Commit
·
4320ea0
1
Parent(s):
5d74d76
yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor
Browse files- README.md +17 -19
- all_results.json +11 -11
- config.json +1 -1
- eval_results.json +6 -6
- pytorch_model.bin +1 -1
- train_results.json +6 -6
- trainer_state.json +355 -487
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor](https://huggingface.co/yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -38,35 +38,33 @@ More information needed
|
|
| 38 |
|
| 39 |
The following hyperparameters were used during training:
|
| 40 |
- learning_rate: 5e-05
|
| 41 |
-
- train_batch_size:
|
| 42 |
- eval_batch_size: 8
|
| 43 |
- seed: 42
|
| 44 |
- gradient_accumulation_steps: 8
|
| 45 |
-
- total_train_batch_size:
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
-
- num_epochs:
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
-
| Training Loss | Epoch | Step
|
| 53 |
-
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.
|
| 61 |
-
| 0.
|
| 62 |
-
| 0.
|
| 63 |
-
| 0.0137 | 52.81 | 10000 | 0.3257 | 0.9352 |
|
| 64 |
-
| 0.0174 | 58.09 | 11000 | 0.3097 | 0.9340 |
|
| 65 |
|
| 66 |
|
| 67 |
### Framework versions
|
| 68 |
|
| 69 |
-
- Transformers 4.
|
| 70 |
- Pytorch 2.0.0
|
| 71 |
- Datasets 2.11.0
|
| 72 |
- Tokenizers 0.13.3
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor](https://huggingface.co/yashcode00/wav2vec2-large-xlsr-indian-language-classification-featureExtractor) on an unknown dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.2965
|
| 21 |
+
- Accuracy: 0.9319
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 38 |
|
| 39 |
The following hyperparameters were used during training:
|
| 40 |
- learning_rate: 5e-05
|
| 41 |
+
- train_batch_size: 16
|
| 42 |
- eval_batch_size: 8
|
| 43 |
- seed: 42
|
| 44 |
- gradient_accumulation_steps: 8
|
| 45 |
+
- total_train_batch_size: 128
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
+
- num_epochs: 100
|
| 49 |
|
| 50 |
### Training results
|
| 51 |
|
| 52 |
+
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 53 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 54 |
+
| 0.0272 | 10.55 | 1000 | 0.2915 | 0.9257 |
|
| 55 |
+
| 0.0172 | 21.11 | 2000 | 0.2769 | 0.9332 |
|
| 56 |
+
| 0.0178 | 31.66 | 3000 | 0.2928 | 0.9323 |
|
| 57 |
+
| 0.0145 | 42.22 | 4000 | 0.2718 | 0.9356 |
|
| 58 |
+
| 0.0147 | 52.77 | 5000 | 0.2676 | 0.9348 |
|
| 59 |
+
| 0.0135 | 63.32 | 6000 | 0.2731 | 0.9398 |
|
| 60 |
+
| 0.0105 | 73.88 | 7000 | 0.3145 | 0.9336 |
|
| 61 |
+
| 0.0075 | 84.43 | 8000 | 0.2971 | 0.9319 |
|
| 62 |
+
| 0.0078 | 94.99 | 9000 | 0.2950 | 0.9328 |
|
|
|
|
|
|
|
| 63 |
|
| 64 |
|
| 65 |
### Framework versions
|
| 66 |
|
| 67 |
+
- Transformers 4.33.0
|
| 68 |
- Pytorch 2.0.0
|
| 69 |
- Datasets 2.11.0
|
| 70 |
- Tokenizers 0.13.3
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 50.
|
| 6 |
"eval_samples": 2424,
|
| 7 |
-
"eval_samples_per_second": 47.
|
| 8 |
-
"eval_steps_per_second": 5.
|
| 9 |
-
"total_flos":
|
| 10 |
-
"train_loss": 0.
|
| 11 |
-
"train_runtime":
|
| 12 |
"train_samples": 12120,
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second": 0.
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 99.21,
|
| 3 |
+
"eval_accuracy": 0.9319307208061218,
|
| 4 |
+
"eval_loss": 0.29647526144981384,
|
| 5 |
+
"eval_runtime": 50.5435,
|
| 6 |
"eval_samples": 2424,
|
| 7 |
+
"eval_samples_per_second": 47.959,
|
| 8 |
+
"eval_steps_per_second": 5.995,
|
| 9 |
+
"total_flos": 3.653391792237703e+19,
|
| 10 |
+
"train_loss": 0.01680715578667661,
|
| 11 |
+
"train_runtime": 36294.494,
|
| 12 |
"train_samples": 12120,
|
| 13 |
+
"train_samples_per_second": 33.393,
|
| 14 |
+
"train_steps_per_second": 0.259
|
| 15 |
}
|
config.json
CHANGED
|
@@ -139,7 +139,7 @@
|
|
| 139 |
1
|
| 140 |
],
|
| 141 |
"torch_dtype": "float32",
|
| 142 |
-
"transformers_version": "4.
|
| 143 |
"use_weighted_layer_sum": false,
|
| 144 |
"vocab_size": 32,
|
| 145 |
"xvector_output_dim": 512
|
|
|
|
| 139 |
1
|
| 140 |
],
|
| 141 |
"torch_dtype": "float32",
|
| 142 |
+
"transformers_version": "4.33.0",
|
| 143 |
"use_weighted_layer_sum": false,
|
| 144 |
"vocab_size": 32,
|
| 145 |
"xvector_output_dim": 512
|
eval_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_accuracy": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 50.
|
| 6 |
"eval_samples": 2424,
|
| 7 |
-
"eval_samples_per_second": 47.
|
| 8 |
-
"eval_steps_per_second": 5.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 99.21,
|
| 3 |
+
"eval_accuracy": 0.9319307208061218,
|
| 4 |
+
"eval_loss": 0.29647526144981384,
|
| 5 |
+
"eval_runtime": 50.5435,
|
| 6 |
"eval_samples": 2424,
|
| 7 |
+
"eval_samples_per_second": 47.959,
|
| 8 |
+
"eval_steps_per_second": 5.995
|
| 9 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1266146037
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:360b8fd02ab317b95f64242ee0a98ef6b12d72844557ec2da4385fe18fe39db6
|
| 3 |
size 1266146037
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 12120,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 99.21,
|
| 3 |
+
"total_flos": 3.653391792237703e+19,
|
| 4 |
+
"train_loss": 0.01680715578667661,
|
| 5 |
+
"train_runtime": 36294.494,
|
| 6 |
"train_samples": 12120,
|
| 7 |
+
"train_samples_per_second": 33.393,
|
| 8 |
+
"train_steps_per_second": 0.259
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,814 +1,682 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 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":
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 0.
|
| 15 |
"step": 100
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch":
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss": 0.
|
| 21 |
"step": 200
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch":
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 300
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch":
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 400
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch":
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch":
|
| 43 |
-
"learning_rate": 4.
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 600
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch":
|
| 49 |
-
"learning_rate": 4.
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 700
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch":
|
| 55 |
-
"learning_rate": 4.
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 800
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch":
|
| 61 |
-
"learning_rate": 4.
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 900
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch":
|
| 67 |
-
"learning_rate": 4.
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 1000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
-
"eval_accuracy": 0.
|
| 74 |
-
"eval_loss": 0.
|
| 75 |
-
"eval_runtime":
|
| 76 |
-
"eval_samples_per_second":
|
| 77 |
-
"eval_steps_per_second":
|
| 78 |
"step": 1000
|
| 79 |
},
|
| 80 |
{
|
| 81 |
-
"epoch":
|
| 82 |
-
"learning_rate": 4.
|
| 83 |
-
"loss": 0.
|
| 84 |
"step": 1100
|
| 85 |
},
|
| 86 |
{
|
| 87 |
-
"epoch":
|
| 88 |
-
"learning_rate": 4.
|
| 89 |
-
"loss": 0.
|
| 90 |
"step": 1200
|
| 91 |
},
|
| 92 |
{
|
| 93 |
-
"epoch":
|
| 94 |
-
"learning_rate": 4.
|
| 95 |
-
"loss": 0.
|
| 96 |
"step": 1300
|
| 97 |
},
|
| 98 |
{
|
| 99 |
-
"epoch":
|
| 100 |
-
"learning_rate": 4.
|
| 101 |
-
"loss": 0.
|
| 102 |
"step": 1400
|
| 103 |
},
|
| 104 |
{
|
| 105 |
-
"epoch":
|
| 106 |
-
"learning_rate": 4.
|
| 107 |
-
"loss": 0.
|
| 108 |
"step": 1500
|
| 109 |
},
|
| 110 |
{
|
| 111 |
-
"epoch":
|
| 112 |
-
"learning_rate": 4.
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 1600
|
| 115 |
},
|
| 116 |
{
|
| 117 |
-
"epoch":
|
| 118 |
-
"learning_rate": 4.
|
| 119 |
-
"loss": 0.
|
| 120 |
"step": 1700
|
| 121 |
},
|
| 122 |
{
|
| 123 |
-
"epoch":
|
| 124 |
-
"learning_rate": 4.
|
| 125 |
-
"loss": 0.
|
| 126 |
"step": 1800
|
| 127 |
},
|
| 128 |
{
|
| 129 |
-
"epoch":
|
| 130 |
-
"learning_rate":
|
| 131 |
-
"loss": 0.
|
| 132 |
"step": 1900
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"learning_rate":
|
| 137 |
-
"loss": 0.
|
| 138 |
"step": 2000
|
| 139 |
},
|
| 140 |
{
|
| 141 |
-
"epoch":
|
| 142 |
-
"eval_accuracy": 0.
|
| 143 |
-
"eval_loss": 0.
|
| 144 |
-
"eval_runtime": 51.
|
| 145 |
-
"eval_samples_per_second": 47.
|
| 146 |
-
"eval_steps_per_second": 5.
|
| 147 |
"step": 2000
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
-
"learning_rate":
|
| 152 |
-
"loss": 0.
|
| 153 |
"step": 2100
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 2200
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"learning_rate": 3.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 2300
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"learning_rate": 3.
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 2400
|
| 172 |
},
|
| 173 |
{
|
| 174 |
-
"epoch":
|
| 175 |
-
"learning_rate": 3.
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 2500
|
| 178 |
},
|
| 179 |
{
|
| 180 |
-
"epoch":
|
| 181 |
-
"learning_rate": 3.
|
| 182 |
-
"loss": 0.
|
| 183 |
"step": 2600
|
| 184 |
},
|
| 185 |
{
|
| 186 |
-
"epoch":
|
| 187 |
-
"learning_rate": 3.
|
| 188 |
-
"loss": 0.
|
| 189 |
"step": 2700
|
| 190 |
},
|
| 191 |
{
|
| 192 |
-
"epoch":
|
| 193 |
-
"learning_rate": 3.
|
| 194 |
-
"loss": 0.
|
| 195 |
"step": 2800
|
| 196 |
},
|
| 197 |
{
|
| 198 |
-
"epoch":
|
| 199 |
-
"learning_rate": 3.
|
| 200 |
-
"loss": 0.
|
| 201 |
"step": 2900
|
| 202 |
},
|
| 203 |
{
|
| 204 |
-
"epoch":
|
| 205 |
-
"learning_rate": 3.
|
| 206 |
-
"loss": 0.
|
| 207 |
"step": 3000
|
| 208 |
},
|
| 209 |
{
|
| 210 |
-
"epoch":
|
| 211 |
-
"eval_accuracy": 0.
|
| 212 |
-
"eval_loss": 0.
|
| 213 |
-
"eval_runtime":
|
| 214 |
-
"eval_samples_per_second": 47.
|
| 215 |
-
"eval_steps_per_second": 5.
|
| 216 |
"step": 3000
|
| 217 |
},
|
| 218 |
{
|
| 219 |
-
"epoch":
|
| 220 |
-
"learning_rate": 3.
|
| 221 |
-
"loss": 0.
|
| 222 |
"step": 3100
|
| 223 |
},
|
| 224 |
{
|
| 225 |
-
"epoch":
|
| 226 |
-
"learning_rate": 3.
|
| 227 |
-
"loss": 0.
|
| 228 |
"step": 3200
|
| 229 |
},
|
| 230 |
{
|
| 231 |
-
"epoch":
|
| 232 |
-
"learning_rate": 3.
|
| 233 |
-
"loss": 0.
|
| 234 |
"step": 3300
|
| 235 |
},
|
| 236 |
{
|
| 237 |
-
"epoch":
|
| 238 |
-
"learning_rate": 3.
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 3400
|
| 241 |
},
|
| 242 |
{
|
| 243 |
-
"epoch":
|
| 244 |
-
"learning_rate": 3.
|
| 245 |
-
"loss": 0.
|
| 246 |
"step": 3500
|
| 247 |
},
|
| 248 |
{
|
| 249 |
-
"epoch":
|
| 250 |
-
"learning_rate": 3.
|
| 251 |
-
"loss": 0.
|
| 252 |
"step": 3600
|
| 253 |
},
|
| 254 |
{
|
| 255 |
-
"epoch":
|
| 256 |
-
"learning_rate": 3.
|
| 257 |
-
"loss": 0.
|
| 258 |
"step": 3700
|
| 259 |
},
|
| 260 |
{
|
| 261 |
-
"epoch":
|
| 262 |
-
"learning_rate":
|
| 263 |
-
"loss": 0.
|
| 264 |
"step": 3800
|
| 265 |
},
|
| 266 |
{
|
| 267 |
-
"epoch":
|
| 268 |
-
"learning_rate":
|
| 269 |
-
"loss": 0.
|
| 270 |
"step": 3900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
-
"epoch":
|
| 274 |
-
"learning_rate":
|
| 275 |
-
"loss": 0.
|
| 276 |
"step": 4000
|
| 277 |
},
|
| 278 |
{
|
| 279 |
-
"epoch":
|
| 280 |
-
"eval_accuracy": 0.
|
| 281 |
-
"eval_loss": 0.
|
| 282 |
-
"eval_runtime": 50.
|
| 283 |
-
"eval_samples_per_second":
|
| 284 |
-
"eval_steps_per_second":
|
| 285 |
"step": 4000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
-
"epoch":
|
| 289 |
-
"learning_rate":
|
| 290 |
-
"loss": 0.
|
| 291 |
"step": 4100
|
| 292 |
},
|
| 293 |
{
|
| 294 |
-
"epoch":
|
| 295 |
-
"learning_rate":
|
| 296 |
-
"loss": 0.
|
| 297 |
"step": 4200
|
| 298 |
},
|
| 299 |
{
|
| 300 |
-
"epoch":
|
| 301 |
-
"learning_rate":
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 4300
|
| 304 |
},
|
| 305 |
{
|
| 306 |
-
"epoch":
|
| 307 |
-
"learning_rate":
|
| 308 |
-
"loss": 0.
|
| 309 |
"step": 4400
|
| 310 |
},
|
| 311 |
{
|
| 312 |
-
"epoch":
|
| 313 |
-
"learning_rate":
|
| 314 |
-
"loss": 0.
|
| 315 |
"step": 4500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
-
"epoch":
|
| 319 |
-
"learning_rate": 2.
|
| 320 |
-
"loss": 0.
|
| 321 |
"step": 4600
|
| 322 |
},
|
| 323 |
{
|
| 324 |
-
"epoch":
|
| 325 |
-
"learning_rate": 2.
|
| 326 |
-
"loss": 0.
|
| 327 |
"step": 4700
|
| 328 |
},
|
| 329 |
{
|
| 330 |
-
"epoch":
|
| 331 |
-
"learning_rate": 2.
|
| 332 |
-
"loss": 0.
|
| 333 |
"step": 4800
|
| 334 |
},
|
| 335 |
{
|
| 336 |
-
"epoch":
|
| 337 |
-
"learning_rate": 2.
|
| 338 |
-
"loss": 0.
|
| 339 |
"step": 4900
|
| 340 |
},
|
| 341 |
{
|
| 342 |
-
"epoch":
|
| 343 |
-
"learning_rate": 2.
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 5000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
-
"epoch":
|
| 349 |
-
"eval_accuracy": 0.
|
| 350 |
-
"eval_loss": 0.
|
| 351 |
-
"eval_runtime": 49.
|
| 352 |
-
"eval_samples_per_second":
|
| 353 |
-
"eval_steps_per_second": 6.
|
| 354 |
"step": 5000
|
| 355 |
},
|
| 356 |
{
|
| 357 |
-
"epoch":
|
| 358 |
-
"learning_rate": 2.
|
| 359 |
-
"loss": 0.
|
| 360 |
"step": 5100
|
| 361 |
},
|
| 362 |
{
|
| 363 |
-
"epoch":
|
| 364 |
-
"learning_rate": 2.
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 5200
|
| 367 |
},
|
| 368 |
{
|
| 369 |
-
"epoch":
|
| 370 |
-
"learning_rate": 2.
|
| 371 |
-
"loss": 0.
|
| 372 |
"step": 5300
|
| 373 |
},
|
| 374 |
{
|
| 375 |
-
"epoch":
|
| 376 |
-
"learning_rate": 2.
|
| 377 |
-
"loss": 0.
|
| 378 |
"step": 5400
|
| 379 |
},
|
| 380 |
{
|
| 381 |
-
"epoch":
|
| 382 |
-
"learning_rate": 2.
|
| 383 |
-
"loss": 0.
|
| 384 |
"step": 5500
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch":
|
| 388 |
-
"learning_rate": 2.
|
| 389 |
-
"loss": 0.
|
| 390 |
"step": 5600
|
| 391 |
},
|
| 392 |
{
|
| 393 |
-
"epoch":
|
| 394 |
-
"learning_rate":
|
| 395 |
-
"loss": 0.
|
| 396 |
"step": 5700
|
| 397 |
},
|
| 398 |
{
|
| 399 |
-
"epoch":
|
| 400 |
-
"learning_rate":
|
| 401 |
-
"loss": 0.
|
| 402 |
"step": 5800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
-
"epoch":
|
| 406 |
-
"learning_rate":
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 5900
|
| 409 |
},
|
| 410 |
{
|
| 411 |
-
"epoch":
|
| 412 |
-
"learning_rate":
|
| 413 |
-
"loss": 0.
|
| 414 |
"step": 6000
|
| 415 |
},
|
| 416 |
{
|
| 417 |
-
"epoch":
|
| 418 |
-
"eval_accuracy": 0.
|
| 419 |
-
"eval_loss": 0.
|
| 420 |
-
"eval_runtime": 49.
|
| 421 |
-
"eval_samples_per_second": 48.
|
| 422 |
-
"eval_steps_per_second": 6.
|
| 423 |
"step": 6000
|
| 424 |
},
|
| 425 |
{
|
| 426 |
-
"epoch":
|
| 427 |
-
"learning_rate":
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 6100
|
| 430 |
},
|
| 431 |
{
|
| 432 |
-
"epoch":
|
| 433 |
-
"learning_rate":
|
| 434 |
-
"loss": 0.
|
| 435 |
"step": 6200
|
| 436 |
},
|
| 437 |
{
|
| 438 |
-
"epoch":
|
| 439 |
-
"learning_rate":
|
| 440 |
-
"loss": 0.
|
| 441 |
"step": 6300
|
| 442 |
},
|
| 443 |
{
|
| 444 |
-
"epoch":
|
| 445 |
-
"learning_rate":
|
| 446 |
-
"loss": 0.
|
| 447 |
"step": 6400
|
| 448 |
},
|
| 449 |
{
|
| 450 |
-
"epoch":
|
| 451 |
-
"learning_rate":
|
| 452 |
-
"loss": 0.
|
| 453 |
"step": 6500
|
| 454 |
},
|
| 455 |
{
|
| 456 |
-
"epoch":
|
| 457 |
-
"learning_rate":
|
| 458 |
-
"loss": 0.
|
| 459 |
"step": 6600
|
| 460 |
},
|
| 461 |
{
|
| 462 |
-
"epoch":
|
| 463 |
-
"learning_rate":
|
| 464 |
-
"loss": 0.
|
| 465 |
"step": 6700
|
| 466 |
},
|
| 467 |
{
|
| 468 |
-
"epoch":
|
| 469 |
-
"learning_rate":
|
| 470 |
-
"loss": 0.
|
| 471 |
"step": 6800
|
| 472 |
},
|
| 473 |
{
|
| 474 |
-
"epoch":
|
| 475 |
-
"learning_rate": 1.
|
| 476 |
-
"loss": 0.
|
| 477 |
"step": 6900
|
| 478 |
},
|
| 479 |
{
|
| 480 |
-
"epoch":
|
| 481 |
-
"learning_rate": 1.
|
| 482 |
-
"loss": 0.
|
| 483 |
"step": 7000
|
| 484 |
},
|
| 485 |
{
|
| 486 |
-
"epoch":
|
| 487 |
-
"eval_accuracy": 0.
|
| 488 |
-
"eval_loss": 0.
|
| 489 |
-
"eval_runtime":
|
| 490 |
-
"eval_samples_per_second":
|
| 491 |
-
"eval_steps_per_second": 6.
|
| 492 |
"step": 7000
|
| 493 |
},
|
| 494 |
{
|
| 495 |
-
"epoch":
|
| 496 |
-
"learning_rate": 1.
|
| 497 |
-
"loss": 0.
|
| 498 |
"step": 7100
|
| 499 |
},
|
| 500 |
{
|
| 501 |
-
"epoch":
|
| 502 |
-
"learning_rate": 1.
|
| 503 |
-
"loss": 0.
|
| 504 |
"step": 7200
|
| 505 |
},
|
| 506 |
{
|
| 507 |
-
"epoch":
|
| 508 |
-
"learning_rate": 1.
|
| 509 |
-
"loss": 0.
|
| 510 |
"step": 7300
|
| 511 |
},
|
| 512 |
{
|
| 513 |
-
"epoch":
|
| 514 |
-
"learning_rate": 1.
|
| 515 |
-
"loss": 0.
|
| 516 |
"step": 7400
|
| 517 |
},
|
| 518 |
{
|
| 519 |
-
"epoch":
|
| 520 |
-
"learning_rate": 1.
|
| 521 |
-
"loss": 0.
|
| 522 |
"step": 7500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
-
"epoch":
|
| 526 |
-
"learning_rate":
|
| 527 |
-
"loss": 0.
|
| 528 |
"step": 7600
|
| 529 |
},
|
| 530 |
{
|
| 531 |
-
"epoch":
|
| 532 |
-
"learning_rate":
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 7700
|
| 535 |
},
|
| 536 |
{
|
| 537 |
-
"epoch":
|
| 538 |
-
"learning_rate":
|
| 539 |
-
"loss": 0.
|
| 540 |
"step": 7800
|
| 541 |
},
|
| 542 |
{
|
| 543 |
-
"epoch":
|
| 544 |
-
"learning_rate":
|
| 545 |
-
"loss": 0.
|
| 546 |
"step": 7900
|
| 547 |
},
|
| 548 |
{
|
| 549 |
-
"epoch":
|
| 550 |
-
"learning_rate":
|
| 551 |
-
"loss": 0.
|
| 552 |
"step": 8000
|
| 553 |
},
|
| 554 |
{
|
| 555 |
-
"epoch":
|
| 556 |
-
"eval_accuracy": 0.
|
| 557 |
-
"eval_loss": 0.
|
| 558 |
-
"eval_runtime":
|
| 559 |
-
"eval_samples_per_second": 48.
|
| 560 |
-
"eval_steps_per_second": 6.
|
| 561 |
"step": 8000
|
| 562 |
},
|
| 563 |
{
|
| 564 |
-
"epoch":
|
| 565 |
-
"learning_rate":
|
| 566 |
-
"loss": 0.
|
| 567 |
"step": 8100
|
| 568 |
},
|
| 569 |
{
|
| 570 |
-
"epoch":
|
| 571 |
-
"learning_rate":
|
| 572 |
-
"loss": 0.
|
| 573 |
"step": 8200
|
| 574 |
},
|
| 575 |
{
|
| 576 |
-
"epoch":
|
| 577 |
-
"learning_rate":
|
| 578 |
-
"loss": 0.
|
| 579 |
"step": 8300
|
| 580 |
},
|
| 581 |
{
|
| 582 |
-
"epoch":
|
| 583 |
-
"learning_rate":
|
| 584 |
-
"loss": 0.
|
| 585 |
"step": 8400
|
| 586 |
},
|
| 587 |
{
|
| 588 |
-
"epoch":
|
| 589 |
-
"learning_rate":
|
| 590 |
-
"loss": 0.
|
| 591 |
"step": 8500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
-
"epoch":
|
| 595 |
-
"learning_rate":
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 8600
|
| 598 |
},
|
| 599 |
{
|
| 600 |
-
"epoch":
|
| 601 |
-
"learning_rate":
|
| 602 |
-
"loss": 0.
|
| 603 |
"step": 8700
|
| 604 |
},
|
| 605 |
{
|
| 606 |
-
"epoch":
|
| 607 |
-
"learning_rate":
|
| 608 |
-
"loss": 0.
|
| 609 |
"step": 8800
|
| 610 |
},
|
| 611 |
{
|
| 612 |
-
"epoch":
|
| 613 |
-
"learning_rate":
|
| 614 |
-
"loss": 0.
|
| 615 |
"step": 8900
|
| 616 |
},
|
| 617 |
{
|
| 618 |
-
"epoch":
|
| 619 |
-
"learning_rate":
|
| 620 |
-
"loss": 0.
|
| 621 |
"step": 9000
|
| 622 |
},
|
| 623 |
{
|
| 624 |
-
"epoch":
|
| 625 |
-
"eval_accuracy": 0.
|
| 626 |
-
"eval_loss": 0.
|
| 627 |
-
"eval_runtime":
|
| 628 |
-
"eval_samples_per_second":
|
| 629 |
-
"eval_steps_per_second":
|
| 630 |
"step": 9000
|
| 631 |
},
|
| 632 |
{
|
| 633 |
-
"epoch":
|
| 634 |
-
"learning_rate":
|
| 635 |
-
"loss": 0.
|
| 636 |
"step": 9100
|
| 637 |
},
|
| 638 |
{
|
| 639 |
-
"epoch":
|
| 640 |
-
"learning_rate":
|
| 641 |
-
"loss": 0.
|
| 642 |
"step": 9200
|
| 643 |
},
|
| 644 |
{
|
| 645 |
-
"epoch":
|
| 646 |
-
"learning_rate":
|
| 647 |
-
"loss": 0.
|
| 648 |
"step": 9300
|
| 649 |
},
|
| 650 |
{
|
| 651 |
-
"epoch":
|
| 652 |
-
"learning_rate":
|
| 653 |
-
"loss": 0.
|
| 654 |
"step": 9400
|
| 655 |
},
|
| 656 |
{
|
| 657 |
-
"epoch":
|
| 658 |
-
"
|
| 659 |
-
"
|
| 660 |
-
"
|
|
|
|
|
|
|
|
|
|
| 661 |
},
|
| 662 |
{
|
| 663 |
-
"epoch":
|
| 664 |
-
"
|
| 665 |
-
"
|
| 666 |
-
"
|
| 667 |
-
|
| 668 |
-
|
| 669 |
-
"
|
| 670 |
-
"learning_rate": 7.253086419753087e-06,
|
| 671 |
-
"loss": 0.012,
|
| 672 |
-
"step": 9700
|
| 673 |
-
},
|
| 674 |
-
{
|
| 675 |
-
"epoch": 51.75,
|
| 676 |
-
"learning_rate": 6.812169312169313e-06,
|
| 677 |
-
"loss": 0.0112,
|
| 678 |
-
"step": 9800
|
| 679 |
-
},
|
| 680 |
-
{
|
| 681 |
-
"epoch": 52.28,
|
| 682 |
-
"learning_rate": 6.371252204585539e-06,
|
| 683 |
-
"loss": 0.0193,
|
| 684 |
-
"step": 9900
|
| 685 |
-
},
|
| 686 |
-
{
|
| 687 |
-
"epoch": 52.81,
|
| 688 |
-
"learning_rate": 5.930335097001764e-06,
|
| 689 |
-
"loss": 0.0137,
|
| 690 |
-
"step": 10000
|
| 691 |
-
},
|
| 692 |
-
{
|
| 693 |
-
"epoch": 52.81,
|
| 694 |
-
"eval_accuracy": 0.9352310299873352,
|
| 695 |
-
"eval_loss": 0.3257134258747101,
|
| 696 |
-
"eval_runtime": 50.886,
|
| 697 |
-
"eval_samples_per_second": 47.636,
|
| 698 |
-
"eval_steps_per_second": 5.954,
|
| 699 |
-
"step": 10000
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 53.33,
|
| 703 |
-
"learning_rate": 5.489417989417989e-06,
|
| 704 |
-
"loss": 0.0171,
|
| 705 |
-
"step": 10100
|
| 706 |
-
},
|
| 707 |
-
{
|
| 708 |
-
"epoch": 53.86,
|
| 709 |
-
"learning_rate": 5.048500881834215e-06,
|
| 710 |
-
"loss": 0.0169,
|
| 711 |
-
"step": 10200
|
| 712 |
-
},
|
| 713 |
-
{
|
| 714 |
-
"epoch": 54.39,
|
| 715 |
-
"learning_rate": 4.611992945326279e-06,
|
| 716 |
-
"loss": 0.0136,
|
| 717 |
-
"step": 10300
|
| 718 |
-
},
|
| 719 |
-
{
|
| 720 |
-
"epoch": 54.92,
|
| 721 |
-
"learning_rate": 4.171075837742505e-06,
|
| 722 |
-
"loss": 0.0152,
|
| 723 |
-
"step": 10400
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 55.45,
|
| 727 |
-
"learning_rate": 3.7301587301587305e-06,
|
| 728 |
-
"loss": 0.015,
|
| 729 |
-
"step": 10500
|
| 730 |
-
},
|
| 731 |
-
{
|
| 732 |
-
"epoch": 55.97,
|
| 733 |
-
"learning_rate": 3.289241622574956e-06,
|
| 734 |
-
"loss": 0.0136,
|
| 735 |
-
"step": 10600
|
| 736 |
-
},
|
| 737 |
-
{
|
| 738 |
-
"epoch": 56.5,
|
| 739 |
-
"learning_rate": 2.848324514991182e-06,
|
| 740 |
-
"loss": 0.0156,
|
| 741 |
-
"step": 10700
|
| 742 |
-
},
|
| 743 |
-
{
|
| 744 |
-
"epoch": 57.03,
|
| 745 |
-
"learning_rate": 2.4074074074074075e-06,
|
| 746 |
-
"loss": 0.0137,
|
| 747 |
-
"step": 10800
|
| 748 |
-
},
|
| 749 |
-
{
|
| 750 |
-
"epoch": 57.56,
|
| 751 |
-
"learning_rate": 1.9664902998236335e-06,
|
| 752 |
-
"loss": 0.0147,
|
| 753 |
-
"step": 10900
|
| 754 |
-
},
|
| 755 |
-
{
|
| 756 |
-
"epoch": 58.09,
|
| 757 |
-
"learning_rate": 1.525573192239859e-06,
|
| 758 |
-
"loss": 0.0174,
|
| 759 |
-
"step": 11000
|
| 760 |
-
},
|
| 761 |
-
{
|
| 762 |
-
"epoch": 58.09,
|
| 763 |
-
"eval_accuracy": 0.933993399143219,
|
| 764 |
-
"eval_loss": 0.30968689918518066,
|
| 765 |
-
"eval_runtime": 51.0931,
|
| 766 |
-
"eval_samples_per_second": 47.443,
|
| 767 |
-
"eval_steps_per_second": 5.93,
|
| 768 |
-
"step": 11000
|
| 769 |
-
},
|
| 770 |
-
{
|
| 771 |
-
"epoch": 58.61,
|
| 772 |
-
"learning_rate": 1.0846560846560847e-06,
|
| 773 |
-
"loss": 0.0163,
|
| 774 |
-
"step": 11100
|
| 775 |
-
},
|
| 776 |
-
{
|
| 777 |
-
"epoch": 59.14,
|
| 778 |
-
"learning_rate": 6.437389770723105e-07,
|
| 779 |
-
"loss": 0.0096,
|
| 780 |
-
"step": 11200
|
| 781 |
-
},
|
| 782 |
-
{
|
| 783 |
-
"epoch": 59.67,
|
| 784 |
-
"learning_rate": 2.0282186948853617e-07,
|
| 785 |
-
"loss": 0.0121,
|
| 786 |
-
"step": 11300
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 59.88,
|
| 790 |
-
"step": 11340,
|
| 791 |
-
"total_flos": 2.20514636224512e+19,
|
| 792 |
-
"train_loss": 0.030625741817122836,
|
| 793 |
-
"train_runtime": 25244.9182,
|
| 794 |
-
"train_samples_per_second": 28.806,
|
| 795 |
-
"train_steps_per_second": 0.449
|
| 796 |
-
},
|
| 797 |
-
{
|
| 798 |
-
"epoch": 59.88,
|
| 799 |
-
"eval_accuracy": 0.9323432445526123,
|
| 800 |
-
"eval_loss": 0.3117374777793884,
|
| 801 |
-
"eval_runtime": 50.8983,
|
| 802 |
-
"eval_samples_per_second": 47.624,
|
| 803 |
-
"eval_steps_per_second": 5.953,
|
| 804 |
-
"step": 11340
|
| 805 |
}
|
| 806 |
],
|
| 807 |
"logging_steps": 100,
|
| 808 |
-
"max_steps":
|
| 809 |
-
"num_train_epochs":
|
| 810 |
"save_steps": 2000,
|
| 811 |
-
"total_flos":
|
| 812 |
"trial_name": null,
|
| 813 |
"trial_params": null
|
| 814 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 99.2084432717678,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 9400,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 1.06,
|
| 13 |
+
"learning_rate": 4.946808510638298e-05,
|
| 14 |
+
"loss": 0.0411,
|
| 15 |
"step": 100
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 2.11,
|
| 19 |
+
"learning_rate": 4.893617021276596e-05,
|
| 20 |
+
"loss": 0.033,
|
| 21 |
"step": 200
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 3.17,
|
| 25 |
+
"learning_rate": 4.840425531914894e-05,
|
| 26 |
+
"loss": 0.0286,
|
| 27 |
"step": 300
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 4.22,
|
| 31 |
+
"learning_rate": 4.787234042553192e-05,
|
| 32 |
+
"loss": 0.0371,
|
| 33 |
"step": 400
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 5.28,
|
| 37 |
+
"learning_rate": 4.734042553191489e-05,
|
| 38 |
+
"loss": 0.0315,
|
| 39 |
"step": 500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 6.33,
|
| 43 |
+
"learning_rate": 4.680851063829788e-05,
|
| 44 |
+
"loss": 0.0245,
|
| 45 |
"step": 600
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 7.39,
|
| 49 |
+
"learning_rate": 4.627659574468085e-05,
|
| 50 |
+
"loss": 0.0301,
|
| 51 |
"step": 700
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 8.44,
|
| 55 |
+
"learning_rate": 4.574468085106383e-05,
|
| 56 |
+
"loss": 0.0276,
|
| 57 |
"step": 800
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 9.5,
|
| 61 |
+
"learning_rate": 4.5212765957446815e-05,
|
| 62 |
+
"loss": 0.027,
|
| 63 |
"step": 900
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 10.55,
|
| 67 |
+
"learning_rate": 4.468085106382979e-05,
|
| 68 |
+
"loss": 0.0272,
|
| 69 |
"step": 1000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 10.55,
|
| 73 |
+
"eval_accuracy": 0.9257425665855408,
|
| 74 |
+
"eval_loss": 0.2915326654911041,
|
| 75 |
+
"eval_runtime": 50.3973,
|
| 76 |
+
"eval_samples_per_second": 48.098,
|
| 77 |
+
"eval_steps_per_second": 6.012,
|
| 78 |
"step": 1000
|
| 79 |
},
|
| 80 |
{
|
| 81 |
+
"epoch": 11.61,
|
| 82 |
+
"learning_rate": 4.414893617021277e-05,
|
| 83 |
+
"loss": 0.0235,
|
| 84 |
"step": 1100
|
| 85 |
},
|
| 86 |
{
|
| 87 |
+
"epoch": 12.66,
|
| 88 |
+
"learning_rate": 4.3617021276595746e-05,
|
| 89 |
+
"loss": 0.026,
|
| 90 |
"step": 1200
|
| 91 |
},
|
| 92 |
{
|
| 93 |
+
"epoch": 13.72,
|
| 94 |
+
"learning_rate": 4.3085106382978725e-05,
|
| 95 |
+
"loss": 0.0274,
|
| 96 |
"step": 1300
|
| 97 |
},
|
| 98 |
{
|
| 99 |
+
"epoch": 14.78,
|
| 100 |
+
"learning_rate": 4.2558510638297876e-05,
|
| 101 |
+
"loss": 0.0213,
|
| 102 |
"step": 1400
|
| 103 |
},
|
| 104 |
{
|
| 105 |
+
"epoch": 15.83,
|
| 106 |
+
"learning_rate": 4.2026595744680855e-05,
|
| 107 |
+
"loss": 0.0235,
|
| 108 |
"step": 1500
|
| 109 |
},
|
| 110 |
{
|
| 111 |
+
"epoch": 16.89,
|
| 112 |
+
"learning_rate": 4.1494680851063834e-05,
|
| 113 |
+
"loss": 0.0269,
|
| 114 |
"step": 1600
|
| 115 |
},
|
| 116 |
{
|
| 117 |
+
"epoch": 17.94,
|
| 118 |
+
"learning_rate": 4.096276595744681e-05,
|
| 119 |
+
"loss": 0.0271,
|
| 120 |
"step": 1700
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"epoch": 19.0,
|
| 124 |
+
"learning_rate": 4.0436170212765964e-05,
|
| 125 |
+
"loss": 0.0222,
|
| 126 |
"step": 1800
|
| 127 |
},
|
| 128 |
{
|
| 129 |
+
"epoch": 20.05,
|
| 130 |
+
"learning_rate": 3.990425531914894e-05,
|
| 131 |
+
"loss": 0.0213,
|
| 132 |
"step": 1900
|
| 133 |
},
|
| 134 |
{
|
| 135 |
+
"epoch": 21.11,
|
| 136 |
+
"learning_rate": 3.9372340425531916e-05,
|
| 137 |
+
"loss": 0.0172,
|
| 138 |
"step": 2000
|
| 139 |
},
|
| 140 |
{
|
| 141 |
+
"epoch": 21.11,
|
| 142 |
+
"eval_accuracy": 0.9331682920455933,
|
| 143 |
+
"eval_loss": 0.27688100934028625,
|
| 144 |
+
"eval_runtime": 51.2178,
|
| 145 |
+
"eval_samples_per_second": 47.327,
|
| 146 |
+
"eval_steps_per_second": 5.916,
|
| 147 |
"step": 2000
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 22.16,
|
| 151 |
+
"learning_rate": 3.8840425531914895e-05,
|
| 152 |
+
"loss": 0.0252,
|
| 153 |
"step": 2100
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 23.22,
|
| 157 |
+
"learning_rate": 3.8308510638297874e-05,
|
| 158 |
+
"loss": 0.0294,
|
| 159 |
"step": 2200
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 24.27,
|
| 163 |
+
"learning_rate": 3.7776595744680853e-05,
|
| 164 |
+
"loss": 0.0249,
|
| 165 |
"step": 2300
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 25.33,
|
| 169 |
+
"learning_rate": 3.7244680851063826e-05,
|
| 170 |
+
"loss": 0.0252,
|
| 171 |
"step": 2400
|
| 172 |
},
|
| 173 |
{
|
| 174 |
+
"epoch": 26.39,
|
| 175 |
+
"learning_rate": 3.671276595744681e-05,
|
| 176 |
+
"loss": 0.0203,
|
| 177 |
"step": 2500
|
| 178 |
},
|
| 179 |
{
|
| 180 |
+
"epoch": 27.44,
|
| 181 |
+
"learning_rate": 3.618085106382979e-05,
|
| 182 |
+
"loss": 0.0195,
|
| 183 |
"step": 2600
|
| 184 |
},
|
| 185 |
{
|
| 186 |
+
"epoch": 28.5,
|
| 187 |
+
"learning_rate": 3.5648936170212764e-05,
|
| 188 |
+
"loss": 0.0209,
|
| 189 |
"step": 2700
|
| 190 |
},
|
| 191 |
{
|
| 192 |
+
"epoch": 29.55,
|
| 193 |
+
"learning_rate": 3.511702127659575e-05,
|
| 194 |
+
"loss": 0.023,
|
| 195 |
"step": 2800
|
| 196 |
},
|
| 197 |
{
|
| 198 |
+
"epoch": 30.61,
|
| 199 |
+
"learning_rate": 3.458510638297873e-05,
|
| 200 |
+
"loss": 0.0173,
|
| 201 |
"step": 2900
|
| 202 |
},
|
| 203 |
{
|
| 204 |
+
"epoch": 31.66,
|
| 205 |
+
"learning_rate": 3.40531914893617e-05,
|
| 206 |
+
"loss": 0.0178,
|
| 207 |
"step": 3000
|
| 208 |
},
|
| 209 |
{
|
| 210 |
+
"epoch": 31.66,
|
| 211 |
+
"eval_accuracy": 0.9323432445526123,
|
| 212 |
+
"eval_loss": 0.29276978969573975,
|
| 213 |
+
"eval_runtime": 51.5058,
|
| 214 |
+
"eval_samples_per_second": 47.063,
|
| 215 |
+
"eval_steps_per_second": 5.883,
|
| 216 |
"step": 3000
|
| 217 |
},
|
| 218 |
{
|
| 219 |
+
"epoch": 32.72,
|
| 220 |
+
"learning_rate": 3.352127659574468e-05,
|
| 221 |
+
"loss": 0.0191,
|
| 222 |
"step": 3100
|
| 223 |
},
|
| 224 |
{
|
| 225 |
+
"epoch": 33.77,
|
| 226 |
+
"learning_rate": 3.2989361702127666e-05,
|
| 227 |
+
"loss": 0.0226,
|
| 228 |
"step": 3200
|
| 229 |
},
|
| 230 |
{
|
| 231 |
+
"epoch": 34.83,
|
| 232 |
+
"learning_rate": 3.245744680851064e-05,
|
| 233 |
+
"loss": 0.0197,
|
| 234 |
"step": 3300
|
| 235 |
},
|
| 236 |
{
|
| 237 |
+
"epoch": 35.88,
|
| 238 |
+
"learning_rate": 3.192553191489362e-05,
|
| 239 |
+
"loss": 0.0246,
|
| 240 |
"step": 3400
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 36.94,
|
| 244 |
+
"learning_rate": 3.13936170212766e-05,
|
| 245 |
+
"loss": 0.0183,
|
| 246 |
"step": 3500
|
| 247 |
},
|
| 248 |
{
|
| 249 |
+
"epoch": 37.99,
|
| 250 |
+
"learning_rate": 3.0861702127659576e-05,
|
| 251 |
+
"loss": 0.0161,
|
| 252 |
"step": 3600
|
| 253 |
},
|
| 254 |
{
|
| 255 |
+
"epoch": 39.05,
|
| 256 |
+
"learning_rate": 3.0329787234042556e-05,
|
| 257 |
+
"loss": 0.0156,
|
| 258 |
"step": 3700
|
| 259 |
},
|
| 260 |
{
|
| 261 |
+
"epoch": 40.11,
|
| 262 |
+
"learning_rate": 2.979787234042553e-05,
|
| 263 |
+
"loss": 0.014,
|
| 264 |
"step": 3800
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"epoch": 41.16,
|
| 268 |
+
"learning_rate": 2.926595744680851e-05,
|
| 269 |
+
"loss": 0.0138,
|
| 270 |
"step": 3900
|
| 271 |
},
|
| 272 |
{
|
| 273 |
+
"epoch": 42.22,
|
| 274 |
+
"learning_rate": 2.8734042553191493e-05,
|
| 275 |
+
"loss": 0.0145,
|
| 276 |
"step": 4000
|
| 277 |
},
|
| 278 |
{
|
| 279 |
+
"epoch": 42.22,
|
| 280 |
+
"eval_accuracy": 0.9356435537338257,
|
| 281 |
+
"eval_loss": 0.27180883288383484,
|
| 282 |
+
"eval_runtime": 50.6123,
|
| 283 |
+
"eval_samples_per_second": 47.894,
|
| 284 |
+
"eval_steps_per_second": 5.987,
|
| 285 |
"step": 4000
|
| 286 |
},
|
| 287 |
{
|
| 288 |
+
"epoch": 43.27,
|
| 289 |
+
"learning_rate": 2.820212765957447e-05,
|
| 290 |
+
"loss": 0.0176,
|
| 291 |
"step": 4100
|
| 292 |
},
|
| 293 |
{
|
| 294 |
+
"epoch": 44.33,
|
| 295 |
+
"learning_rate": 2.7670212765957448e-05,
|
| 296 |
+
"loss": 0.0135,
|
| 297 |
"step": 4200
|
| 298 |
},
|
| 299 |
{
|
| 300 |
+
"epoch": 45.38,
|
| 301 |
+
"learning_rate": 2.713829787234043e-05,
|
| 302 |
+
"loss": 0.0179,
|
| 303 |
"step": 4300
|
| 304 |
},
|
| 305 |
{
|
| 306 |
+
"epoch": 46.44,
|
| 307 |
+
"learning_rate": 2.6606382978723403e-05,
|
| 308 |
+
"loss": 0.0164,
|
| 309 |
"step": 4400
|
| 310 |
},
|
| 311 |
{
|
| 312 |
+
"epoch": 47.49,
|
| 313 |
+
"learning_rate": 2.6074468085106386e-05,
|
| 314 |
+
"loss": 0.019,
|
| 315 |
"step": 4500
|
| 316 |
},
|
| 317 |
{
|
| 318 |
+
"epoch": 48.55,
|
| 319 |
+
"learning_rate": 2.5542553191489365e-05,
|
| 320 |
+
"loss": 0.0136,
|
| 321 |
"step": 4600
|
| 322 |
},
|
| 323 |
{
|
| 324 |
+
"epoch": 49.6,
|
| 325 |
+
"learning_rate": 2.501063829787234e-05,
|
| 326 |
+
"loss": 0.0159,
|
| 327 |
"step": 4700
|
| 328 |
},
|
| 329 |
{
|
| 330 |
+
"epoch": 50.66,
|
| 331 |
+
"learning_rate": 2.447872340425532e-05,
|
| 332 |
+
"loss": 0.0163,
|
| 333 |
"step": 4800
|
| 334 |
},
|
| 335 |
{
|
| 336 |
+
"epoch": 51.72,
|
| 337 |
+
"learning_rate": 2.39468085106383e-05,
|
| 338 |
+
"loss": 0.0148,
|
| 339 |
"step": 4900
|
| 340 |
},
|
| 341 |
{
|
| 342 |
+
"epoch": 52.77,
|
| 343 |
+
"learning_rate": 2.341489361702128e-05,
|
| 344 |
+
"loss": 0.0147,
|
| 345 |
"step": 5000
|
| 346 |
},
|
| 347 |
{
|
| 348 |
+
"epoch": 52.77,
|
| 349 |
+
"eval_accuracy": 0.9348185062408447,
|
| 350 |
+
"eval_loss": 0.26764675974845886,
|
| 351 |
+
"eval_runtime": 49.7336,
|
| 352 |
+
"eval_samples_per_second": 48.74,
|
| 353 |
+
"eval_steps_per_second": 6.092,
|
| 354 |
"step": 5000
|
| 355 |
},
|
| 356 |
{
|
| 357 |
+
"epoch": 53.83,
|
| 358 |
+
"learning_rate": 2.2882978723404254e-05,
|
| 359 |
+
"loss": 0.0127,
|
| 360 |
"step": 5100
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"epoch": 54.88,
|
| 364 |
+
"learning_rate": 2.2351063829787237e-05,
|
| 365 |
+
"loss": 0.014,
|
| 366 |
"step": 5200
|
| 367 |
},
|
| 368 |
{
|
| 369 |
+
"epoch": 55.94,
|
| 370 |
+
"learning_rate": 2.1819148936170213e-05,
|
| 371 |
+
"loss": 0.0142,
|
| 372 |
"step": 5300
|
| 373 |
},
|
| 374 |
{
|
| 375 |
+
"epoch": 56.99,
|
| 376 |
+
"learning_rate": 2.1287234042553192e-05,
|
| 377 |
+
"loss": 0.016,
|
| 378 |
"step": 5400
|
| 379 |
},
|
| 380 |
{
|
| 381 |
+
"epoch": 58.05,
|
| 382 |
+
"learning_rate": 2.075531914893617e-05,
|
| 383 |
+
"loss": 0.0148,
|
| 384 |
"step": 5500
|
| 385 |
},
|
| 386 |
{
|
| 387 |
+
"epoch": 59.1,
|
| 388 |
+
"learning_rate": 2.022340425531915e-05,
|
| 389 |
+
"loss": 0.0125,
|
| 390 |
"step": 5600
|
| 391 |
},
|
| 392 |
{
|
| 393 |
+
"epoch": 60.16,
|
| 394 |
+
"learning_rate": 1.969148936170213e-05,
|
| 395 |
+
"loss": 0.0123,
|
| 396 |
"step": 5700
|
| 397 |
},
|
| 398 |
{
|
| 399 |
+
"epoch": 61.21,
|
| 400 |
+
"learning_rate": 1.9159574468085105e-05,
|
| 401 |
+
"loss": 0.0153,
|
| 402 |
"step": 5800
|
| 403 |
},
|
| 404 |
{
|
| 405 |
+
"epoch": 62.27,
|
| 406 |
+
"learning_rate": 1.8627659574468088e-05,
|
| 407 |
+
"loss": 0.0105,
|
| 408 |
"step": 5900
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"epoch": 63.32,
|
| 412 |
+
"learning_rate": 1.8095744680851064e-05,
|
| 413 |
+
"loss": 0.0135,
|
| 414 |
"step": 6000
|
| 415 |
},
|
| 416 |
{
|
| 417 |
+
"epoch": 63.32,
|
| 418 |
+
"eval_accuracy": 0.9397689700126648,
|
| 419 |
+
"eval_loss": 0.27310383319854736,
|
| 420 |
+
"eval_runtime": 49.9787,
|
| 421 |
+
"eval_samples_per_second": 48.501,
|
| 422 |
+
"eval_steps_per_second": 6.063,
|
| 423 |
"step": 6000
|
| 424 |
},
|
| 425 |
{
|
| 426 |
+
"epoch": 64.38,
|
| 427 |
+
"learning_rate": 1.7563829787234043e-05,
|
| 428 |
+
"loss": 0.0126,
|
| 429 |
"step": 6100
|
| 430 |
},
|
| 431 |
{
|
| 432 |
+
"epoch": 65.44,
|
| 433 |
+
"learning_rate": 1.7031914893617022e-05,
|
| 434 |
+
"loss": 0.0141,
|
| 435 |
"step": 6200
|
| 436 |
},
|
| 437 |
{
|
| 438 |
+
"epoch": 66.49,
|
| 439 |
+
"learning_rate": 1.6505319148936173e-05,
|
| 440 |
+
"loss": 0.0093,
|
| 441 |
"step": 6300
|
| 442 |
},
|
| 443 |
{
|
| 444 |
+
"epoch": 67.55,
|
| 445 |
+
"learning_rate": 1.597340425531915e-05,
|
| 446 |
+
"loss": 0.0109,
|
| 447 |
"step": 6400
|
| 448 |
},
|
| 449 |
{
|
| 450 |
+
"epoch": 68.6,
|
| 451 |
+
"learning_rate": 1.5441489361702128e-05,
|
| 452 |
+
"loss": 0.011,
|
| 453 |
"step": 6500
|
| 454 |
},
|
| 455 |
{
|
| 456 |
+
"epoch": 69.66,
|
| 457 |
+
"learning_rate": 1.4909574468085109e-05,
|
| 458 |
+
"loss": 0.0127,
|
| 459 |
"step": 6600
|
| 460 |
},
|
| 461 |
{
|
| 462 |
+
"epoch": 70.71,
|
| 463 |
+
"learning_rate": 1.4377659574468086e-05,
|
| 464 |
+
"loss": 0.009,
|
| 465 |
"step": 6700
|
| 466 |
},
|
| 467 |
{
|
| 468 |
+
"epoch": 71.77,
|
| 469 |
+
"learning_rate": 1.3845744680851064e-05,
|
| 470 |
+
"loss": 0.0109,
|
| 471 |
"step": 6800
|
| 472 |
},
|
| 473 |
{
|
| 474 |
+
"epoch": 72.82,
|
| 475 |
+
"learning_rate": 1.3313829787234045e-05,
|
| 476 |
+
"loss": 0.0114,
|
| 477 |
"step": 6900
|
| 478 |
},
|
| 479 |
{
|
| 480 |
+
"epoch": 73.88,
|
| 481 |
+
"learning_rate": 1.2781914893617022e-05,
|
| 482 |
+
"loss": 0.0105,
|
| 483 |
"step": 7000
|
| 484 |
},
|
| 485 |
{
|
| 486 |
+
"epoch": 73.88,
|
| 487 |
+
"eval_accuracy": 0.9335808753967285,
|
| 488 |
+
"eval_loss": 0.314525306224823,
|
| 489 |
+
"eval_runtime": 50.1755,
|
| 490 |
+
"eval_samples_per_second": 48.31,
|
| 491 |
+
"eval_steps_per_second": 6.039,
|
| 492 |
"step": 7000
|
| 493 |
},
|
| 494 |
{
|
| 495 |
+
"epoch": 74.93,
|
| 496 |
+
"learning_rate": 1.225e-05,
|
| 497 |
+
"loss": 0.0124,
|
| 498 |
"step": 7100
|
| 499 |
},
|
| 500 |
{
|
| 501 |
+
"epoch": 75.99,
|
| 502 |
+
"learning_rate": 1.171808510638298e-05,
|
| 503 |
+
"loss": 0.0091,
|
| 504 |
"step": 7200
|
| 505 |
},
|
| 506 |
{
|
| 507 |
+
"epoch": 77.04,
|
| 508 |
+
"learning_rate": 1.1186170212765958e-05,
|
| 509 |
+
"loss": 0.0118,
|
| 510 |
"step": 7300
|
| 511 |
},
|
| 512 |
{
|
| 513 |
+
"epoch": 78.1,
|
| 514 |
+
"learning_rate": 1.0654255319148937e-05,
|
| 515 |
+
"loss": 0.0098,
|
| 516 |
"step": 7400
|
| 517 |
},
|
| 518 |
{
|
| 519 |
+
"epoch": 79.16,
|
| 520 |
+
"learning_rate": 1.0122340425531915e-05,
|
| 521 |
+
"loss": 0.01,
|
| 522 |
"step": 7500
|
| 523 |
},
|
| 524 |
{
|
| 525 |
+
"epoch": 80.21,
|
| 526 |
+
"learning_rate": 9.590425531914894e-06,
|
| 527 |
+
"loss": 0.0103,
|
| 528 |
"step": 7600
|
| 529 |
},
|
| 530 |
{
|
| 531 |
+
"epoch": 81.27,
|
| 532 |
+
"learning_rate": 9.058510638297871e-06,
|
| 533 |
+
"loss": 0.0071,
|
| 534 |
"step": 7700
|
| 535 |
},
|
| 536 |
{
|
| 537 |
+
"epoch": 82.32,
|
| 538 |
+
"learning_rate": 8.52659574468085e-06,
|
| 539 |
+
"loss": 0.0114,
|
| 540 |
"step": 7800
|
| 541 |
},
|
| 542 |
{
|
| 543 |
+
"epoch": 83.38,
|
| 544 |
+
"learning_rate": 7.994680851063832e-06,
|
| 545 |
+
"loss": 0.0094,
|
| 546 |
"step": 7900
|
| 547 |
},
|
| 548 |
{
|
| 549 |
+
"epoch": 84.43,
|
| 550 |
+
"learning_rate": 7.462765957446809e-06,
|
| 551 |
+
"loss": 0.0075,
|
| 552 |
"step": 8000
|
| 553 |
},
|
| 554 |
{
|
| 555 |
+
"epoch": 84.43,
|
| 556 |
+
"eval_accuracy": 0.9319307208061218,
|
| 557 |
+
"eval_loss": 0.29711028933525085,
|
| 558 |
+
"eval_runtime": 50.3531,
|
| 559 |
+
"eval_samples_per_second": 48.14,
|
| 560 |
+
"eval_steps_per_second": 6.018,
|
| 561 |
"step": 8000
|
| 562 |
},
|
| 563 |
{
|
| 564 |
+
"epoch": 85.49,
|
| 565 |
+
"learning_rate": 6.930851063829788e-06,
|
| 566 |
+
"loss": 0.0089,
|
| 567 |
"step": 8100
|
| 568 |
},
|
| 569 |
{
|
| 570 |
+
"epoch": 86.54,
|
| 571 |
+
"learning_rate": 6.398936170212766e-06,
|
| 572 |
+
"loss": 0.0096,
|
| 573 |
"step": 8200
|
| 574 |
},
|
| 575 |
{
|
| 576 |
+
"epoch": 87.6,
|
| 577 |
+
"learning_rate": 5.867021276595745e-06,
|
| 578 |
+
"loss": 0.0095,
|
| 579 |
"step": 8300
|
| 580 |
},
|
| 581 |
{
|
| 582 |
+
"epoch": 88.65,
|
| 583 |
+
"learning_rate": 5.335106382978724e-06,
|
| 584 |
+
"loss": 0.0143,
|
| 585 |
"step": 8400
|
| 586 |
},
|
| 587 |
{
|
| 588 |
+
"epoch": 89.71,
|
| 589 |
+
"learning_rate": 4.8031914893617025e-06,
|
| 590 |
+
"loss": 0.0097,
|
| 591 |
"step": 8500
|
| 592 |
},
|
| 593 |
{
|
| 594 |
+
"epoch": 90.77,
|
| 595 |
+
"learning_rate": 4.271276595744681e-06,
|
| 596 |
+
"loss": 0.0094,
|
| 597 |
"step": 8600
|
| 598 |
},
|
| 599 |
{
|
| 600 |
+
"epoch": 91.82,
|
| 601 |
+
"learning_rate": 3.7393617021276596e-06,
|
| 602 |
+
"loss": 0.0088,
|
| 603 |
"step": 8700
|
| 604 |
},
|
| 605 |
{
|
| 606 |
+
"epoch": 92.88,
|
| 607 |
+
"learning_rate": 3.2074468085106384e-06,
|
| 608 |
+
"loss": 0.0052,
|
| 609 |
"step": 8800
|
| 610 |
},
|
| 611 |
{
|
| 612 |
+
"epoch": 93.93,
|
| 613 |
+
"learning_rate": 2.6755319148936168e-06,
|
| 614 |
+
"loss": 0.0102,
|
| 615 |
"step": 8900
|
| 616 |
},
|
| 617 |
{
|
| 618 |
+
"epoch": 94.99,
|
| 619 |
+
"learning_rate": 2.143617021276596e-06,
|
| 620 |
+
"loss": 0.0078,
|
| 621 |
"step": 9000
|
| 622 |
},
|
| 623 |
{
|
| 624 |
+
"epoch": 94.99,
|
| 625 |
+
"eval_accuracy": 0.9327557682991028,
|
| 626 |
+
"eval_loss": 0.2949957847595215,
|
| 627 |
+
"eval_runtime": 51.5836,
|
| 628 |
+
"eval_samples_per_second": 46.992,
|
| 629 |
+
"eval_steps_per_second": 5.874,
|
| 630 |
"step": 9000
|
| 631 |
},
|
| 632 |
{
|
| 633 |
+
"epoch": 96.04,
|
| 634 |
+
"learning_rate": 1.6117021276595745e-06,
|
| 635 |
+
"loss": 0.0066,
|
| 636 |
"step": 9100
|
| 637 |
},
|
| 638 |
{
|
| 639 |
+
"epoch": 97.1,
|
| 640 |
+
"learning_rate": 1.0797872340425531e-06,
|
| 641 |
+
"loss": 0.0081,
|
| 642 |
"step": 9200
|
| 643 |
},
|
| 644 |
{
|
| 645 |
+
"epoch": 98.15,
|
| 646 |
+
"learning_rate": 5.478723404255319e-07,
|
| 647 |
+
"loss": 0.0094,
|
| 648 |
"step": 9300
|
| 649 |
},
|
| 650 |
{
|
| 651 |
+
"epoch": 99.21,
|
| 652 |
+
"learning_rate": 1.595744680851064e-08,
|
| 653 |
+
"loss": 0.0095,
|
| 654 |
"step": 9400
|
| 655 |
},
|
| 656 |
{
|
| 657 |
+
"epoch": 99.21,
|
| 658 |
+
"step": 9400,
|
| 659 |
+
"total_flos": 3.653391792237703e+19,
|
| 660 |
+
"train_loss": 0.01680715578667661,
|
| 661 |
+
"train_runtime": 36294.494,
|
| 662 |
+
"train_samples_per_second": 33.393,
|
| 663 |
+
"train_steps_per_second": 0.259
|
| 664 |
},
|
| 665 |
{
|
| 666 |
+
"epoch": 99.21,
|
| 667 |
+
"eval_accuracy": 0.9319307208061218,
|
| 668 |
+
"eval_loss": 0.29647526144981384,
|
| 669 |
+
"eval_runtime": 50.5435,
|
| 670 |
+
"eval_samples_per_second": 47.959,
|
| 671 |
+
"eval_steps_per_second": 5.995,
|
| 672 |
+
"step": 9400
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 673 |
}
|
| 674 |
],
|
| 675 |
"logging_steps": 100,
|
| 676 |
+
"max_steps": 9400,
|
| 677 |
+
"num_train_epochs": 100,
|
| 678 |
"save_steps": 2000,
|
| 679 |
+
"total_flos": 3.653391792237703e+19,
|
| 680 |
"trial_name": null,
|
| 681 |
"trial_params": null
|
| 682 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4155
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26c98e7aa7407b828227bd572ea7475f0910a407b91381f0d71f441503edc41e
|
| 3 |
size 4155
|