End of training
Browse files- README.md +9 -7
- all_results.json +15 -15
- eval_results.json +10 -10
- train_results.json +6 -6
- trainer_state.json +356 -483
README.md
CHANGED
|
@@ -3,6 +3,8 @@ library_name: transformers
|
|
| 3 |
license: other
|
| 4 |
base_model: nvidia/mit-b0
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- generator
|
|
@@ -16,16 +18,16 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# autocrop-bilder
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Mean Iou: 0.
|
| 23 |
-
- Mean Accuracy: 0.
|
| 24 |
-
- Overall Accuracy: 0.
|
| 25 |
- Accuracy Background: nan
|
| 26 |
-
- Accuracy Crop: 0.
|
| 27 |
- Iou Background: 0.0
|
| 28 |
-
- Iou Crop: 0.
|
| 29 |
|
| 30 |
## Model description
|
| 31 |
|
|
|
|
| 3 |
license: other
|
| 4 |
base_model: nvidia/mit-b0
|
| 5 |
tags:
|
| 6 |
+
- image-segmentation
|
| 7 |
+
- vision
|
| 8 |
- generated_from_trainer
|
| 9 |
datasets:
|
| 10 |
- generator
|
|
|
|
| 18 |
|
| 19 |
# autocrop-bilder
|
| 20 |
|
| 21 |
+
This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the /mnt/disk1/autocrop-data/datasets/bilder/ dataset.
|
| 22 |
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 0.1087
|
| 24 |
+
- Mean Iou: 0.4908
|
| 25 |
+
- Mean Accuracy: 0.9816
|
| 26 |
+
- Overall Accuracy: 0.9816
|
| 27 |
- Accuracy Background: nan
|
| 28 |
+
- Accuracy Crop: 0.9816
|
| 29 |
- Iou Background: 0.0
|
| 30 |
+
- Iou Crop: 0.9816
|
| 31 |
|
| 32 |
## Model description
|
| 33 |
|
all_results.json
CHANGED
|
@@ -1,19 +1,19 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"eval_accuracy_background": NaN,
|
| 4 |
-
"eval_accuracy_crop": 0.
|
| 5 |
"eval_iou_background": 0.0,
|
| 6 |
-
"eval_iou_crop": 0.
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_mean_accuracy": 0.
|
| 9 |
-
"eval_mean_iou": 0.
|
| 10 |
-
"eval_overall_accuracy": 0.
|
| 11 |
-
"eval_runtime":
|
| 12 |
-
"eval_samples_per_second":
|
| 13 |
-
"eval_steps_per_second":
|
| 14 |
-
"total_flos": 2.
|
| 15 |
-
"train_loss": 0.
|
| 16 |
-
"train_runtime":
|
| 17 |
-
"train_samples_per_second":
|
| 18 |
-
"train_steps_per_second":
|
| 19 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 31.0,
|
| 3 |
"eval_accuracy_background": NaN,
|
| 4 |
+
"eval_accuracy_crop": 0.9816317254805985,
|
| 5 |
"eval_iou_background": 0.0,
|
| 6 |
+
"eval_iou_crop": 0.9816317254805985,
|
| 7 |
+
"eval_loss": 0.1087288111448288,
|
| 8 |
+
"eval_mean_accuracy": 0.9816317254805985,
|
| 9 |
+
"eval_mean_iou": 0.49081586274029926,
|
| 10 |
+
"eval_overall_accuracy": 0.9816317254805985,
|
| 11 |
+
"eval_runtime": 0.6022,
|
| 12 |
+
"eval_samples_per_second": 16.605,
|
| 13 |
+
"eval_steps_per_second": 3.321,
|
| 14 |
+
"total_flos": 2.7711697679548416e+16,
|
| 15 |
+
"train_loss": 0.13814654548047325,
|
| 16 |
+
"train_runtime": 56.09,
|
| 17 |
+
"train_samples_per_second": 45.463,
|
| 18 |
+
"train_steps_per_second": 6.24
|
| 19 |
}
|
eval_results.json
CHANGED
|
@@ -1,14 +1,14 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
"eval_accuracy_background": NaN,
|
| 4 |
-
"eval_accuracy_crop": 0.
|
| 5 |
"eval_iou_background": 0.0,
|
| 6 |
-
"eval_iou_crop": 0.
|
| 7 |
-
"eval_loss": 0.
|
| 8 |
-
"eval_mean_accuracy": 0.
|
| 9 |
-
"eval_mean_iou": 0.
|
| 10 |
-
"eval_overall_accuracy": 0.
|
| 11 |
-
"eval_runtime":
|
| 12 |
-
"eval_samples_per_second":
|
| 13 |
-
"eval_steps_per_second":
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 31.0,
|
| 3 |
"eval_accuracy_background": NaN,
|
| 4 |
+
"eval_accuracy_crop": 0.9816317254805985,
|
| 5 |
"eval_iou_background": 0.0,
|
| 6 |
+
"eval_iou_crop": 0.9816317254805985,
|
| 7 |
+
"eval_loss": 0.1087288111448288,
|
| 8 |
+
"eval_mean_accuracy": 0.9816317254805985,
|
| 9 |
+
"eval_mean_iou": 0.49081586274029926,
|
| 10 |
+
"eval_overall_accuracy": 0.9816317254805985,
|
| 11 |
+
"eval_runtime": 0.6022,
|
| 12 |
+
"eval_samples_per_second": 16.605,
|
| 13 |
+
"eval_steps_per_second": 3.321
|
| 14 |
}
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 2.
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 31.0,
|
| 3 |
+
"total_flos": 2.7711697679548416e+16,
|
| 4 |
+
"train_loss": 0.13814654548047325,
|
| 5 |
+
"train_runtime": 56.09,
|
| 6 |
+
"train_samples_per_second": 45.463,
|
| 7 |
+
"train_steps_per_second": 6.24
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,618 +1,491 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "trainer_output/checkpoint-
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.8,
|
| 14 |
-
"grad_norm": 1.1524319648742676,
|
| 15 |
-
"learning_rate": 9.5808e-06,
|
| 16 |
-
"loss": 0.43339532470703124,
|
| 17 |
-
"step": 500
|
| 18 |
-
},
|
| 19 |
{
|
| 20 |
"epoch": 1.0,
|
| 21 |
"eval_accuracy_background": NaN,
|
| 22 |
-
"eval_accuracy_crop": 0.
|
| 23 |
"eval_iou_background": 0.0,
|
| 24 |
-
"eval_iou_crop": 0.
|
| 25 |
-
"eval_loss": 0.
|
| 26 |
-
"eval_mean_accuracy": 0.
|
| 27 |
-
"eval_mean_iou": 0.
|
| 28 |
-
"eval_overall_accuracy": 0.
|
| 29 |
-
"eval_runtime":
|
| 30 |
-
"eval_samples_per_second":
|
| 31 |
-
"eval_steps_per_second":
|
| 32 |
-
"step":
|
| 33 |
-
},
|
| 34 |
-
{
|
| 35 |
-
"epoch": 1.6,
|
| 36 |
-
"grad_norm": 0.8539880514144897,
|
| 37 |
-
"learning_rate": 1.91808e-05,
|
| 38 |
-
"loss": 0.11356404113769532,
|
| 39 |
-
"step": 1000
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 2.0,
|
| 43 |
"eval_accuracy_background": NaN,
|
| 44 |
-
"eval_accuracy_crop": 0.
|
| 45 |
"eval_iou_background": 0.0,
|
| 46 |
-
"eval_iou_crop": 0.
|
| 47 |
-
"eval_loss": 0.
|
| 48 |
-
"eval_mean_accuracy": 0.
|
| 49 |
-
"eval_mean_iou": 0.
|
| 50 |
-
"eval_overall_accuracy": 0.
|
| 51 |
-
"eval_runtime":
|
| 52 |
-
"eval_samples_per_second":
|
| 53 |
-
"eval_steps_per_second":
|
| 54 |
-
"step":
|
| 55 |
-
},
|
| 56 |
-
{
|
| 57 |
-
"epoch": 2.4,
|
| 58 |
-
"grad_norm": 0.45036041736602783,
|
| 59 |
-
"learning_rate": 2.87808e-05,
|
| 60 |
-
"loss": 0.06263476181030274,
|
| 61 |
-
"step": 1500
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 3.0,
|
| 65 |
"eval_accuracy_background": NaN,
|
| 66 |
-
"eval_accuracy_crop": 0.
|
| 67 |
"eval_iou_background": 0.0,
|
| 68 |
-
"eval_iou_crop": 0.
|
| 69 |
-
"eval_loss": 0.
|
| 70 |
-
"eval_mean_accuracy": 0.
|
| 71 |
-
"eval_mean_iou": 0.
|
| 72 |
-
"eval_overall_accuracy": 0.
|
| 73 |
-
"eval_runtime":
|
| 74 |
-
"eval_samples_per_second":
|
| 75 |
-
"eval_steps_per_second":
|
| 76 |
-
"step":
|
| 77 |
-
},
|
| 78 |
-
{
|
| 79 |
-
"epoch": 3.2,
|
| 80 |
-
"grad_norm": 0.3090341091156006,
|
| 81 |
-
"learning_rate": 3.8380800000000004e-05,
|
| 82 |
-
"loss": 0.04462788009643555,
|
| 83 |
-
"step": 2000
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 4.0,
|
| 87 |
-
"grad_norm": 0.24643999338150024,
|
| 88 |
-
"learning_rate": 4.79808e-05,
|
| 89 |
-
"loss": 0.03756232452392578,
|
| 90 |
-
"step": 2500
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"epoch": 4.0,
|
| 94 |
"eval_accuracy_background": NaN,
|
| 95 |
-
"eval_accuracy_crop": 0.
|
| 96 |
"eval_iou_background": 0.0,
|
| 97 |
-
"eval_iou_crop": 0.
|
| 98 |
-
"eval_loss": 0.
|
| 99 |
-
"eval_mean_accuracy": 0.
|
| 100 |
-
"eval_mean_iou": 0.
|
| 101 |
-
"eval_overall_accuracy": 0.
|
| 102 |
-
"eval_runtime":
|
| 103 |
-
"eval_samples_per_second":
|
| 104 |
-
"eval_steps_per_second":
|
| 105 |
-
"step":
|
| 106 |
-
},
|
| 107 |
-
{
|
| 108 |
-
"epoch": 4.8,
|
| 109 |
-
"grad_norm": 0.5337187647819519,
|
| 110 |
-
"learning_rate": 5.75808e-05,
|
| 111 |
-
"loss": 0.03314236831665039,
|
| 112 |
-
"step": 3000
|
| 113 |
},
|
| 114 |
{
|
| 115 |
"epoch": 5.0,
|
| 116 |
"eval_accuracy_background": NaN,
|
| 117 |
-
"eval_accuracy_crop": 0.
|
| 118 |
"eval_iou_background": 0.0,
|
| 119 |
-
"eval_iou_crop": 0.
|
| 120 |
-
"eval_loss": 0.
|
| 121 |
-
"eval_mean_accuracy": 0.
|
| 122 |
-
"eval_mean_iou": 0.
|
| 123 |
-
"eval_overall_accuracy": 0.
|
| 124 |
-
"eval_runtime":
|
| 125 |
-
"eval_samples_per_second":
|
| 126 |
-
"eval_steps_per_second":
|
| 127 |
-
"step":
|
| 128 |
-
},
|
| 129 |
-
{
|
| 130 |
-
"epoch": 5.6,
|
| 131 |
-
"grad_norm": 0.49186381697654724,
|
| 132 |
-
"learning_rate": 5.9973825042635214e-05,
|
| 133 |
-
"loss": 0.031177324295043947,
|
| 134 |
-
"step": 3500
|
| 135 |
},
|
| 136 |
{
|
| 137 |
"epoch": 6.0,
|
| 138 |
"eval_accuracy_background": NaN,
|
| 139 |
-
"eval_accuracy_crop": 0.
|
| 140 |
"eval_iou_background": 0.0,
|
| 141 |
-
"eval_iou_crop": 0.
|
| 142 |
-
"eval_loss": 0.
|
| 143 |
-
"eval_mean_accuracy": 0.
|
| 144 |
-
"eval_mean_iou": 0.
|
| 145 |
-
"eval_overall_accuracy": 0.
|
| 146 |
-
"eval_runtime":
|
| 147 |
-
"eval_samples_per_second":
|
| 148 |
-
"eval_steps_per_second":
|
| 149 |
-
"step":
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 6.4,
|
| 153 |
-
"grad_norm": 0.26261016726493835,
|
| 154 |
-
"learning_rate": 5.985714881662627e-05,
|
| 155 |
-
"loss": 0.028994924545288087,
|
| 156 |
-
"step": 4000
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 7.0,
|
| 160 |
"eval_accuracy_background": NaN,
|
| 161 |
-
"eval_accuracy_crop": 0.
|
| 162 |
"eval_iou_background": 0.0,
|
| 163 |
-
"eval_iou_crop": 0.
|
| 164 |
-
"eval_loss": 0.
|
| 165 |
-
"eval_mean_accuracy": 0.
|
| 166 |
-
"eval_mean_iou": 0.
|
| 167 |
-
"eval_overall_accuracy": 0.
|
| 168 |
-
"eval_runtime":
|
| 169 |
-
"eval_samples_per_second":
|
| 170 |
-
"eval_steps_per_second":
|
| 171 |
-
"step":
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 7.2,
|
| 175 |
-
"grad_norm": 0.3545992374420166,
|
| 176 |
-
"learning_rate": 5.9647363918899715e-05,
|
| 177 |
-
"loss": 0.028564287185668946,
|
| 178 |
-
"step": 4500
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 8.0,
|
| 182 |
-
"grad_norm": 0.3758007884025574,
|
| 183 |
-
"learning_rate": 5.9345124557708774e-05,
|
| 184 |
-
"loss": 0.027177539825439453,
|
| 185 |
-
"step": 5000
|
| 186 |
},
|
| 187 |
{
|
| 188 |
"epoch": 8.0,
|
| 189 |
"eval_accuracy_background": NaN,
|
| 190 |
-
"eval_accuracy_crop": 0.
|
| 191 |
"eval_iou_background": 0.0,
|
| 192 |
-
"eval_iou_crop": 0.
|
| 193 |
-
"eval_loss": 0.
|
| 194 |
-
"eval_mean_accuracy": 0.
|
| 195 |
-
"eval_mean_iou": 0.
|
| 196 |
-
"eval_overall_accuracy": 0.
|
| 197 |
-
"eval_runtime":
|
| 198 |
-
"eval_samples_per_second":
|
| 199 |
-
"eval_steps_per_second":
|
| 200 |
-
"step":
|
| 201 |
-
},
|
| 202 |
-
{
|
| 203 |
-
"epoch": 8.8,
|
| 204 |
-
"grad_norm": 0.22928953170776367,
|
| 205 |
-
"learning_rate": 5.895137325792861e-05,
|
| 206 |
-
"loss": 0.026362817764282228,
|
| 207 |
-
"step": 5500
|
| 208 |
},
|
| 209 |
{
|
| 210 |
"epoch": 9.0,
|
| 211 |
"eval_accuracy_background": NaN,
|
| 212 |
-
"eval_accuracy_crop": 0.
|
| 213 |
"eval_iou_background": 0.0,
|
| 214 |
-
"eval_iou_crop": 0.
|
| 215 |
-
"eval_loss": 0.
|
| 216 |
-
"eval_mean_accuracy": 0.
|
| 217 |
-
"eval_mean_iou": 0.
|
| 218 |
-
"eval_overall_accuracy": 0.
|
| 219 |
-
"eval_runtime":
|
| 220 |
-
"eval_samples_per_second":
|
| 221 |
-
"eval_steps_per_second":
|
| 222 |
-
"step":
|
| 223 |
-
},
|
| 224 |
-
{
|
| 225 |
-
"epoch": 9.6,
|
| 226 |
-
"grad_norm": 0.3915350139141083,
|
| 227 |
-
"learning_rate": 5.846733792181926e-05,
|
| 228 |
-
"loss": 0.02600261688232422,
|
| 229 |
-
"step": 6000
|
| 230 |
},
|
| 231 |
{
|
| 232 |
"epoch": 10.0,
|
| 233 |
"eval_accuracy_background": NaN,
|
| 234 |
-
"eval_accuracy_crop": 0.
|
| 235 |
"eval_iou_background": 0.0,
|
| 236 |
-
"eval_iou_crop": 0.
|
| 237 |
-
"eval_loss": 0.
|
| 238 |
-
"eval_mean_accuracy": 0.
|
| 239 |
-
"eval_mean_iou": 0.
|
| 240 |
-
"eval_overall_accuracy": 0.
|
| 241 |
-
"eval_runtime":
|
| 242 |
-
"eval_samples_per_second":
|
| 243 |
-
"eval_steps_per_second":
|
| 244 |
-
"step":
|
| 245 |
-
},
|
| 246 |
-
{
|
| 247 |
-
"epoch": 10.4,
|
| 248 |
-
"grad_norm": 0.26642245054244995,
|
| 249 |
-
"learning_rate": 5.78945279998472e-05,
|
| 250 |
-
"loss": 0.025672937393188478,
|
| 251 |
-
"step": 6500
|
| 252 |
},
|
| 253 |
{
|
| 254 |
"epoch": 11.0,
|
| 255 |
"eval_accuracy_background": NaN,
|
| 256 |
-
"eval_accuracy_crop": 0.
|
| 257 |
"eval_iou_background": 0.0,
|
| 258 |
-
"eval_iou_crop": 0.
|
| 259 |
-
"eval_loss": 0.
|
| 260 |
-
"eval_mean_accuracy": 0.
|
| 261 |
-
"eval_mean_iou": 0.
|
| 262 |
-
"eval_overall_accuracy": 0.
|
| 263 |
-
"eval_runtime":
|
| 264 |
-
"eval_samples_per_second":
|
| 265 |
-
"eval_steps_per_second":
|
| 266 |
-
"step":
|
| 267 |
-
},
|
| 268 |
-
{
|
| 269 |
-
"epoch": 11.2,
|
| 270 |
-
"grad_norm": 0.5744491815567017,
|
| 271 |
-
"learning_rate": 5.723472978350694e-05,
|
| 272 |
-
"loss": 0.024709026336669923,
|
| 273 |
-
"step": 7000
|
| 274 |
-
},
|
| 275 |
-
{
|
| 276 |
-
"epoch": 12.0,
|
| 277 |
-
"grad_norm": 0.16804039478302002,
|
| 278 |
-
"learning_rate": 5.649000083482167e-05,
|
| 279 |
-
"loss": 0.024350887298583986,
|
| 280 |
-
"step": 7500
|
| 281 |
},
|
| 282 |
{
|
| 283 |
"epoch": 12.0,
|
| 284 |
"eval_accuracy_background": NaN,
|
| 285 |
-
"eval_accuracy_crop": 0.
|
| 286 |
"eval_iou_background": 0.0,
|
| 287 |
-
"eval_iou_crop": 0.
|
| 288 |
-
"eval_loss": 0.
|
| 289 |
-
"eval_mean_accuracy": 0.
|
| 290 |
-
"eval_mean_iou": 0.
|
| 291 |
-
"eval_overall_accuracy": 0.
|
| 292 |
-
"eval_runtime":
|
| 293 |
-
"eval_samples_per_second":
|
| 294 |
-
"eval_steps_per_second":
|
| 295 |
-
"step":
|
| 296 |
-
},
|
| 297 |
-
{
|
| 298 |
-
"epoch": 12.8,
|
| 299 |
-
"grad_norm": 0.2182096242904663,
|
| 300 |
-
"learning_rate": 5.566266356989437e-05,
|
| 301 |
-
"loss": 0.02396395683288574,
|
| 302 |
-
"step": 8000
|
| 303 |
},
|
| 304 |
{
|
| 305 |
"epoch": 13.0,
|
| 306 |
"eval_accuracy_background": NaN,
|
| 307 |
-
"eval_accuracy_crop": 0.
|
| 308 |
"eval_iou_background": 0.0,
|
| 309 |
-
"eval_iou_crop": 0.
|
| 310 |
-
"eval_loss": 0.
|
| 311 |
-
"eval_mean_accuracy": 0.
|
| 312 |
-
"eval_mean_iou": 0.
|
| 313 |
-
"eval_overall_accuracy": 0.
|
| 314 |
-
"eval_runtime":
|
| 315 |
-
"eval_samples_per_second":
|
| 316 |
-
"eval_steps_per_second":
|
| 317 |
-
"step":
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 13.6,
|
| 321 |
-
"grad_norm": 0.2837521731853485,
|
| 322 |
-
"learning_rate": 5.4755298016519184e-05,
|
| 323 |
-
"loss": 0.023011560440063476,
|
| 324 |
-
"step": 8500
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 14.0,
|
| 328 |
"eval_accuracy_background": NaN,
|
| 329 |
-
"eval_accuracy_crop": 0.
|
| 330 |
"eval_iou_background": 0.0,
|
| 331 |
-
"eval_iou_crop": 0.
|
| 332 |
-
"eval_loss": 0.
|
| 333 |
-
"eval_mean_accuracy": 0.
|
| 334 |
-
"eval_mean_iou": 0.
|
| 335 |
-
"eval_overall_accuracy": 0.
|
| 336 |
-
"eval_runtime":
|
| 337 |
-
"eval_samples_per_second":
|
| 338 |
-
"eval_steps_per_second":
|
| 339 |
-
"step":
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 14.4,
|
| 343 |
-
"grad_norm": 0.2068459540605545,
|
| 344 |
-
"learning_rate": 5.3770733768437804e-05,
|
| 345 |
-
"loss": 0.022984485626220703,
|
| 346 |
-
"step": 9000
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 15.0,
|
| 350 |
"eval_accuracy_background": NaN,
|
| 351 |
-
"eval_accuracy_crop": 0.
|
| 352 |
"eval_iou_background": 0.0,
|
| 353 |
-
"eval_iou_crop": 0.
|
| 354 |
-
"eval_loss": 0.
|
| 355 |
-
"eval_mean_accuracy": 0.
|
| 356 |
-
"eval_mean_iou": 0.
|
| 357 |
-
"eval_overall_accuracy": 0.
|
| 358 |
-
"eval_runtime":
|
| 359 |
-
"eval_samples_per_second":
|
| 360 |
-
"eval_steps_per_second":
|
| 361 |
-
"step":
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 15.2,
|
| 365 |
-
"grad_norm": 0.28135085105895996,
|
| 366 |
-
"learning_rate": 5.271204116133174e-05,
|
| 367 |
-
"loss": 0.022900413513183594,
|
| 368 |
-
"step": 9500
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 16.0,
|
| 372 |
-
"grad_norm": 0.3188186287879944,
|
| 373 |
-
"learning_rate": 5.158252169806754e-05,
|
| 374 |
-
"loss": 0.02281988716125488,
|
| 375 |
-
"step": 10000
|
| 376 |
},
|
| 377 |
{
|
| 378 |
"epoch": 16.0,
|
| 379 |
"eval_accuracy_background": NaN,
|
| 380 |
-
"eval_accuracy_crop": 0.
|
| 381 |
"eval_iou_background": 0.0,
|
| 382 |
-
"eval_iou_crop": 0.
|
| 383 |
-
"eval_loss": 0.
|
| 384 |
-
"eval_mean_accuracy": 0.
|
| 385 |
-
"eval_mean_iou": 0.
|
| 386 |
-
"eval_overall_accuracy": 0.
|
| 387 |
-
"eval_runtime":
|
| 388 |
-
"eval_samples_per_second":
|
| 389 |
-
"eval_steps_per_second":
|
| 390 |
-
"step":
|
| 391 |
-
},
|
| 392 |
-
{
|
| 393 |
-
"epoch": 16.8,
|
| 394 |
-
"grad_norm": 0.2763194143772125,
|
| 395 |
-
"learning_rate": 5.038569775305373e-05,
|
| 396 |
-
"loss": 0.02196409034729004,
|
| 397 |
-
"step": 10500
|
| 398 |
},
|
| 399 |
{
|
| 400 |
"epoch": 17.0,
|
| 401 |
"eval_accuracy_background": NaN,
|
| 402 |
-
"eval_accuracy_crop": 0.
|
| 403 |
"eval_iou_background": 0.0,
|
| 404 |
-
"eval_iou_crop": 0.
|
| 405 |
-
"eval_loss": 0.
|
| 406 |
-
"eval_mean_accuracy": 0.
|
| 407 |
-
"eval_mean_iou": 0.
|
| 408 |
-
"eval_overall_accuracy": 0.
|
| 409 |
-
"eval_runtime":
|
| 410 |
-
"eval_samples_per_second":
|
| 411 |
-
"eval_steps_per_second":
|
| 412 |
-
"step":
|
| 413 |
-
},
|
| 414 |
-
{
|
| 415 |
-
"epoch": 17.6,
|
| 416 |
-
"grad_norm": 0.20888900756835938,
|
| 417 |
-
"learning_rate": 4.912530158781619e-05,
|
| 418 |
-
"loss": 0.02165799331665039,
|
| 419 |
-
"step": 11000
|
| 420 |
},
|
| 421 |
{
|
| 422 |
"epoch": 18.0,
|
| 423 |
"eval_accuracy_background": NaN,
|
| 424 |
-
"eval_accuracy_crop": 0.
|
| 425 |
"eval_iou_background": 0.0,
|
| 426 |
-
"eval_iou_crop": 0.
|
| 427 |
-
"eval_loss": 0.
|
| 428 |
-
"eval_mean_accuracy": 0.
|
| 429 |
-
"eval_mean_iou": 0.
|
| 430 |
-
"eval_overall_accuracy": 0.
|
| 431 |
-
"eval_runtime":
|
| 432 |
-
"eval_samples_per_second":
|
| 433 |
-
"eval_steps_per_second":
|
| 434 |
-
"step":
|
| 435 |
-
},
|
| 436 |
-
{
|
| 437 |
-
"epoch": 18.4,
|
| 438 |
-
"grad_norm": 0.23501233756542206,
|
| 439 |
-
"learning_rate": 4.780526371204638e-05,
|
| 440 |
-
"loss": 0.021172378540039064,
|
| 441 |
-
"step": 11500
|
| 442 |
},
|
| 443 |
{
|
| 444 |
"epoch": 19.0,
|
| 445 |
"eval_accuracy_background": NaN,
|
| 446 |
-
"eval_accuracy_crop": 0.
|
| 447 |
"eval_iou_background": 0.0,
|
| 448 |
-
"eval_iou_crop": 0.
|
| 449 |
-
"eval_loss": 0.
|
| 450 |
-
"eval_mean_accuracy": 0.
|
| 451 |
-
"eval_mean_iou": 0.
|
| 452 |
-
"eval_overall_accuracy": 0.
|
| 453 |
-
"eval_runtime":
|
| 454 |
-
"eval_samples_per_second":
|
| 455 |
-
"eval_steps_per_second":
|
| 456 |
-
"step":
|
| 457 |
-
},
|
| 458 |
-
{
|
| 459 |
-
"epoch": 19.2,
|
| 460 |
-
"grad_norm": 1.4721801280975342,
|
| 461 |
-
"learning_rate": 4.642970062641836e-05,
|
| 462 |
-
"loss": 0.02130653762817383,
|
| 463 |
-
"step": 12000
|
| 464 |
-
},
|
| 465 |
-
{
|
| 466 |
-
"epoch": 20.0,
|
| 467 |
-
"grad_norm": 0.5824402570724487,
|
| 468 |
-
"learning_rate": 4.5002901985398264e-05,
|
| 469 |
-
"loss": 0.02092706298828125,
|
| 470 |
-
"step": 12500
|
| 471 |
},
|
| 472 |
{
|
| 473 |
"epoch": 20.0,
|
| 474 |
"eval_accuracy_background": NaN,
|
| 475 |
-
"eval_accuracy_crop": 0.
|
| 476 |
"eval_iou_background": 0.0,
|
| 477 |
-
"eval_iou_crop": 0.
|
| 478 |
-
"eval_loss": 0.
|
| 479 |
-
"eval_mean_accuracy": 0.
|
| 480 |
-
"eval_mean_iou": 0.
|
| 481 |
-
"eval_overall_accuracy": 0.
|
| 482 |
-
"eval_runtime":
|
| 483 |
-
"eval_samples_per_second":
|
| 484 |
-
"eval_steps_per_second":
|
| 485 |
-
"step":
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 20.8,
|
| 489 |
-
"grad_norm": 0.17007386684417725,
|
| 490 |
-
"learning_rate": 4.3529317220078455e-05,
|
| 491 |
-
"loss": 0.020286674499511718,
|
| 492 |
-
"step": 13000
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 21.0,
|
| 496 |
"eval_accuracy_background": NaN,
|
| 497 |
-
"eval_accuracy_crop": 0.
|
| 498 |
"eval_iou_background": 0.0,
|
| 499 |
-
"eval_iou_crop": 0.
|
| 500 |
-
"eval_loss": 0.
|
| 501 |
-
"eval_mean_accuracy": 0.
|
| 502 |
-
"eval_mean_iou": 0.
|
| 503 |
-
"eval_overall_accuracy": 0.
|
| 504 |
-
"eval_runtime":
|
| 505 |
-
"eval_samples_per_second":
|
| 506 |
-
"eval_steps_per_second":
|
| 507 |
-
"step":
|
| 508 |
-
},
|
| 509 |
-
{
|
| 510 |
-
"epoch": 21.6,
|
| 511 |
-
"grad_norm": 0.4324241876602173,
|
| 512 |
-
"learning_rate": 4.2013541662752865e-05,
|
| 513 |
-
"loss": 0.020156242370605468,
|
| 514 |
-
"step": 13500
|
| 515 |
},
|
| 516 |
{
|
| 517 |
"epoch": 22.0,
|
| 518 |
"eval_accuracy_background": NaN,
|
| 519 |
-
"eval_accuracy_crop": 0.
|
| 520 |
"eval_iou_background": 0.0,
|
| 521 |
-
"eval_iou_crop": 0.
|
| 522 |
-
"eval_loss": 0.
|
| 523 |
-
"eval_mean_accuracy": 0.
|
| 524 |
-
"eval_mean_iou": 0.
|
| 525 |
-
"eval_overall_accuracy": 0.
|
| 526 |
-
"eval_runtime":
|
| 527 |
-
"eval_samples_per_second":
|
| 528 |
-
"eval_steps_per_second":
|
| 529 |
-
"step":
|
| 530 |
-
},
|
| 531 |
-
{
|
| 532 |
-
"epoch": 22.4,
|
| 533 |
-
"grad_norm": 0.1799526959657669,
|
| 534 |
-
"learning_rate": 4.0460302216503615e-05,
|
| 535 |
-
"loss": 0.020241693496704103,
|
| 536 |
-
"step": 14000
|
| 537 |
},
|
| 538 |
{
|
| 539 |
"epoch": 23.0,
|
| 540 |
"eval_accuracy_background": NaN,
|
| 541 |
-
"eval_accuracy_crop": 0.
|
| 542 |
"eval_iou_background": 0.0,
|
| 543 |
-
"eval_iou_crop": 0.
|
| 544 |
-
"eval_loss": 0.
|
| 545 |
-
"eval_mean_accuracy": 0.
|
| 546 |
-
"eval_mean_iou": 0.
|
| 547 |
-
"eval_overall_accuracy": 0.
|
| 548 |
-
"eval_runtime":
|
| 549 |
-
"eval_samples_per_second":
|
| 550 |
-
"eval_steps_per_second":
|
| 551 |
-
"step":
|
| 552 |
-
},
|
| 553 |
-
{
|
| 554 |
-
"epoch": 23.2,
|
| 555 |
-
"grad_norm": 0.19619110226631165,
|
| 556 |
-
"learning_rate": 3.887444261448782e-05,
|
| 557 |
-
"loss": 0.019864578247070313,
|
| 558 |
-
"step": 14500
|
| 559 |
-
},
|
| 560 |
-
{
|
| 561 |
-
"epoch": 24.0,
|
| 562 |
-
"grad_norm": 0.24014325439929962,
|
| 563 |
-
"learning_rate": 3.726090831489309e-05,
|
| 564 |
-
"loss": 0.01975071907043457,
|
| 565 |
-
"step": 15000
|
| 566 |
},
|
| 567 |
{
|
| 568 |
"epoch": 24.0,
|
| 569 |
"eval_accuracy_background": NaN,
|
| 570 |
-
"eval_accuracy_crop": 0.
|
| 571 |
"eval_iou_background": 0.0,
|
| 572 |
-
"eval_iou_crop": 0.
|
| 573 |
-
"eval_loss": 0.
|
| 574 |
-
"eval_mean_accuracy": 0.
|
| 575 |
-
"eval_mean_iou": 0.
|
| 576 |
-
"eval_overall_accuracy": 0.
|
| 577 |
-
"eval_runtime":
|
| 578 |
-
"eval_samples_per_second":
|
| 579 |
-
"eval_steps_per_second":
|
| 580 |
-
"step":
|
| 581 |
-
},
|
| 582 |
-
{
|
| 583 |
-
"epoch": 24.8,
|
| 584 |
-
"grad_norm": 0.17266370356082916,
|
| 585 |
-
"learning_rate": 3.5624731078666494e-05,
|
| 586 |
-
"loss": 0.01938687515258789,
|
| 587 |
-
"step": 15500
|
| 588 |
},
|
| 589 |
{
|
| 590 |
"epoch": 25.0,
|
| 591 |
"eval_accuracy_background": NaN,
|
| 592 |
-
"eval_accuracy_crop": 0.
|
| 593 |
"eval_iou_background": 0.0,
|
| 594 |
-
"eval_iou_crop": 0.
|
| 595 |
-
"eval_loss": 0.
|
| 596 |
-
"eval_mean_accuracy": 0.
|
| 597 |
-
"eval_mean_iou": 0.
|
| 598 |
-
"eval_overall_accuracy": 0.
|
| 599 |
-
"eval_runtime":
|
| 600 |
-
"eval_samples_per_second":
|
| 601 |
-
"eval_steps_per_second":
|
| 602 |
-
"step":
|
| 603 |
-
},
|
| 604 |
-
{
|
| 605 |
-
"epoch":
|
| 606 |
-
"
|
| 607 |
-
"
|
| 608 |
-
"
|
| 609 |
-
"
|
| 610 |
-
"
|
| 611 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 612 |
}
|
| 613 |
],
|
| 614 |
"logging_steps": 500,
|
| 615 |
-
"max_steps":
|
| 616 |
"num_input_tokens_seen": 0,
|
| 617 |
"num_train_epochs": 50,
|
| 618 |
"save_steps": 500,
|
|
@@ -637,7 +510,7 @@
|
|
| 637 |
"attributes": {}
|
| 638 |
}
|
| 639 |
},
|
| 640 |
-
"total_flos": 2.
|
| 641 |
"train_batch_size": 8,
|
| 642 |
"trial_name": null,
|
| 643 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 196,
|
| 3 |
+
"best_metric": 0.1087288111448288,
|
| 4 |
+
"best_model_checkpoint": "trainer_output/checkpoint-196",
|
| 5 |
+
"epoch": 31.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 217,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
{
|
| 13 |
"epoch": 1.0,
|
| 14 |
"eval_accuracy_background": NaN,
|
| 15 |
+
"eval_accuracy_crop": 0.6798034240092841,
|
| 16 |
"eval_iou_background": 0.0,
|
| 17 |
+
"eval_iou_crop": 0.6798034240092841,
|
| 18 |
+
"eval_loss": 0.6694175601005554,
|
| 19 |
+
"eval_mean_accuracy": 0.6798034240092841,
|
| 20 |
+
"eval_mean_iou": 0.33990171200464203,
|
| 21 |
+
"eval_overall_accuracy": 0.6798034240092841,
|
| 22 |
+
"eval_runtime": 0.9318,
|
| 23 |
+
"eval_samples_per_second": 10.732,
|
| 24 |
+
"eval_steps_per_second": 2.146,
|
| 25 |
+
"step": 7
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"epoch": 2.0,
|
| 29 |
"eval_accuracy_background": NaN,
|
| 30 |
+
"eval_accuracy_crop": 0.9246364650670945,
|
| 31 |
"eval_iou_background": 0.0,
|
| 32 |
+
"eval_iou_crop": 0.9246364650670945,
|
| 33 |
+
"eval_loss": 0.5930418968200684,
|
| 34 |
+
"eval_mean_accuracy": 0.9246364650670945,
|
| 35 |
+
"eval_mean_iou": 0.46231823253354726,
|
| 36 |
+
"eval_overall_accuracy": 0.9246364650670945,
|
| 37 |
+
"eval_runtime": 0.5622,
|
| 38 |
+
"eval_samples_per_second": 17.787,
|
| 39 |
+
"eval_steps_per_second": 3.557,
|
| 40 |
+
"step": 14
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"epoch": 3.0,
|
| 44 |
"eval_accuracy_background": NaN,
|
| 45 |
+
"eval_accuracy_crop": 0.9411854547997284,
|
| 46 |
"eval_iou_background": 0.0,
|
| 47 |
+
"eval_iou_crop": 0.9411854547997284,
|
| 48 |
+
"eval_loss": 0.4609629511833191,
|
| 49 |
+
"eval_mean_accuracy": 0.9411854547997284,
|
| 50 |
+
"eval_mean_iou": 0.4705927273998642,
|
| 51 |
+
"eval_overall_accuracy": 0.9411854547997284,
|
| 52 |
+
"eval_runtime": 0.5875,
|
| 53 |
+
"eval_samples_per_second": 17.022,
|
| 54 |
+
"eval_steps_per_second": 3.404,
|
| 55 |
+
"step": 21
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 56 |
},
|
| 57 |
{
|
| 58 |
"epoch": 4.0,
|
| 59 |
"eval_accuracy_background": NaN,
|
| 60 |
+
"eval_accuracy_crop": 0.9410732062778207,
|
| 61 |
"eval_iou_background": 0.0,
|
| 62 |
+
"eval_iou_crop": 0.9410732062778207,
|
| 63 |
+
"eval_loss": 0.3075406551361084,
|
| 64 |
+
"eval_mean_accuracy": 0.9410732062778207,
|
| 65 |
+
"eval_mean_iou": 0.4705366031389103,
|
| 66 |
+
"eval_overall_accuracy": 0.9410732062778207,
|
| 67 |
+
"eval_runtime": 0.5543,
|
| 68 |
+
"eval_samples_per_second": 18.04,
|
| 69 |
+
"eval_steps_per_second": 3.608,
|
| 70 |
+
"step": 28
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
},
|
| 72 |
{
|
| 73 |
"epoch": 5.0,
|
| 74 |
"eval_accuracy_background": NaN,
|
| 75 |
+
"eval_accuracy_crop": 0.9417221622132219,
|
| 76 |
"eval_iou_background": 0.0,
|
| 77 |
+
"eval_iou_crop": 0.9417221622132219,
|
| 78 |
+
"eval_loss": 0.20374695956707,
|
| 79 |
+
"eval_mean_accuracy": 0.9417221622132219,
|
| 80 |
+
"eval_mean_iou": 0.47086108110661096,
|
| 81 |
+
"eval_overall_accuracy": 0.9417221622132219,
|
| 82 |
+
"eval_runtime": 0.5695,
|
| 83 |
+
"eval_samples_per_second": 17.559,
|
| 84 |
+
"eval_steps_per_second": 3.512,
|
| 85 |
+
"step": 35
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 86 |
},
|
| 87 |
{
|
| 88 |
"epoch": 6.0,
|
| 89 |
"eval_accuracy_background": NaN,
|
| 90 |
+
"eval_accuracy_crop": 0.9323693454807089,
|
| 91 |
"eval_iou_background": 0.0,
|
| 92 |
+
"eval_iou_crop": 0.9323693454807089,
|
| 93 |
+
"eval_loss": 0.16684319078922272,
|
| 94 |
+
"eval_mean_accuracy": 0.9323693454807089,
|
| 95 |
+
"eval_mean_iou": 0.46618467274035447,
|
| 96 |
+
"eval_overall_accuracy": 0.9323693454807089,
|
| 97 |
+
"eval_runtime": 0.5262,
|
| 98 |
+
"eval_samples_per_second": 19.005,
|
| 99 |
+
"eval_steps_per_second": 3.801,
|
| 100 |
+
"step": 42
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 7.0,
|
| 104 |
"eval_accuracy_background": NaN,
|
| 105 |
+
"eval_accuracy_crop": 0.9503174547678327,
|
| 106 |
"eval_iou_background": 0.0,
|
| 107 |
+
"eval_iou_crop": 0.9503174547678327,
|
| 108 |
+
"eval_loss": 0.1420690268278122,
|
| 109 |
+
"eval_mean_accuracy": 0.9503174547678327,
|
| 110 |
+
"eval_mean_iou": 0.47515872738391635,
|
| 111 |
+
"eval_overall_accuracy": 0.9503174547678327,
|
| 112 |
+
"eval_runtime": 0.5946,
|
| 113 |
+
"eval_samples_per_second": 16.818,
|
| 114 |
+
"eval_steps_per_second": 3.364,
|
| 115 |
+
"step": 49
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
},
|
| 117 |
{
|
| 118 |
"epoch": 8.0,
|
| 119 |
"eval_accuracy_background": NaN,
|
| 120 |
+
"eval_accuracy_crop": 0.9546798126247078,
|
| 121 |
"eval_iou_background": 0.0,
|
| 122 |
+
"eval_iou_crop": 0.9546798126247078,
|
| 123 |
+
"eval_loss": 0.13820314407348633,
|
| 124 |
+
"eval_mean_accuracy": 0.9546798126247078,
|
| 125 |
+
"eval_mean_iou": 0.4773399063123539,
|
| 126 |
+
"eval_overall_accuracy": 0.9546798126247078,
|
| 127 |
+
"eval_runtime": 0.5971,
|
| 128 |
+
"eval_samples_per_second": 16.748,
|
| 129 |
+
"eval_steps_per_second": 3.35,
|
| 130 |
+
"step": 56
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 131 |
},
|
| 132 |
{
|
| 133 |
"epoch": 9.0,
|
| 134 |
"eval_accuracy_background": NaN,
|
| 135 |
+
"eval_accuracy_crop": 0.9473266143698963,
|
| 136 |
"eval_iou_background": 0.0,
|
| 137 |
+
"eval_iou_crop": 0.9473266143698963,
|
| 138 |
+
"eval_loss": 0.15878096222877502,
|
| 139 |
+
"eval_mean_accuracy": 0.9473266143698963,
|
| 140 |
+
"eval_mean_iou": 0.47366330718494815,
|
| 141 |
+
"eval_overall_accuracy": 0.9473266143698963,
|
| 142 |
+
"eval_runtime": 0.5896,
|
| 143 |
+
"eval_samples_per_second": 16.962,
|
| 144 |
+
"eval_steps_per_second": 3.392,
|
| 145 |
+
"step": 63
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 146 |
},
|
| 147 |
{
|
| 148 |
"epoch": 10.0,
|
| 149 |
"eval_accuracy_background": NaN,
|
| 150 |
+
"eval_accuracy_crop": 0.9690187945735507,
|
| 151 |
"eval_iou_background": 0.0,
|
| 152 |
+
"eval_iou_crop": 0.9690187945735507,
|
| 153 |
+
"eval_loss": 0.13168753683567047,
|
| 154 |
+
"eval_mean_accuracy": 0.9690187945735507,
|
| 155 |
+
"eval_mean_iou": 0.48450939728677533,
|
| 156 |
+
"eval_overall_accuracy": 0.9690187945735507,
|
| 157 |
+
"eval_runtime": 0.5435,
|
| 158 |
+
"eval_samples_per_second": 18.4,
|
| 159 |
+
"eval_steps_per_second": 3.68,
|
| 160 |
+
"step": 70
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 161 |
},
|
| 162 |
{
|
| 163 |
"epoch": 11.0,
|
| 164 |
"eval_accuracy_background": NaN,
|
| 165 |
+
"eval_accuracy_crop": 0.9671203837795366,
|
| 166 |
"eval_iou_background": 0.0,
|
| 167 |
+
"eval_iou_crop": 0.9671203837795366,
|
| 168 |
+
"eval_loss": 0.13073985278606415,
|
| 169 |
+
"eval_mean_accuracy": 0.9671203837795366,
|
| 170 |
+
"eval_mean_iou": 0.4835601918897683,
|
| 171 |
+
"eval_overall_accuracy": 0.9671203837795366,
|
| 172 |
+
"eval_runtime": 0.5417,
|
| 173 |
+
"eval_samples_per_second": 18.459,
|
| 174 |
+
"eval_steps_per_second": 3.692,
|
| 175 |
+
"step": 77
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 176 |
},
|
| 177 |
{
|
| 178 |
"epoch": 12.0,
|
| 179 |
"eval_accuracy_background": NaN,
|
| 180 |
+
"eval_accuracy_crop": 0.9797731843801581,
|
| 181 |
"eval_iou_background": 0.0,
|
| 182 |
+
"eval_iou_crop": 0.9797731843801581,
|
| 183 |
+
"eval_loss": 0.13279679417610168,
|
| 184 |
+
"eval_mean_accuracy": 0.9797731843801581,
|
| 185 |
+
"eval_mean_iou": 0.4898865921900791,
|
| 186 |
+
"eval_overall_accuracy": 0.9797731843801581,
|
| 187 |
+
"eval_runtime": 0.5591,
|
| 188 |
+
"eval_samples_per_second": 17.887,
|
| 189 |
+
"eval_steps_per_second": 3.577,
|
| 190 |
+
"step": 84
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 191 |
},
|
| 192 |
{
|
| 193 |
"epoch": 13.0,
|
| 194 |
"eval_accuracy_background": NaN,
|
| 195 |
+
"eval_accuracy_crop": 0.9725408219658703,
|
| 196 |
"eval_iou_background": 0.0,
|
| 197 |
+
"eval_iou_crop": 0.9725408219658703,
|
| 198 |
+
"eval_loss": 0.12647856771945953,
|
| 199 |
+
"eval_mean_accuracy": 0.9725408219658703,
|
| 200 |
+
"eval_mean_iou": 0.48627041098293516,
|
| 201 |
+
"eval_overall_accuracy": 0.9725408219658703,
|
| 202 |
+
"eval_runtime": 0.5658,
|
| 203 |
+
"eval_samples_per_second": 17.675,
|
| 204 |
+
"eval_steps_per_second": 3.535,
|
| 205 |
+
"step": 91
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 14.0,
|
| 209 |
"eval_accuracy_background": NaN,
|
| 210 |
+
"eval_accuracy_crop": 0.9771607993812224,
|
| 211 |
"eval_iou_background": 0.0,
|
| 212 |
+
"eval_iou_crop": 0.9771607993812224,
|
| 213 |
+
"eval_loss": 0.12825439870357513,
|
| 214 |
+
"eval_mean_accuracy": 0.9771607993812224,
|
| 215 |
+
"eval_mean_iou": 0.4885803996906112,
|
| 216 |
+
"eval_overall_accuracy": 0.9771607993812224,
|
| 217 |
+
"eval_runtime": 0.5494,
|
| 218 |
+
"eval_samples_per_second": 18.201,
|
| 219 |
+
"eval_steps_per_second": 3.64,
|
| 220 |
+
"step": 98
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 221 |
},
|
| 222 |
{
|
| 223 |
"epoch": 15.0,
|
| 224 |
"eval_accuracy_background": NaN,
|
| 225 |
+
"eval_accuracy_crop": 0.977466875951889,
|
| 226 |
"eval_iou_background": 0.0,
|
| 227 |
+
"eval_iou_crop": 0.977466875951889,
|
| 228 |
+
"eval_loss": 0.1286262422800064,
|
| 229 |
+
"eval_mean_accuracy": 0.977466875951889,
|
| 230 |
+
"eval_mean_iou": 0.4887334379759445,
|
| 231 |
+
"eval_overall_accuracy": 0.977466875951889,
|
| 232 |
+
"eval_runtime": 0.5651,
|
| 233 |
+
"eval_samples_per_second": 17.697,
|
| 234 |
+
"eval_steps_per_second": 3.539,
|
| 235 |
+
"step": 105
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 236 |
},
|
| 237 |
{
|
| 238 |
"epoch": 16.0,
|
| 239 |
"eval_accuracy_background": NaN,
|
| 240 |
+
"eval_accuracy_crop": 0.9788175384941892,
|
| 241 |
"eval_iou_background": 0.0,
|
| 242 |
+
"eval_iou_crop": 0.9788175384941892,
|
| 243 |
+
"eval_loss": 0.12346808612346649,
|
| 244 |
+
"eval_mean_accuracy": 0.9788175384941892,
|
| 245 |
+
"eval_mean_iou": 0.4894087692470946,
|
| 246 |
+
"eval_overall_accuracy": 0.9788175384941892,
|
| 247 |
+
"eval_runtime": 0.5641,
|
| 248 |
+
"eval_samples_per_second": 17.727,
|
| 249 |
+
"eval_steps_per_second": 3.545,
|
| 250 |
+
"step": 112
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 251 |
},
|
| 252 |
{
|
| 253 |
"epoch": 17.0,
|
| 254 |
"eval_accuracy_background": NaN,
|
| 255 |
+
"eval_accuracy_crop": 0.9795486873363426,
|
| 256 |
"eval_iou_background": 0.0,
|
| 257 |
+
"eval_iou_crop": 0.9795486873363426,
|
| 258 |
+
"eval_loss": 0.12134357541799545,
|
| 259 |
+
"eval_mean_accuracy": 0.9795486873363426,
|
| 260 |
+
"eval_mean_iou": 0.4897743436681713,
|
| 261 |
+
"eval_overall_accuracy": 0.9795486873363426,
|
| 262 |
+
"eval_runtime": 0.5708,
|
| 263 |
+
"eval_samples_per_second": 17.52,
|
| 264 |
+
"eval_steps_per_second": 3.504,
|
| 265 |
+
"step": 119
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 266 |
},
|
| 267 |
{
|
| 268 |
"epoch": 18.0,
|
| 269 |
"eval_accuracy_background": NaN,
|
| 270 |
+
"eval_accuracy_crop": 0.9820856266074387,
|
| 271 |
"eval_iou_background": 0.0,
|
| 272 |
+
"eval_iou_crop": 0.9820856266074387,
|
| 273 |
+
"eval_loss": 0.1223326101899147,
|
| 274 |
+
"eval_mean_accuracy": 0.9820856266074387,
|
| 275 |
+
"eval_mean_iou": 0.49104281330371935,
|
| 276 |
+
"eval_overall_accuracy": 0.9820856266074387,
|
| 277 |
+
"eval_runtime": 0.5809,
|
| 278 |
+
"eval_samples_per_second": 17.214,
|
| 279 |
+
"eval_steps_per_second": 3.443,
|
| 280 |
+
"step": 126
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 281 |
},
|
| 282 |
{
|
| 283 |
"epoch": 19.0,
|
| 284 |
"eval_accuracy_background": NaN,
|
| 285 |
+
"eval_accuracy_crop": 0.9763290562352828,
|
| 286 |
"eval_iou_background": 0.0,
|
| 287 |
+
"eval_iou_crop": 0.9763290562352828,
|
| 288 |
+
"eval_loss": 0.11792740970849991,
|
| 289 |
+
"eval_mean_accuracy": 0.9763290562352828,
|
| 290 |
+
"eval_mean_iou": 0.4881645281176414,
|
| 291 |
+
"eval_overall_accuracy": 0.9763290562352828,
|
| 292 |
+
"eval_runtime": 0.5496,
|
| 293 |
+
"eval_samples_per_second": 18.195,
|
| 294 |
+
"eval_steps_per_second": 3.639,
|
| 295 |
+
"step": 133
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 296 |
},
|
| 297 |
{
|
| 298 |
"epoch": 20.0,
|
| 299 |
"eval_accuracy_background": NaN,
|
| 300 |
+
"eval_accuracy_crop": 0.9828934479372341,
|
| 301 |
"eval_iou_background": 0.0,
|
| 302 |
+
"eval_iou_crop": 0.9828934479372341,
|
| 303 |
+
"eval_loss": 0.11694834381341934,
|
| 304 |
+
"eval_mean_accuracy": 0.9828934479372341,
|
| 305 |
+
"eval_mean_iou": 0.49144672396861705,
|
| 306 |
+
"eval_overall_accuracy": 0.9828934479372341,
|
| 307 |
+
"eval_runtime": 0.5655,
|
| 308 |
+
"eval_samples_per_second": 17.684,
|
| 309 |
+
"eval_steps_per_second": 3.537,
|
| 310 |
+
"step": 140
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 21.0,
|
| 314 |
"eval_accuracy_background": NaN,
|
| 315 |
+
"eval_accuracy_crop": 0.9816378592796099,
|
| 316 |
"eval_iou_background": 0.0,
|
| 317 |
+
"eval_iou_crop": 0.9816378592796099,
|
| 318 |
+
"eval_loss": 0.11534647643566132,
|
| 319 |
+
"eval_mean_accuracy": 0.9816378592796099,
|
| 320 |
+
"eval_mean_iou": 0.49081892963980495,
|
| 321 |
+
"eval_overall_accuracy": 0.9816378592796099,
|
| 322 |
+
"eval_runtime": 0.5715,
|
| 323 |
+
"eval_samples_per_second": 17.497,
|
| 324 |
+
"eval_steps_per_second": 3.499,
|
| 325 |
+
"step": 147
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 326 |
},
|
| 327 |
{
|
| 328 |
"epoch": 22.0,
|
| 329 |
"eval_accuracy_background": NaN,
|
| 330 |
+
"eval_accuracy_crop": 0.9803834973817879,
|
| 331 |
"eval_iou_background": 0.0,
|
| 332 |
+
"eval_iou_crop": 0.9803834973817879,
|
| 333 |
+
"eval_loss": 0.11552004516124725,
|
| 334 |
+
"eval_mean_accuracy": 0.9803834973817879,
|
| 335 |
+
"eval_mean_iou": 0.49019174869089394,
|
| 336 |
+
"eval_overall_accuracy": 0.9803834973817879,
|
| 337 |
+
"eval_runtime": 0.5635,
|
| 338 |
+
"eval_samples_per_second": 17.745,
|
| 339 |
+
"eval_steps_per_second": 3.549,
|
| 340 |
+
"step": 154
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 341 |
},
|
| 342 |
{
|
| 343 |
"epoch": 23.0,
|
| 344 |
"eval_accuracy_background": NaN,
|
| 345 |
+
"eval_accuracy_crop": 0.9838576811418189,
|
| 346 |
"eval_iou_background": 0.0,
|
| 347 |
+
"eval_iou_crop": 0.9838576811418189,
|
| 348 |
+
"eval_loss": 0.11426303535699844,
|
| 349 |
+
"eval_mean_accuracy": 0.9838576811418189,
|
| 350 |
+
"eval_mean_iou": 0.49192884057090946,
|
| 351 |
+
"eval_overall_accuracy": 0.9838576811418189,
|
| 352 |
+
"eval_runtime": 0.5699,
|
| 353 |
+
"eval_samples_per_second": 17.546,
|
| 354 |
+
"eval_steps_per_second": 3.509,
|
| 355 |
+
"step": 161
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 356 |
},
|
| 357 |
{
|
| 358 |
"epoch": 24.0,
|
| 359 |
"eval_accuracy_background": NaN,
|
| 360 |
+
"eval_accuracy_crop": 0.9825499551925982,
|
| 361 |
"eval_iou_background": 0.0,
|
| 362 |
+
"eval_iou_crop": 0.9825499551925982,
|
| 363 |
+
"eval_loss": 0.11150294542312622,
|
| 364 |
+
"eval_mean_accuracy": 0.9825499551925982,
|
| 365 |
+
"eval_mean_iou": 0.4912749775962991,
|
| 366 |
+
"eval_overall_accuracy": 0.9825499551925982,
|
| 367 |
+
"eval_runtime": 0.5714,
|
| 368 |
+
"eval_samples_per_second": 17.502,
|
| 369 |
+
"eval_steps_per_second": 3.5,
|
| 370 |
+
"step": 168
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 371 |
},
|
| 372 |
{
|
| 373 |
"epoch": 25.0,
|
| 374 |
"eval_accuracy_background": NaN,
|
| 375 |
+
"eval_accuracy_crop": 0.9840711373474141,
|
| 376 |
"eval_iou_background": 0.0,
|
| 377 |
+
"eval_iou_crop": 0.9840711373474141,
|
| 378 |
+
"eval_loss": 0.11130404472351074,
|
| 379 |
+
"eval_mean_accuracy": 0.9840711373474141,
|
| 380 |
+
"eval_mean_iou": 0.49203556867370707,
|
| 381 |
+
"eval_overall_accuracy": 0.9840711373474141,
|
| 382 |
+
"eval_runtime": 0.571,
|
| 383 |
+
"eval_samples_per_second": 17.514,
|
| 384 |
+
"eval_steps_per_second": 3.503,
|
| 385 |
+
"step": 175
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 26.0,
|
| 389 |
+
"eval_accuracy_background": NaN,
|
| 390 |
+
"eval_accuracy_crop": 0.984569201827136,
|
| 391 |
+
"eval_iou_background": 0.0,
|
| 392 |
+
"eval_iou_crop": 0.984569201827136,
|
| 393 |
+
"eval_loss": 0.11055157333612442,
|
| 394 |
+
"eval_mean_accuracy": 0.984569201827136,
|
| 395 |
+
"eval_mean_iou": 0.492284600913568,
|
| 396 |
+
"eval_overall_accuracy": 0.984569201827136,
|
| 397 |
+
"eval_runtime": 0.5673,
|
| 398 |
+
"eval_samples_per_second": 17.626,
|
| 399 |
+
"eval_steps_per_second": 3.525,
|
| 400 |
+
"step": 182
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 27.0,
|
| 404 |
+
"eval_accuracy_background": NaN,
|
| 405 |
+
"eval_accuracy_crop": 0.9813649052236045,
|
| 406 |
+
"eval_iou_background": 0.0,
|
| 407 |
+
"eval_iou_crop": 0.9813649052236045,
|
| 408 |
+
"eval_loss": 0.10974361002445221,
|
| 409 |
+
"eval_mean_accuracy": 0.9813649052236045,
|
| 410 |
+
"eval_mean_iou": 0.49068245261180227,
|
| 411 |
+
"eval_overall_accuracy": 0.9813649052236045,
|
| 412 |
+
"eval_runtime": 0.5718,
|
| 413 |
+
"eval_samples_per_second": 17.49,
|
| 414 |
+
"eval_steps_per_second": 3.498,
|
| 415 |
+
"step": 189
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 28.0,
|
| 419 |
+
"eval_accuracy_background": NaN,
|
| 420 |
+
"eval_accuracy_crop": 0.9816317254805985,
|
| 421 |
+
"eval_iou_background": 0.0,
|
| 422 |
+
"eval_iou_crop": 0.9816317254805985,
|
| 423 |
+
"eval_loss": 0.1087288111448288,
|
| 424 |
+
"eval_mean_accuracy": 0.9816317254805985,
|
| 425 |
+
"eval_mean_iou": 0.49081586274029926,
|
| 426 |
+
"eval_overall_accuracy": 0.9816317254805985,
|
| 427 |
+
"eval_runtime": 0.5686,
|
| 428 |
+
"eval_samples_per_second": 17.588,
|
| 429 |
+
"eval_steps_per_second": 3.518,
|
| 430 |
+
"step": 196
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 29.0,
|
| 434 |
+
"eval_accuracy_background": NaN,
|
| 435 |
+
"eval_accuracy_crop": 0.9839582754456052,
|
| 436 |
+
"eval_iou_background": 0.0,
|
| 437 |
+
"eval_iou_crop": 0.9839582754456052,
|
| 438 |
+
"eval_loss": 0.10943397134542465,
|
| 439 |
+
"eval_mean_accuracy": 0.9839582754456052,
|
| 440 |
+
"eval_mean_iou": 0.4919791377228026,
|
| 441 |
+
"eval_overall_accuracy": 0.9839582754456052,
|
| 442 |
+
"eval_runtime": 0.5769,
|
| 443 |
+
"eval_samples_per_second": 17.334,
|
| 444 |
+
"eval_steps_per_second": 3.467,
|
| 445 |
+
"step": 203
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 30.0,
|
| 449 |
+
"eval_accuracy_background": NaN,
|
| 450 |
+
"eval_accuracy_crop": 0.9821169089823966,
|
| 451 |
+
"eval_iou_background": 0.0,
|
| 452 |
+
"eval_iou_crop": 0.9821169089823966,
|
| 453 |
+
"eval_loss": 0.10927335917949677,
|
| 454 |
+
"eval_mean_accuracy": 0.9821169089823966,
|
| 455 |
+
"eval_mean_iou": 0.4910584544911983,
|
| 456 |
+
"eval_overall_accuracy": 0.9821169089823966,
|
| 457 |
+
"eval_runtime": 0.5836,
|
| 458 |
+
"eval_samples_per_second": 17.134,
|
| 459 |
+
"eval_steps_per_second": 3.427,
|
| 460 |
+
"step": 210
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 31.0,
|
| 464 |
+
"eval_accuracy_background": NaN,
|
| 465 |
+
"eval_accuracy_crop": 0.9813661319834068,
|
| 466 |
+
"eval_iou_background": 0.0,
|
| 467 |
+
"eval_iou_crop": 0.9813661319834068,
|
| 468 |
+
"eval_loss": 0.10906684398651123,
|
| 469 |
+
"eval_mean_accuracy": 0.9813661319834068,
|
| 470 |
+
"eval_mean_iou": 0.4906830659917034,
|
| 471 |
+
"eval_overall_accuracy": 0.9813661319834068,
|
| 472 |
+
"eval_runtime": 0.5612,
|
| 473 |
+
"eval_samples_per_second": 17.82,
|
| 474 |
+
"eval_steps_per_second": 3.564,
|
| 475 |
+
"step": 217
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 31.0,
|
| 479 |
+
"step": 217,
|
| 480 |
+
"total_flos": 2.7711697679548416e+16,
|
| 481 |
+
"train_loss": 0.13814654548047325,
|
| 482 |
+
"train_runtime": 56.09,
|
| 483 |
+
"train_samples_per_second": 45.463,
|
| 484 |
+
"train_steps_per_second": 6.24
|
| 485 |
}
|
| 486 |
],
|
| 487 |
"logging_steps": 500,
|
| 488 |
+
"max_steps": 350,
|
| 489 |
"num_input_tokens_seen": 0,
|
| 490 |
"num_train_epochs": 50,
|
| 491 |
"save_steps": 500,
|
|
|
|
| 510 |
"attributes": {}
|
| 511 |
}
|
| 512 |
},
|
| 513 |
+
"total_flos": 2.7711697679548416e+16,
|
| 514 |
"train_batch_size": 8,
|
| 515 |
"trial_name": null,
|
| 516 |
"trial_params": null
|