390 |
{
|
| 391 |
"epoch": 1.56,
|
| 392 |
-
"grad_norm":
|
| 393 |
-
"learning_rate": 9.
|
| 394 |
-
"loss": 0.
|
| 395 |
"step": 500
|
| 396 |
},
|
| 397 |
{
|
| 398 |
"epoch": 1.56,
|
| 399 |
-
"eval_accuracy": 0.
|
| 400 |
-
"eval_loss": 0.
|
| 401 |
-
"eval_runtime": 39.
|
| 402 |
-
"eval_samples_per_second":
|
| 403 |
-
"eval_steps_per_second": 9.
|
| 404 |
"step": 500
|
| 405 |
},
|
| 406 |
-
{
|
| 407 |
-
"epoch": 1.59,
|
| 408 |
-
"grad_norm": 4.186713695526123,
|
| 409 |
-
"learning_rate": 9.841433021806854e-05,
|
| 410 |
-
"loss": 0.4802,
|
| 411 |
-
"step": 510
|
| 412 |
-
},
|
| 413 |
-
{
|
| 414 |
-
"epoch": 1.62,
|
| 415 |
-
"grad_norm": 4.469476699829102,
|
| 416 |
-
"learning_rate": 9.838317757009346e-05,
|
| 417 |
-
"loss": 0.4118,
|
| 418 |
-
"step": 520
|
| 419 |
-
},
|
| 420 |
-
{
|
| 421 |
-
"epoch": 1.65,
|
| 422 |
-
"grad_norm": 7.559543609619141,
|
| 423 |
-
"learning_rate": 9.835202492211837e-05,
|
| 424 |
-
"loss": 0.4605,
|
| 425 |
-
"step": 530
|
| 426 |
-
},
|
| 427 |
-
{
|
| 428 |
-
"epoch": 1.68,
|
| 429 |
-
"grad_norm": 2.5254476070404053,
|
| 430 |
-
"learning_rate": 9.83208722741433e-05,
|
| 431 |
-
"loss": 0.2878,
|
| 432 |
-
"step": 540
|
| 433 |
-
},
|
| 434 |
-
{
|
| 435 |
-
"epoch": 1.71,
|
| 436 |
-
"grad_norm": 4.755306720733643,
|
| 437 |
-
"learning_rate": 9.828971962616823e-05,
|
| 438 |
-
"loss": 0.3881,
|
| 439 |
-
"step": 550
|
| 440 |
-
},
|
| 441 |
-
{
|
| 442 |
-
"epoch": 1.74,
|
| 443 |
-
"grad_norm": 3.8046302795410156,
|
| 444 |
-
"learning_rate": 9.825856697819316e-05,
|
| 445 |
-
"loss": 0.3511,
|
| 446 |
-
"step": 560
|
| 447 |
-
},
|
| 448 |
-
{
|
| 449 |
-
"epoch": 1.78,
|
| 450 |
-
"grad_norm": 2.603055238723755,
|
| 451 |
-
"learning_rate": 9.822741433021808e-05,
|
| 452 |
-
"loss": 0.3354,
|
| 453 |
-
"step": 570
|
| 454 |
-
},
|
| 455 |
-
{
|
| 456 |
-
"epoch": 1.81,
|
| 457 |
-
"grad_norm": 3.5978047847747803,
|
| 458 |
-
"learning_rate": 9.819626168224299e-05,
|
| 459 |
-
"loss": 0.2885,
|
| 460 |
-
"step": 580
|
| 461 |
-
},
|
| 462 |
-
{
|
| 463 |
-
"epoch": 1.84,
|
| 464 |
-
"grad_norm": 5.625058174133301,
|
| 465 |
-
"learning_rate": 9.816510903426792e-05,
|
| 466 |
-
"loss": 0.392,
|
| 467 |
-
"step": 590
|
| 468 |
-
},
|
| 469 |
{
|
| 470 |
"epoch": 1.87,
|
| 471 |
-
"grad_norm": 4.
|
| 472 |
-
"learning_rate": 9.
|
| 473 |
-
"loss": 0.
|
| 474 |
"step": 600
|
| 475 |
},
|
| 476 |
{
|
| 477 |
"epoch": 1.87,
|
| 478 |
"eval_accuracy": 0.794382801664355,
|
| 479 |
-
"eval_loss": 0.
|
| 480 |
-
"eval_runtime": 39.
|
| 481 |
-
"eval_samples_per_second":
|
| 482 |
-
"eval_steps_per_second": 9.
|
| 483 |
"step": 600
|
| 484 |
},
|
| 485 |
-
{
|
| 486 |
-
"epoch": 1.9,
|
| 487 |
-
"grad_norm": 6.1720123291015625,
|
| 488 |
-
"learning_rate": 9.810280373831777e-05,
|
| 489 |
-
"loss": 0.5169,
|
| 490 |
-
"step": 610
|
| 491 |
-
},
|
| 492 |
-
{
|
| 493 |
-
"epoch": 1.93,
|
| 494 |
-
"grad_norm": 3.5033581256866455,
|
| 495 |
-
"learning_rate": 9.807165109034268e-05,
|
| 496 |
-
"loss": 0.4374,
|
| 497 |
-
"step": 620
|
| 498 |
-
},
|
| 499 |
-
{
|
| 500 |
-
"epoch": 1.96,
|
| 501 |
-
"grad_norm": 4.63729190826416,
|
| 502 |
-
"learning_rate": 9.80404984423676e-05,
|
| 503 |
-
"loss": 0.2866,
|
| 504 |
-
"step": 630
|
| 505 |
-
},
|
| 506 |
-
{
|
| 507 |
-
"epoch": 1.99,
|
| 508 |
-
"grad_norm": 3.4656624794006348,
|
| 509 |
-
"learning_rate": 9.800934579439253e-05,
|
| 510 |
-
"loss": 0.3323,
|
| 511 |
-
"step": 640
|
| 512 |
-
},
|
| 513 |
-
{
|
| 514 |
-
"epoch": 2.02,
|
| 515 |
-
"grad_norm": 3.2712185382843018,
|
| 516 |
-
"learning_rate": 9.797819314641746e-05,
|
| 517 |
-
"loss": 0.2275,
|
| 518 |
-
"step": 650
|
| 519 |
-
},
|
| 520 |
-
{
|
| 521 |
-
"epoch": 2.06,
|
| 522 |
-
"grad_norm": 3.9931576251983643,
|
| 523 |
-
"learning_rate": 9.794704049844237e-05,
|
| 524 |
-
"loss": 0.21,
|
| 525 |
-
"step": 660
|
| 526 |
-
},
|
| 527 |
-
{
|
| 528 |
-
"epoch": 2.09,
|
| 529 |
-
"grad_norm": 1.7855778932571411,
|
| 530 |
-
"learning_rate": 9.791588785046729e-05,
|
| 531 |
-
"loss": 0.2996,
|
| 532 |
-
"step": 670
|
| 533 |
-
},
|
| 534 |
-
{
|
| 535 |
-
"epoch": 2.12,
|
| 536 |
-
"grad_norm": 5.909547328948975,
|
| 537 |
-
"learning_rate": 9.788473520249222e-05,
|
| 538 |
-
"loss": 0.2081,
|
| 539 |
-
"step": 680
|
| 540 |
-
},
|
| 541 |
-
{
|
| 542 |
-
"epoch": 2.15,
|
| 543 |
-
"grad_norm": 1.9149627685546875,
|
| 544 |
-
"learning_rate": 9.785358255451714e-05,
|
| 545 |
-
"loss": 0.216,
|
| 546 |
-
"step": 690
|
| 547 |
-
},
|
| 548 |
{
|
| 549 |
"epoch": 2.18,
|
| 550 |
-
"grad_norm":
|
| 551 |
-
"learning_rate": 9.
|
| 552 |
-
"loss": 0.
|
| 553 |
"step": 700
|
| 554 |
},
|
| 555 |
{
|
| 556 |
"epoch": 2.18,
|
| 557 |
-
"eval_accuracy": 0.
|
| 558 |
-
"eval_loss": 0.
|
| 559 |
-
"eval_runtime": 39.
|
| 560 |
-
"eval_samples_per_second": 73.
|
| 561 |
-
"eval_steps_per_second": 9.
|
| 562 |
"step": 700
|
| 563 |
},
|
| 564 |
{
|
| 565 |
-
"epoch": 2.
|
| 566 |
-
"
|
| 567 |
-
"
|
| 568 |
-
"
|
| 569 |
-
"
|
| 570 |
-
|
| 571 |
-
|
| 572 |
-
"epoch": 2.24,
|
| 573 |
-
"grad_norm": 6.487290382385254,
|
| 574 |
-
"learning_rate": 9.77601246105919e-05,
|
| 575 |
-
"loss": 0.3093,
|
| 576 |
-
"step": 720
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 2.27,
|
| 580 |
-
"grad_norm": 6.3305230140686035,
|
| 581 |
-
"learning_rate": 9.772897196261682e-05,
|
| 582 |
-
"loss": 0.3298,
|
| 583 |
-
"step": 730
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 2.31,
|
| 587 |
-
"grad_norm": 1.8739694356918335,
|
| 588 |
-
"learning_rate": 9.769781931464175e-05,
|
| 589 |
-
"loss": 0.2836,
|
| 590 |
-
"step": 740
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 2.34,
|
| 594 |
-
"grad_norm": 6.215338230133057,
|
| 595 |
-
"learning_rate": 9.766666666666668e-05,
|
| 596 |
-
"loss": 0.213,
|
| 597 |
-
"step": 750
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 2.37,
|
| 601 |
-
"grad_norm": 0.6346792578697205,
|
| 602 |
-
"learning_rate": 9.76355140186916e-05,
|
| 603 |
-
"loss": 0.1537,
|
| 604 |
-
"step": 760
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 2.4,
|
| 608 |
-
"grad_norm": 6.488032341003418,
|
| 609 |
-
"learning_rate": 9.760436137071651e-05,
|
| 610 |
-
"loss": 0.2713,
|
| 611 |
-
"step": 770
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 2.43,
|
| 615 |
-
"grad_norm": 3.9325110912323,
|
| 616 |
-
"learning_rate": 9.757320872274144e-05,
|
| 617 |
-
"loss": 0.2861,
|
| 618 |
-
"step": 780
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 2.46,
|
| 622 |
-
"grad_norm": 4.505757808685303,
|
| 623 |
-
"learning_rate": 9.754205607476637e-05,
|
| 624 |
-
"loss": 0.2625,
|
| 625 |
-
"step": 790
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 2.49,
|
| 629 |
-
"grad_norm": 1.1799720525741577,
|
| 630 |
-
"learning_rate": 9.751090342679128e-05,
|
| 631 |
-
"loss": 0.2159,
|
| 632 |
-
"step": 800
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 2.49,
|
| 636 |
-
"eval_accuracy": 0.8359916782246879,
|
| 637 |
-
"eval_loss": 0.5187094211578369,
|
| 638 |
-
"eval_runtime": 38.9237,
|
| 639 |
-
"eval_samples_per_second": 74.094,
|
| 640 |
-
"eval_steps_per_second": 9.275,
|
| 641 |
-
"step": 800
|
| 642 |
-
},
|
| 643 |
-
{
|
| 644 |
-
"epoch": 2.52,
|
| 645 |
-
"grad_norm": 0.662136971950531,
|
| 646 |
-
"learning_rate": 9.74797507788162e-05,
|
| 647 |
-
"loss": 0.2325,
|
| 648 |
-
"step": 810
|
| 649 |
-
},
|
| 650 |
-
{
|
| 651 |
-
"epoch": 2.55,
|
| 652 |
-
"grad_norm": 9.680724143981934,
|
| 653 |
-
"learning_rate": 9.744859813084113e-05,
|
| 654 |
-
"loss": 0.2579,
|
| 655 |
-
"step": 820
|
| 656 |
-
},
|
| 657 |
-
{
|
| 658 |
-
"epoch": 2.59,
|
| 659 |
-
"grad_norm": 0.5741639137268066,
|
| 660 |
-
"learning_rate": 9.741744548286604e-05,
|
| 661 |
-
"loss": 0.1271,
|
| 662 |
-
"step": 830
|
| 663 |
-
},
|
| 664 |
-
{
|
| 665 |
-
"epoch": 2.62,
|
| 666 |
-
"grad_norm": 9.787598609924316,
|
| 667 |
-
"learning_rate": 9.738629283489097e-05,
|
| 668 |
-
"loss": 0.2705,
|
| 669 |
-
"step": 840
|
| 670 |
-
},
|
| 671 |
-
{
|
| 672 |
-
"epoch": 2.65,
|
| 673 |
-
"grad_norm": 3.5786948204040527,
|
| 674 |
-
"learning_rate": 9.73551401869159e-05,
|
| 675 |
-
"loss": 0.1892,
|
| 676 |
-
"step": 850
|
| 677 |
-
},
|
| 678 |
-
{
|
| 679 |
-
"epoch": 2.68,
|
| 680 |
-
"grad_norm": 3.534965753555298,
|
| 681 |
-
"learning_rate": 9.732398753894082e-05,
|
| 682 |
-
"loss": 0.1629,
|
| 683 |
-
"step": 860
|
| 684 |
-
},
|
| 685 |
-
{
|
| 686 |
-
"epoch": 2.71,
|
| 687 |
-
"grad_norm": 3.6333532333374023,
|
| 688 |
-
"learning_rate": 9.729283489096573e-05,
|
| 689 |
-
"loss": 0.3307,
|
| 690 |
-
"step": 870
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 2.74,
|
| 694 |
-
"grad_norm": 7.089163780212402,
|
| 695 |
-
"learning_rate": 9.726168224299066e-05,
|
| 696 |
-
"loss": 0.3159,
|
| 697 |
-
"step": 880
|
| 698 |
-
},
|
| 699 |
-
{
|
| 700 |
-
"epoch": 2.77,
|
| 701 |
-
"grad_norm": 0.845410168170929,
|
| 702 |
-
"learning_rate": 9.723052959501559e-05,
|
| 703 |
-
"loss": 0.1572,
|
| 704 |
-
"step": 890
|
| 705 |
-
},
|
| 706 |
-
{
|
| 707 |
-
"epoch": 2.8,
|
| 708 |
-
"grad_norm": 6.727290630340576,
|
| 709 |
-
"learning_rate": 9.71993769470405e-05,
|
| 710 |
-
"loss": 0.1659,
|
| 711 |
-
"step": 900
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 2.8,
|
| 715 |
-
"eval_accuracy": 0.8318307905686546,
|
| 716 |
-
"eval_loss": 0.5848782062530518,
|
| 717 |
-
"eval_runtime": 38.908,
|
| 718 |
-
"eval_samples_per_second": 74.124,
|
| 719 |
-
"eval_steps_per_second": 9.278,
|
| 720 |
-
"step": 900
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 2.83,
|
| 724 |
-
"grad_norm": 4.609592914581299,
|
| 725 |
-
"learning_rate": 9.716822429906542e-05,
|
| 726 |
-
"loss": 0.3052,
|
| 727 |
-
"step": 910
|
| 728 |
-
},
|
| 729 |
-
{
|
| 730 |
-
"epoch": 2.87,
|
| 731 |
-
"grad_norm": 4.7995285987854,
|
| 732 |
-
"learning_rate": 9.713707165109035e-05,
|
| 733 |
-
"loss": 0.3944,
|
| 734 |
-
"step": 920
|
| 735 |
-
},
|
| 736 |
-
{
|
| 737 |
-
"epoch": 2.9,
|
| 738 |
-
"grad_norm": 4.78571891784668,
|
| 739 |
-
"learning_rate": 9.710591900311527e-05,
|
| 740 |
-
"loss": 0.3118,
|
| 741 |
-
"step": 930
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 2.93,
|
| 745 |
-
"grad_norm": 7.084194183349609,
|
| 746 |
-
"learning_rate": 9.70747663551402e-05,
|
| 747 |
-
"loss": 0.2874,
|
| 748 |
-
"step": 940
|
| 749 |
-
},
|
| 750 |
-
{
|
| 751 |
-
"epoch": 2.96,
|
| 752 |
-
"grad_norm": 2.6003012657165527,
|
| 753 |
-
"learning_rate": 9.704361370716511e-05,
|
| 754 |
-
"loss": 0.2024,
|
| 755 |
-
"step": 950
|
| 756 |
-
},
|
| 757 |
-
{
|
| 758 |
-
"epoch": 2.99,
|
| 759 |
-
"grad_norm": 3.7201614379882812,
|
| 760 |
-
"learning_rate": 9.701246105919004e-05,
|
| 761 |
-
"loss": 0.224,
|
| 762 |
-
"step": 960
|
| 763 |
-
},
|
| 764 |
-
{
|
| 765 |
-
"epoch": 3.02,
|
| 766 |
-
"grad_norm": 3.0651230812072754,
|
| 767 |
-
"learning_rate": 9.698130841121495e-05,
|
| 768 |
-
"loss": 0.1076,
|
| 769 |
-
"step": 970
|
| 770 |
-
},
|
| 771 |
-
{
|
| 772 |
-
"epoch": 3.05,
|
| 773 |
-
"grad_norm": 4.620060920715332,
|
| 774 |
-
"learning_rate": 9.695015576323988e-05,
|
| 775 |
-
"loss": 0.1312,
|
| 776 |
-
"step": 980
|
| 777 |
-
},
|
| 778 |
-
{
|
| 779 |
-
"epoch": 3.08,
|
| 780 |
-
"grad_norm": 0.5105383992195129,
|
| 781 |
-
"learning_rate": 9.691900311526481e-05,
|
| 782 |
-
"loss": 0.1479,
|
| 783 |
-
"step": 990
|
| 784 |
-
},
|
| 785 |
-
{
|
| 786 |
-
"epoch": 3.12,
|
| 787 |
-
"grad_norm": 1.0817281007766724,
|
| 788 |
-
"learning_rate": 9.688785046728971e-05,
|
| 789 |
-
"loss": 0.0554,
|
| 790 |
-
"step": 1000
|
| 791 |
-
},
|
| 792 |
-
{
|
| 793 |
-
"epoch": 3.12,
|
| 794 |
-
"eval_accuracy": 0.8543689320388349,
|
| 795 |
-
"eval_loss": 0.5457024574279785,
|
| 796 |
-
"eval_runtime": 39.7583,
|
| 797 |
-
"eval_samples_per_second": 72.538,
|
| 798 |
-
"eval_steps_per_second": 9.08,
|
| 799 |
-
"step": 1000
|
| 800 |
-
},
|
| 801 |
-
{
|
| 802 |
-
"epoch": 3.15,
|
| 803 |
-
"grad_norm": 1.351655125617981,
|
| 804 |
-
"learning_rate": 9.685669781931464e-05,
|
| 805 |
-
"loss": 0.0568,
|
| 806 |
-
"step": 1010
|
| 807 |
-
},
|
| 808 |
-
{
|
| 809 |
-
"epoch": 3.18,
|
| 810 |
-
"grad_norm": 3.9608170986175537,
|
| 811 |
-
"learning_rate": 9.682554517133957e-05,
|
| 812 |
-
"loss": 0.1162,
|
| 813 |
-
"step": 1020
|
| 814 |
-
},
|
| 815 |
-
{
|
| 816 |
-
"epoch": 3.21,
|
| 817 |
-
"grad_norm": 6.9148783683776855,
|
| 818 |
-
"learning_rate": 9.679439252336449e-05,
|
| 819 |
-
"loss": 0.0587,
|
| 820 |
-
"step": 1030
|
| 821 |
-
},
|
| 822 |
-
{
|
| 823 |
-
"epoch": 3.24,
|
| 824 |
-
"grad_norm": 1.0946722030639648,
|
| 825 |
-
"learning_rate": 9.676323987538942e-05,
|
| 826 |
-
"loss": 0.0572,
|
| 827 |
-
"step": 1040
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 3.27,
|
| 831 |
-
"grad_norm": 9.527204513549805,
|
| 832 |
-
"learning_rate": 9.673208722741433e-05,
|
| 833 |
-
"loss": 0.1847,
|
| 834 |
-
"step": 1050
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 3.3,
|
| 838 |
-
"grad_norm": 7.0584716796875,
|
| 839 |
-
"learning_rate": 9.670093457943926e-05,
|
| 840 |
-
"loss": 0.1296,
|
| 841 |
-
"step": 1060
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 3.33,
|
| 845 |
-
"grad_norm": 7.145915985107422,
|
| 846 |
-
"learning_rate": 9.666978193146418e-05,
|
| 847 |
-
"loss": 0.1188,
|
| 848 |
-
"step": 1070
|
| 849 |
-
},
|
| 850 |
-
{
|
| 851 |
-
"epoch": 3.36,
|
| 852 |
-
"grad_norm": 7.105827808380127,
|
| 853 |
-
"learning_rate": 9.66386292834891e-05,
|
| 854 |
-
"loss": 0.099,
|
| 855 |
-
"step": 1080
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 3.4,
|
| 859 |
-
"grad_norm": 8.103167533874512,
|
| 860 |
-
"learning_rate": 9.660747663551402e-05,
|
| 861 |
-
"loss": 0.0874,
|
| 862 |
-
"step": 1090
|
| 863 |
-
},
|
| 864 |
-
{
|
| 865 |
-
"epoch": 3.43,
|
| 866 |
-
"grad_norm": 1.4485965967178345,
|
| 867 |
-
"learning_rate": 9.657632398753894e-05,
|
| 868 |
-
"loss": 0.1125,
|
| 869 |
-
"step": 1100
|
| 870 |
-
},
|
| 871 |
-
{
|
| 872 |
-
"epoch": 3.43,
|
| 873 |
-
"eval_accuracy": 0.8318307905686546,
|
| 874 |
-
"eval_loss": 0.7323819398880005,
|
| 875 |
-
"eval_runtime": 39.6414,
|
| 876 |
-
"eval_samples_per_second": 72.752,
|
| 877 |
-
"eval_steps_per_second": 9.107,
|
| 878 |
-
"step": 1100
|
| 879 |
-
},
|
| 880 |
-
{
|
| 881 |
-
"epoch": 3.43,
|
| 882 |
-
"step": 1100,
|
| 883 |
-
"total_flos": 1.362060245306622e+18,
|
| 884 |
-
"train_loss": 0.38896639000285754,
|
| 885 |
-
"train_runtime": 855.3091,
|
| 886 |
-
"train_samples_per_second": 599.549,
|
| 887 |
-
"train_steps_per_second": 37.53
|
| 888 |
}
|
| 889 |
],
|
| 890 |
-
"logging_steps":
|
| 891 |
"max_steps": 32100,
|
| 892 |
"num_input_tokens_seen": 0,
|
| 893 |
"num_train_epochs": 100,
|
| 894 |
"save_steps": 100,
|
| 895 |
-
"total_flos":
|
| 896 |
"train_batch_size": 16,
|
| 897 |
"trial_name": null,
|
| 898 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.49744343757629395,
|
| 3 |
+
"best_model_checkpoint": "./vit-lr-0.0001/checkpoint-400",
|
| 4 |
+
"epoch": 2.1806853582554515,
|
| 5 |
"eval_steps": 100,
|
| 6 |
+
"global_step": 700,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.31,
|
| 13 |
+
"grad_norm": 4.504175186157227,
|
| 14 |
+
"learning_rate": 9.999760543724762e-05,
|
| 15 |
+
"loss": 0.7774,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.31,
|
| 20 |
+
"eval_accuracy": 0.7846740638002774,
|
| 21 |
+
"eval_loss": 0.5988100171089172,
|
| 22 |
+
"eval_runtime": 39.924,
|
| 23 |
+
"eval_samples_per_second": 72.237,
|
| 24 |
+
"eval_steps_per_second": 9.042,
|
| 25 |
"step": 100
|
| 26 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
{
|
| 28 |
"epoch": 0.62,
|
| 29 |
+
"grad_norm": 6.5739970207214355,
|
| 30 |
+
"learning_rate": 9.999042197834769e-05,
|
| 31 |
+
"loss": 0.59,
|
| 32 |
"step": 200
|
| 33 |
},
|
| 34 |
{
|
| 35 |
"epoch": 0.62,
|
| 36 |
+
"eval_accuracy": 0.7732316227461858,
|
| 37 |
+
"eval_loss": 0.6567931175231934,
|
| 38 |
+
"eval_runtime": 39.9411,
|
| 39 |
+
"eval_samples_per_second": 72.206,
|
| 40 |
+
"eval_steps_per_second": 9.038,
|
| 41 |
"step": 200
|
| 42 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
{
|
| 44 |
"epoch": 0.93,
|
| 45 |
+
"grad_norm": 6.0171613693237305,
|
| 46 |
+
"learning_rate": 9.997859372626506e-05,
|
| 47 |
+
"loss": 0.584,
|
| 48 |
"step": 300
|
| 49 |
},
|
| 50 |
{
|
| 51 |
"epoch": 0.93,
|
| 52 |
+
"eval_accuracy": 0.7562413314840499,
|
| 53 |
+
"eval_loss": 0.6686193346977234,
|
| 54 |
+
"eval_runtime": 40.0793,
|
| 55 |
+
"eval_samples_per_second": 71.957,
|
| 56 |
+
"eval_steps_per_second": 9.007,
|
| 57 |
"step": 300
|
| 58 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
{
|
| 60 |
"epoch": 1.25,
|
| 61 |
+
"grad_norm": 3.5295469760894775,
|
| 62 |
+
"learning_rate": 9.996188286127832e-05,
|
| 63 |
+
"loss": 0.4275,
|
| 64 |
"step": 400
|
| 65 |
},
|
| 66 |
{
|
| 67 |
"epoch": 1.25,
|
| 68 |
+
"eval_accuracy": 0.8169209431345353,
|
| 69 |
+
"eval_loss": 0.49744343757629395,
|
| 70 |
+
"eval_runtime": 39.8506,
|
| 71 |
+
"eval_samples_per_second": 72.37,
|
| 72 |
+
"eval_steps_per_second": 9.059,
|
| 73 |
"step": 400
|
| 74 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
{
|
| 76 |
"epoch": 1.56,
|
| 77 |
+
"grad_norm": 2.440728187561035,
|
| 78 |
+
"learning_rate": 9.994038652174203e-05,
|
| 79 |
+
"loss": 0.3963,
|
| 80 |
"step": 500
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"epoch": 1.56,
|
| 84 |
+
"eval_accuracy": 0.8287101248266296,
|
| 85 |
+
"eval_loss": 0.507750391960144,
|
| 86 |
+
"eval_runtime": 39.6231,
|
| 87 |
+
"eval_samples_per_second": 72.786,
|
| 88 |
+
"eval_steps_per_second": 9.111,
|
| 89 |
"step": 500
|
| 90 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 91 |
{
|
| 92 |
"epoch": 1.87,
|
| 93 |
+
"grad_norm": 4.625945568084717,
|
| 94 |
+
"learning_rate": 9.991410676662952e-05,
|
| 95 |
+
"loss": 0.3708,
|
| 96 |
"step": 600
|
| 97 |
},
|
| 98 |
{
|
| 99 |
"epoch": 1.87,
|
| 100 |
"eval_accuracy": 0.794382801664355,
|
| 101 |
+
"eval_loss": 0.5709377527236938,
|
| 102 |
+
"eval_runtime": 39.6196,
|
| 103 |
+
"eval_samples_per_second": 72.792,
|
| 104 |
+
"eval_steps_per_second": 9.112,
|
| 105 |
"step": 600
|
| 106 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 107 |
{
|
| 108 |
"epoch": 2.18,
|
| 109 |
+
"grad_norm": 2.9291675090789795,
|
| 110 |
+
"learning_rate": 9.988304611308174e-05,
|
| 111 |
+
"loss": 0.2628,
|
| 112 |
"step": 700
|
| 113 |
},
|
| 114 |
{
|
| 115 |
"epoch": 2.18,
|
| 116 |
+
"eval_accuracy": 0.812760055478502,
|
| 117 |
+
"eval_loss": 0.6022083163261414,
|
| 118 |
+
"eval_runtime": 39.3146,
|
| 119 |
+
"eval_samples_per_second": 73.357,
|
| 120 |
+
"eval_steps_per_second": 9.182,
|
| 121 |
"step": 700
|
| 122 |
},
|
| 123 |
{
|
| 124 |
+
"epoch": 2.18,
|
| 125 |
+
"step": 700,
|
| 126 |
+
"total_flos": 8.667092503134536e+17,
|
| 127 |
+
"train_loss": 0.4869789886474609,
|
| 128 |
+
"train_runtime": 547.2513,
|
| 129 |
+
"train_samples_per_second": 937.047,
|
| 130 |
+
"train_steps_per_second": 58.657
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 131 |
}
|
| 132 |
],
|
| 133 |
+
"logging_steps": 100,
|
| 134 |
"max_steps": 32100,
|
| 135 |
"num_input_tokens_seen": 0,
|
| 136 |
"num_train_epochs": 100,
|
| 137 |
"save_steps": 100,
|
| 138 |
+
"total_flos": 8.667092503134536e+17,
|
| 139 |
"train_batch_size": 16,
|
| 140 |
"trial_name": null,
|
| 141 |
"trial_params": null
|