End of training
Browse files- README.md +11 -13
- all_results.json +6 -6
- model.safetensors +1 -1
- runs/Jun15_14-11-05_9fe18ef6c7ec/events.out.tfevents.1718460680.9fe18ef6c7ec.34.0 +3 -0
- train_results.json +6 -6
- trainer_state.json +205 -279
- 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 [facebook/hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
-
- Loss: 0.
|
| 21 |
-
- Accuracy: 0.
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
@@ -38,8 +38,8 @@ More information needed
|
|
| 38 |
|
| 39 |
The following hyperparameters were used during training:
|
| 40 |
- learning_rate: 1e-05
|
| 41 |
-
- train_batch_size:
|
| 42 |
-
- eval_batch_size:
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
|
@@ -50,15 +50,13 @@ The following hyperparameters were used during training:
|
|
| 50 |
|
| 51 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 52 |
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
| 53 |
-
|
|
| 54 |
-
| 0.
|
| 55 |
-
| 0.
|
| 56 |
-
| 0.
|
| 57 |
-
| 0.
|
| 58 |
-
| 0.
|
| 59 |
-
| 0.
|
| 60 |
-
| 0.3086 | 1.3311 | 4000 | 0.4347 | 0.9275 |
|
| 61 |
-
| 0.1531 | 1.4975 | 4500 | 0.4469 | 0.9260 |
|
| 62 |
|
| 63 |
|
| 64 |
### Framework versions
|
|
|
|
| 17 |
|
| 18 |
This model is a fine-tuned version of [facebook/hubert-large-ls960-ft](https://huggingface.co/facebook/hubert-large-ls960-ft) on the None dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.3169
|
| 21 |
+
- Accuracy: 0.9395
|
| 22 |
|
| 23 |
## Model description
|
| 24 |
|
|
|
|
| 38 |
|
| 39 |
The following hyperparameters were used during training:
|
| 40 |
- learning_rate: 1e-05
|
| 41 |
+
- train_batch_size: 16
|
| 42 |
+
- eval_batch_size: 8
|
| 43 |
- seed: 42
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
|
|
|
| 50 |
|
| 51 |
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 52 |
|:-------------:|:------:|:----:|:---------------:|:--------:|
|
| 53 |
+
| 0.5275 | 0.6649 | 500 | 0.4740 | 0.8483 |
|
| 54 |
+
| 0.2923 | 1.3298 | 1000 | 0.3727 | 0.8999 |
|
| 55 |
+
| 0.2168 | 1.9947 | 1500 | 0.4175 | 0.9073 |
|
| 56 |
+
| 0.1442 | 2.6596 | 2000 | 0.3471 | 0.9312 |
|
| 57 |
+
| 0.1341 | 3.3245 | 2500 | 0.3641 | 0.9275 |
|
| 58 |
+
| 0.1136 | 3.9894 | 3000 | 0.3286 | 0.9432 |
|
| 59 |
+
| 0.1102 | 4.6543 | 3500 | 0.3169 | 0.9395 |
|
|
|
|
|
|
|
| 60 |
|
| 61 |
|
| 62 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 12018,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second":
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 7.814523952420126e+18,
|
| 4 |
+
"train_loss": 0.29202749221882923,
|
| 5 |
+
"train_runtime": 8842.7154,
|
| 6 |
"train_samples": 12018,
|
| 7 |
+
"train_samples_per_second": 6.795,
|
| 8 |
+
"train_steps_per_second": 0.425
|
| 9 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1266025756
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f556ced5fb4077887ee371d074f7dbebe2d2b95d4386e3a614e1f69724268478
|
| 3 |
size 1266025756
|
runs/Jun15_14-11-05_9fe18ef6c7ec/events.out.tfevents.1718460680.9fe18ef6c7ec.34.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:147afdbad27934d690be17e3bbb6539effe150558c95e1c9c0c417bd0bc591c2
|
| 3 |
+
size 16477
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 12018,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second":
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 7.814523952420126e+18,
|
| 4 |
+
"train_loss": 0.29202749221882923,
|
| 5 |
+
"train_runtime": 8842.7154,
|
| 6 |
"train_samples": 12018,
|
| 7 |
+
"train_samples_per_second": 6.795,
|
| 8 |
+
"train_steps_per_second": 0.425
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,421 +1,347 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "/kaggle/working/hubert-agum960-amharic/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 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.
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 9.
|
| 15 |
-
"loss": 1.
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"epoch": 0.
|
| 20 |
-
"grad_norm":
|
| 21 |
-
"learning_rate": 9.
|
| 22 |
-
"loss": 1.
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
-
"epoch": 0.
|
| 27 |
-
"grad_norm":
|
| 28 |
-
"learning_rate": 9.
|
| 29 |
-
"loss": 1.
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
-
"epoch": 0.
|
| 34 |
-
"grad_norm":
|
| 35 |
-
"learning_rate":
|
| 36 |
-
"loss":
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
-
"epoch": 0.
|
| 41 |
-
"grad_norm":
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss":
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"eval_accuracy": 0.
|
| 49 |
-
"eval_loss": 0.
|
| 50 |
-
"eval_runtime":
|
| 51 |
-
"eval_samples_per_second":
|
| 52 |
-
"eval_steps_per_second":
|
| 53 |
"step": 500
|
| 54 |
},
|
| 55 |
{
|
| 56 |
-
"epoch": 0.
|
| 57 |
-
"grad_norm":
|
| 58 |
-
"learning_rate":
|
| 59 |
-
"loss": 0.
|
| 60 |
"step": 600
|
| 61 |
},
|
| 62 |
{
|
| 63 |
-
"epoch": 0.
|
| 64 |
-
"grad_norm":
|
| 65 |
-
"learning_rate":
|
| 66 |
-
"loss": 0.
|
| 67 |
"step": 700
|
| 68 |
},
|
| 69 |
{
|
| 70 |
-
"epoch":
|
| 71 |
-
"grad_norm":
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss": 0.
|
| 74 |
"step": 800
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch":
|
| 78 |
-
"grad_norm":
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 900
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
-
"grad_norm":
|
| 86 |
-
"learning_rate":
|
| 87 |
-
"loss": 0.
|
| 88 |
"step": 1000
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"epoch":
|
| 92 |
-
"eval_accuracy": 0.
|
| 93 |
-
"eval_loss": 0.
|
| 94 |
-
"eval_runtime": 121.
|
| 95 |
-
"eval_samples_per_second":
|
| 96 |
-
"eval_steps_per_second":
|
| 97 |
"step": 1000
|
| 98 |
},
|
| 99 |
{
|
| 100 |
-
"epoch":
|
| 101 |
-
"grad_norm":
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 0.
|
| 104 |
"step": 1100
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"grad_norm":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 1200
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"grad_norm":
|
| 116 |
-
"learning_rate":
|
| 117 |
-
"loss": 0.
|
| 118 |
"step": 1300
|
| 119 |
},
|
| 120 |
{
|
| 121 |
-
"epoch":
|
| 122 |
-
"grad_norm":
|
| 123 |
-
"learning_rate":
|
| 124 |
-
"loss": 0.
|
| 125 |
"step": 1400
|
| 126 |
},
|
| 127 |
{
|
| 128 |
-
"epoch":
|
| 129 |
-
"grad_norm":
|
| 130 |
-
"learning_rate":
|
| 131 |
-
"loss": 0.
|
| 132 |
"step": 1500
|
| 133 |
},
|
| 134 |
{
|
| 135 |
-
"epoch":
|
| 136 |
-
"eval_accuracy": 0.
|
| 137 |
-
"eval_loss": 0.
|
| 138 |
-
"eval_runtime":
|
| 139 |
-
"eval_samples_per_second":
|
| 140 |
-
"eval_steps_per_second":
|
| 141 |
"step": 1500
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"grad_norm":
|
| 146 |
-
"learning_rate":
|
| 147 |
-
"loss": 0.
|
| 148 |
"step": 1600
|
| 149 |
},
|
| 150 |
{
|
| 151 |
-
"epoch":
|
| 152 |
-
"grad_norm":
|
| 153 |
-
"learning_rate":
|
| 154 |
-
"loss": 0.
|
| 155 |
"step": 1700
|
| 156 |
},
|
| 157 |
{
|
| 158 |
-
"epoch":
|
| 159 |
-
"grad_norm":
|
| 160 |
-
"learning_rate":
|
| 161 |
-
"loss": 0.
|
| 162 |
"step": 1800
|
| 163 |
},
|
| 164 |
{
|
| 165 |
-
"epoch":
|
| 166 |
-
"grad_norm":
|
| 167 |
-
"learning_rate":
|
| 168 |
-
"loss": 0.
|
| 169 |
"step": 1900
|
| 170 |
},
|
| 171 |
{
|
| 172 |
-
"epoch":
|
| 173 |
-
"grad_norm":
|
| 174 |
-
"learning_rate":
|
| 175 |
-
"loss": 0.
|
| 176 |
"step": 2000
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch":
|
| 180 |
-
"eval_accuracy": 0.
|
| 181 |
-
"eval_loss": 0.
|
| 182 |
-
"eval_runtime": 121.
|
| 183 |
-
"eval_samples_per_second":
|
| 184 |
-
"eval_steps_per_second":
|
| 185 |
"step": 2000
|
| 186 |
},
|
| 187 |
{
|
| 188 |
-
"epoch":
|
| 189 |
-
"grad_norm":
|
| 190 |
-
"learning_rate":
|
| 191 |
-
"loss": 0.
|
| 192 |
"step": 2100
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"epoch":
|
| 196 |
-
"grad_norm": 0.
|
| 197 |
-
"learning_rate":
|
| 198 |
-
"loss": 0.
|
| 199 |
"step": 2200
|
| 200 |
},
|
| 201 |
{
|
| 202 |
-
"epoch":
|
| 203 |
-
"grad_norm": 0.
|
| 204 |
-
"learning_rate":
|
| 205 |
-
"loss": 0.
|
| 206 |
"step": 2300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
-
"epoch":
|
| 210 |
-
"grad_norm":
|
| 211 |
-
"learning_rate":
|
| 212 |
-
"loss": 0.
|
| 213 |
"step": 2400
|
| 214 |
},
|
| 215 |
{
|
| 216 |
-
"epoch":
|
| 217 |
-
"grad_norm":
|
| 218 |
-
"learning_rate":
|
| 219 |
-
"loss": 0.
|
| 220 |
"step": 2500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
-
"epoch":
|
| 224 |
-
"eval_accuracy": 0.
|
| 225 |
-
"eval_loss": 0.
|
| 226 |
-
"eval_runtime":
|
| 227 |
-
"eval_samples_per_second":
|
| 228 |
-
"eval_steps_per_second":
|
| 229 |
"step": 2500
|
| 230 |
},
|
| 231 |
{
|
| 232 |
-
"epoch":
|
| 233 |
-
"grad_norm":
|
| 234 |
-
"learning_rate":
|
| 235 |
-
"loss": 0.
|
| 236 |
"step": 2600
|
| 237 |
},
|
| 238 |
{
|
| 239 |
-
"epoch":
|
| 240 |
-
"grad_norm":
|
| 241 |
-
"learning_rate":
|
| 242 |
-
"loss": 0.
|
| 243 |
"step": 2700
|
| 244 |
},
|
| 245 |
{
|
| 246 |
-
"epoch":
|
| 247 |
-
"grad_norm":
|
| 248 |
-
"learning_rate":
|
| 249 |
-
"loss": 0.
|
| 250 |
"step": 2800
|
| 251 |
},
|
| 252 |
{
|
| 253 |
-
"epoch":
|
| 254 |
-
"grad_norm":
|
| 255 |
-
"learning_rate":
|
| 256 |
-
"loss": 0.
|
| 257 |
"step": 2900
|
| 258 |
},
|
| 259 |
{
|
| 260 |
-
"epoch":
|
| 261 |
-
"grad_norm":
|
| 262 |
-
"learning_rate":
|
| 263 |
-
"loss": 0.
|
| 264 |
"step": 3000
|
| 265 |
},
|
| 266 |
{
|
| 267 |
-
"epoch":
|
| 268 |
-
"eval_accuracy": 0.
|
| 269 |
-
"eval_loss": 0.
|
| 270 |
-
"eval_runtime": 121.
|
| 271 |
-
"eval_samples_per_second":
|
| 272 |
-
"eval_steps_per_second":
|
| 273 |
"step": 3000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
-
"epoch":
|
| 277 |
-
"grad_norm":
|
| 278 |
-
"learning_rate":
|
| 279 |
-
"loss": 0.
|
| 280 |
"step": 3100
|
| 281 |
},
|
| 282 |
{
|
| 283 |
-
"epoch":
|
| 284 |
-
"grad_norm":
|
| 285 |
-
"learning_rate":
|
| 286 |
-
"loss": 0.
|
| 287 |
"step": 3200
|
| 288 |
},
|
| 289 |
{
|
| 290 |
-
"epoch":
|
| 291 |
-
"grad_norm":
|
| 292 |
-
"learning_rate":
|
| 293 |
-
"loss": 0.
|
| 294 |
"step": 3300
|
| 295 |
},
|
| 296 |
{
|
| 297 |
-
"epoch":
|
| 298 |
-
"grad_norm": 0.
|
| 299 |
-
"learning_rate":
|
| 300 |
-
"loss": 0.
|
| 301 |
"step": 3400
|
| 302 |
},
|
| 303 |
{
|
| 304 |
-
"epoch":
|
| 305 |
-
"grad_norm":
|
| 306 |
-
"learning_rate": 7.
|
| 307 |
-
"loss": 0.
|
| 308 |
"step": 3500
|
| 309 |
},
|
| 310 |
{
|
| 311 |
-
"epoch":
|
| 312 |
-
"eval_accuracy": 0.
|
| 313 |
-
"eval_loss": 0.
|
| 314 |
-
"eval_runtime":
|
| 315 |
-
"eval_samples_per_second":
|
| 316 |
-
"eval_steps_per_second":
|
| 317 |
"step": 3500
|
| 318 |
},
|
| 319 |
{
|
| 320 |
-
"epoch":
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
-
"learning_rate":
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 3600
|
| 325 |
},
|
| 326 |
{
|
| 327 |
-
"epoch":
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate":
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 3700
|
| 332 |
},
|
| 333 |
{
|
| 334 |
-
"epoch":
|
| 335 |
-
"
|
| 336 |
-
"
|
| 337 |
-
"
|
| 338 |
-
"
|
| 339 |
-
|
| 340 |
-
|
| 341 |
-
"epoch": 1.29783693843594,
|
| 342 |
-
"grad_norm": 2.320756673812866,
|
| 343 |
-
"learning_rate": 7.407653910149751e-06,
|
| 344 |
-
"loss": 0.1545,
|
| 345 |
-
"step": 3900
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 1.3311148086522462,
|
| 349 |
-
"grad_norm": 0.02829299308359623,
|
| 350 |
-
"learning_rate": 7.341098169717139e-06,
|
| 351 |
-
"loss": 0.3086,
|
| 352 |
-
"step": 4000
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 1.3311148086522462,
|
| 356 |
-
"eval_accuracy": 0.927503764629364,
|
| 357 |
-
"eval_loss": 0.43469786643981934,
|
| 358 |
-
"eval_runtime": 122.0442,
|
| 359 |
-
"eval_samples_per_second": 10.963,
|
| 360 |
-
"eval_steps_per_second": 2.745,
|
| 361 |
-
"step": 4000
|
| 362 |
-
},
|
| 363 |
-
{
|
| 364 |
-
"epoch": 1.3643926788685525,
|
| 365 |
-
"grad_norm": 0.563818097114563,
|
| 366 |
-
"learning_rate": 7.274542429284527e-06,
|
| 367 |
-
"loss": 0.203,
|
| 368 |
-
"step": 4100
|
| 369 |
-
},
|
| 370 |
-
{
|
| 371 |
-
"epoch": 1.3976705490848587,
|
| 372 |
-
"grad_norm": 0.032793425023555756,
|
| 373 |
-
"learning_rate": 7.207986688851915e-06,
|
| 374 |
-
"loss": 0.2317,
|
| 375 |
-
"step": 4200
|
| 376 |
-
},
|
| 377 |
-
{
|
| 378 |
-
"epoch": 1.4309484193011648,
|
| 379 |
-
"grad_norm": 13.284598350524902,
|
| 380 |
-
"learning_rate": 7.1414309484193024e-06,
|
| 381 |
-
"loss": 0.1721,
|
| 382 |
-
"step": 4300
|
| 383 |
-
},
|
| 384 |
-
{
|
| 385 |
-
"epoch": 1.464226289517471,
|
| 386 |
-
"grad_norm": 7.5517897605896,
|
| 387 |
-
"learning_rate": 7.0748752079866885e-06,
|
| 388 |
-
"loss": 0.2042,
|
| 389 |
-
"step": 4400
|
| 390 |
-
},
|
| 391 |
-
{
|
| 392 |
-
"epoch": 1.497504159733777,
|
| 393 |
-
"grad_norm": 0.04595191031694412,
|
| 394 |
-
"learning_rate": 7.008319467554076e-06,
|
| 395 |
-
"loss": 0.1531,
|
| 396 |
-
"step": 4500
|
| 397 |
-
},
|
| 398 |
-
{
|
| 399 |
-
"epoch": 1.497504159733777,
|
| 400 |
-
"eval_accuracy": 0.926008939743042,
|
| 401 |
-
"eval_loss": 0.4468592703342438,
|
| 402 |
-
"eval_runtime": 122.6586,
|
| 403 |
-
"eval_samples_per_second": 10.908,
|
| 404 |
-
"eval_steps_per_second": 2.731,
|
| 405 |
-
"step": 4500
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 1.497504159733777,
|
| 409 |
-
"step": 4500,
|
| 410 |
-
"total_flos": 2.060256870141264e+18,
|
| 411 |
-
"train_loss": 0.4611078107621935,
|
| 412 |
-
"train_runtime": 3912.8249,
|
| 413 |
-
"train_samples_per_second": 15.357,
|
| 414 |
-
"train_steps_per_second": 3.84
|
| 415 |
}
|
| 416 |
],
|
| 417 |
"logging_steps": 100,
|
| 418 |
-
"max_steps":
|
| 419 |
"num_input_tokens_seen": 0,
|
| 420 |
"num_train_epochs": 5,
|
| 421 |
"save_steps": 500,
|
|
@@ -435,13 +361,13 @@
|
|
| 435 |
"should_evaluate": false,
|
| 436 |
"should_log": false,
|
| 437 |
"should_save": true,
|
| 438 |
-
"should_training_stop":
|
| 439 |
},
|
| 440 |
"attributes": {}
|
| 441 |
}
|
| 442 |
},
|
| 443 |
-
"total_flos":
|
| 444 |
-
"train_batch_size":
|
| 445 |
"trial_name": null,
|
| 446 |
"trial_params": null
|
| 447 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.31690406799316406,
|
| 3 |
+
"best_model_checkpoint": "/kaggle/working/hubert-agum960-amharic/checkpoint-3500",
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3760,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.13297872340425532,
|
| 13 |
+
"grad_norm": 2.615074396133423,
|
| 14 |
+
"learning_rate": 9.73404255319149e-06,
|
| 15 |
+
"loss": 1.5743,
|
| 16 |
"step": 100
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"epoch": 0.26595744680851063,
|
| 20 |
+
"grad_norm": 2.4918744564056396,
|
| 21 |
+
"learning_rate": 9.46808510638298e-06,
|
| 22 |
+
"loss": 1.3461,
|
| 23 |
"step": 200
|
| 24 |
},
|
| 25 |
{
|
| 26 |
+
"epoch": 0.39893617021276595,
|
| 27 |
+
"grad_norm": 2.8701541423797607,
|
| 28 |
+
"learning_rate": 9.20212765957447e-06,
|
| 29 |
+
"loss": 1.0072,
|
| 30 |
"step": 300
|
| 31 |
},
|
| 32 |
{
|
| 33 |
+
"epoch": 0.5319148936170213,
|
| 34 |
+
"grad_norm": 12.351414680480957,
|
| 35 |
+
"learning_rate": 8.938829787234043e-06,
|
| 36 |
+
"loss": 0.7362,
|
| 37 |
"step": 400
|
| 38 |
},
|
| 39 |
{
|
| 40 |
+
"epoch": 0.6648936170212766,
|
| 41 |
+
"grad_norm": 8.005541801452637,
|
| 42 |
+
"learning_rate": 8.672872340425533e-06,
|
| 43 |
+
"loss": 0.5275,
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.6648936170212766,
|
| 48 |
+
"eval_accuracy": 0.8482810258865356,
|
| 49 |
+
"eval_loss": 0.4739724397659302,
|
| 50 |
+
"eval_runtime": 120.9505,
|
| 51 |
+
"eval_samples_per_second": 11.062,
|
| 52 |
+
"eval_steps_per_second": 1.389,
|
| 53 |
"step": 500
|
| 54 |
},
|
| 55 |
{
|
| 56 |
+
"epoch": 0.7978723404255319,
|
| 57 |
+
"grad_norm": 4.672068119049072,
|
| 58 |
+
"learning_rate": 8.406914893617022e-06,
|
| 59 |
+
"loss": 0.4446,
|
| 60 |
"step": 600
|
| 61 |
},
|
| 62 |
{
|
| 63 |
+
"epoch": 0.9308510638297872,
|
| 64 |
+
"grad_norm": 7.660475254058838,
|
| 65 |
+
"learning_rate": 8.140957446808512e-06,
|
| 66 |
+
"loss": 0.3683,
|
| 67 |
"step": 700
|
| 68 |
},
|
| 69 |
{
|
| 70 |
+
"epoch": 1.0638297872340425,
|
| 71 |
+
"grad_norm": 3.2801108360290527,
|
| 72 |
+
"learning_rate": 7.877659574468086e-06,
|
| 73 |
+
"loss": 0.3025,
|
| 74 |
"step": 800
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 1.196808510638298,
|
| 78 |
+
"grad_norm": 11.70166015625,
|
| 79 |
+
"learning_rate": 7.61436170212766e-06,
|
| 80 |
+
"loss": 0.2542,
|
| 81 |
"step": 900
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 1.3297872340425532,
|
| 85 |
+
"grad_norm": 10.051458358764648,
|
| 86 |
+
"learning_rate": 7.348404255319149e-06,
|
| 87 |
+
"loss": 0.2923,
|
| 88 |
"step": 1000
|
| 89 |
},
|
| 90 |
{
|
| 91 |
+
"epoch": 1.3297872340425532,
|
| 92 |
+
"eval_accuracy": 0.8998505473136902,
|
| 93 |
+
"eval_loss": 0.3727114200592041,
|
| 94 |
+
"eval_runtime": 121.4022,
|
| 95 |
+
"eval_samples_per_second": 11.021,
|
| 96 |
+
"eval_steps_per_second": 1.384,
|
| 97 |
"step": 1000
|
| 98 |
},
|
| 99 |
{
|
| 100 |
+
"epoch": 1.4627659574468086,
|
| 101 |
+
"grad_norm": 2.711136817932129,
|
| 102 |
+
"learning_rate": 7.0824468085106394e-06,
|
| 103 |
+
"loss": 0.2151,
|
| 104 |
"step": 1100
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 1.5957446808510638,
|
| 108 |
+
"grad_norm": 3.505892038345337,
|
| 109 |
+
"learning_rate": 6.816489361702127e-06,
|
| 110 |
+
"loss": 0.2303,
|
| 111 |
"step": 1200
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 1.728723404255319,
|
| 115 |
+
"grad_norm": 0.5282111763954163,
|
| 116 |
+
"learning_rate": 6.550531914893618e-06,
|
| 117 |
+
"loss": 0.2323,
|
| 118 |
"step": 1300
|
| 119 |
},
|
| 120 |
{
|
| 121 |
+
"epoch": 1.8617021276595744,
|
| 122 |
+
"grad_norm": 6.366661071777344,
|
| 123 |
+
"learning_rate": 6.284574468085107e-06,
|
| 124 |
+
"loss": 0.2029,
|
| 125 |
"step": 1400
|
| 126 |
},
|
| 127 |
{
|
| 128 |
+
"epoch": 1.9946808510638299,
|
| 129 |
+
"grad_norm": 17.721065521240234,
|
| 130 |
+
"learning_rate": 6.018617021276596e-06,
|
| 131 |
+
"loss": 0.2168,
|
| 132 |
"step": 1500
|
| 133 |
},
|
| 134 |
{
|
| 135 |
+
"epoch": 1.9946808510638299,
|
| 136 |
+
"eval_accuracy": 0.9073243737220764,
|
| 137 |
+
"eval_loss": 0.41751202940940857,
|
| 138 |
+
"eval_runtime": 120.4054,
|
| 139 |
+
"eval_samples_per_second": 11.112,
|
| 140 |
+
"eval_steps_per_second": 1.395,
|
| 141 |
"step": 1500
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 2.127659574468085,
|
| 145 |
+
"grad_norm": 29.569236755371094,
|
| 146 |
+
"learning_rate": 5.752659574468086e-06,
|
| 147 |
+
"loss": 0.181,
|
| 148 |
"step": 1600
|
| 149 |
},
|
| 150 |
{
|
| 151 |
+
"epoch": 2.2606382978723403,
|
| 152 |
+
"grad_norm": 3.7976977825164795,
|
| 153 |
+
"learning_rate": 5.4867021276595745e-06,
|
| 154 |
+
"loss": 0.1902,
|
| 155 |
"step": 1700
|
| 156 |
},
|
| 157 |
{
|
| 158 |
+
"epoch": 2.393617021276596,
|
| 159 |
+
"grad_norm": 5.339010715484619,
|
| 160 |
+
"learning_rate": 5.220744680851064e-06,
|
| 161 |
+
"loss": 0.1711,
|
| 162 |
"step": 1800
|
| 163 |
},
|
| 164 |
{
|
| 165 |
+
"epoch": 2.526595744680851,
|
| 166 |
+
"grad_norm": 20.87778091430664,
|
| 167 |
+
"learning_rate": 4.954787234042554e-06,
|
| 168 |
+
"loss": 0.1889,
|
| 169 |
"step": 1900
|
| 170 |
},
|
| 171 |
{
|
| 172 |
+
"epoch": 2.6595744680851063,
|
| 173 |
+
"grad_norm": 8.609652519226074,
|
| 174 |
+
"learning_rate": 4.6888297872340425e-06,
|
| 175 |
+
"loss": 0.1442,
|
| 176 |
"step": 2000
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 2.6595744680851063,
|
| 180 |
+
"eval_accuracy": 0.9312406778335571,
|
| 181 |
+
"eval_loss": 0.3470732569694519,
|
| 182 |
+
"eval_runtime": 121.4455,
|
| 183 |
+
"eval_samples_per_second": 11.017,
|
| 184 |
+
"eval_steps_per_second": 1.383,
|
| 185 |
"step": 2000
|
| 186 |
},
|
| 187 |
{
|
| 188 |
+
"epoch": 2.7925531914893615,
|
| 189 |
+
"grad_norm": 10.679058074951172,
|
| 190 |
+
"learning_rate": 4.422872340425532e-06,
|
| 191 |
+
"loss": 0.1706,
|
| 192 |
"step": 2100
|
| 193 |
},
|
| 194 |
{
|
| 195 |
+
"epoch": 2.925531914893617,
|
| 196 |
+
"grad_norm": 0.32079407572746277,
|
| 197 |
+
"learning_rate": 4.156914893617022e-06,
|
| 198 |
+
"loss": 0.1333,
|
| 199 |
"step": 2200
|
| 200 |
},
|
| 201 |
{
|
| 202 |
+
"epoch": 3.0585106382978724,
|
| 203 |
+
"grad_norm": 0.19917072355747223,
|
| 204 |
+
"learning_rate": 3.890957446808511e-06,
|
| 205 |
+
"loss": 0.1414,
|
| 206 |
"step": 2300
|
| 207 |
},
|
| 208 |
{
|
| 209 |
+
"epoch": 3.1914893617021276,
|
| 210 |
+
"grad_norm": 3.178858757019043,
|
| 211 |
+
"learning_rate": 3.625e-06,
|
| 212 |
+
"loss": 0.1217,
|
| 213 |
"step": 2400
|
| 214 |
},
|
| 215 |
{
|
| 216 |
+
"epoch": 3.324468085106383,
|
| 217 |
+
"grad_norm": 4.0875020027160645,
|
| 218 |
+
"learning_rate": 3.3590425531914896e-06,
|
| 219 |
+
"loss": 0.1341,
|
| 220 |
"step": 2500
|
| 221 |
},
|
| 222 |
{
|
| 223 |
+
"epoch": 3.324468085106383,
|
| 224 |
+
"eval_accuracy": 0.927503764629364,
|
| 225 |
+
"eval_loss": 0.3641144931316376,
|
| 226 |
+
"eval_runtime": 120.3344,
|
| 227 |
+
"eval_samples_per_second": 11.119,
|
| 228 |
+
"eval_steps_per_second": 1.396,
|
| 229 |
"step": 2500
|
| 230 |
},
|
| 231 |
{
|
| 232 |
+
"epoch": 3.4574468085106385,
|
| 233 |
+
"grad_norm": 2.2575433254241943,
|
| 234 |
+
"learning_rate": 3.0957446808510637e-06,
|
| 235 |
+
"loss": 0.1179,
|
| 236 |
"step": 2600
|
| 237 |
},
|
| 238 |
{
|
| 239 |
+
"epoch": 3.5904255319148937,
|
| 240 |
+
"grad_norm": 2.2771668434143066,
|
| 241 |
+
"learning_rate": 2.8297872340425537e-06,
|
| 242 |
+
"loss": 0.1247,
|
| 243 |
"step": 2700
|
| 244 |
},
|
| 245 |
{
|
| 246 |
+
"epoch": 3.723404255319149,
|
| 247 |
+
"grad_norm": 0.28244930505752563,
|
| 248 |
+
"learning_rate": 2.563829787234043e-06,
|
| 249 |
+
"loss": 0.1186,
|
| 250 |
"step": 2800
|
| 251 |
},
|
| 252 |
{
|
| 253 |
+
"epoch": 3.8563829787234045,
|
| 254 |
+
"grad_norm": 10.92738151550293,
|
| 255 |
+
"learning_rate": 2.297872340425532e-06,
|
| 256 |
+
"loss": 0.1487,
|
| 257 |
"step": 2900
|
| 258 |
},
|
| 259 |
{
|
| 260 |
+
"epoch": 3.9893617021276597,
|
| 261 |
+
"grad_norm": 3.5330650806427,
|
| 262 |
+
"learning_rate": 2.0319148936170213e-06,
|
| 263 |
+
"loss": 0.1136,
|
| 264 |
"step": 3000
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"epoch": 3.9893617021276597,
|
| 268 |
+
"eval_accuracy": 0.9431988000869751,
|
| 269 |
+
"eval_loss": 0.3285652995109558,
|
| 270 |
+
"eval_runtime": 121.4946,
|
| 271 |
+
"eval_samples_per_second": 11.013,
|
| 272 |
+
"eval_steps_per_second": 1.383,
|
| 273 |
"step": 3000
|
| 274 |
},
|
| 275 |
{
|
| 276 |
+
"epoch": 4.122340425531915,
|
| 277 |
+
"grad_norm": 0.24657931923866272,
|
| 278 |
+
"learning_rate": 1.7659574468085109e-06,
|
| 279 |
+
"loss": 0.1157,
|
| 280 |
"step": 3100
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"epoch": 4.25531914893617,
|
| 284 |
+
"grad_norm": 4.8149800300598145,
|
| 285 |
+
"learning_rate": 1.5e-06,
|
| 286 |
+
"loss": 0.1127,
|
| 287 |
"step": 3200
|
| 288 |
},
|
| 289 |
{
|
| 290 |
+
"epoch": 4.388297872340425,
|
| 291 |
+
"grad_norm": 12.910594940185547,
|
| 292 |
+
"learning_rate": 1.2340425531914894e-06,
|
| 293 |
+
"loss": 0.1159,
|
| 294 |
"step": 3300
|
| 295 |
},
|
| 296 |
{
|
| 297 |
+
"epoch": 4.5212765957446805,
|
| 298 |
+
"grad_norm": 0.07863473147153854,
|
| 299 |
+
"learning_rate": 9.680851063829788e-07,
|
| 300 |
+
"loss": 0.1177,
|
| 301 |
"step": 3400
|
| 302 |
},
|
| 303 |
{
|
| 304 |
+
"epoch": 4.654255319148936,
|
| 305 |
+
"grad_norm": 8.249691009521484,
|
| 306 |
+
"learning_rate": 7.021276595744682e-07,
|
| 307 |
+
"loss": 0.1102,
|
| 308 |
"step": 3500
|
| 309 |
},
|
| 310 |
{
|
| 311 |
+
"epoch": 4.654255319148936,
|
| 312 |
+
"eval_accuracy": 0.939461886882782,
|
| 313 |
+
"eval_loss": 0.31690406799316406,
|
| 314 |
+
"eval_runtime": 121.4032,
|
| 315 |
+
"eval_samples_per_second": 11.021,
|
| 316 |
+
"eval_steps_per_second": 1.384,
|
| 317 |
"step": 3500
|
| 318 |
},
|
| 319 |
{
|
| 320 |
+
"epoch": 4.787234042553192,
|
| 321 |
+
"grad_norm": 0.02278612181544304,
|
| 322 |
+
"learning_rate": 4.361702127659575e-07,
|
| 323 |
+
"loss": 0.0957,
|
| 324 |
"step": 3600
|
| 325 |
},
|
| 326 |
{
|
| 327 |
+
"epoch": 4.920212765957447,
|
| 328 |
+
"grad_norm": 11.251379013061523,
|
| 329 |
+
"learning_rate": 1.7021276595744683e-07,
|
| 330 |
+
"loss": 0.1042,
|
| 331 |
"step": 3700
|
| 332 |
},
|
| 333 |
{
|
| 334 |
+
"epoch": 5.0,
|
| 335 |
+
"step": 3760,
|
| 336 |
+
"total_flos": 7.814523952420126e+18,
|
| 337 |
+
"train_loss": 0.29202749221882923,
|
| 338 |
+
"train_runtime": 8842.7154,
|
| 339 |
+
"train_samples_per_second": 6.795,
|
| 340 |
+
"train_steps_per_second": 0.425
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 341 |
}
|
| 342 |
],
|
| 343 |
"logging_steps": 100,
|
| 344 |
+
"max_steps": 3760,
|
| 345 |
"num_input_tokens_seen": 0,
|
| 346 |
"num_train_epochs": 5,
|
| 347 |
"save_steps": 500,
|
|
|
|
| 361 |
"should_evaluate": false,
|
| 362 |
"should_log": false,
|
| 363 |
"should_save": true,
|
| 364 |
+
"should_training_stop": false
|
| 365 |
},
|
| 366 |
"attributes": {}
|
| 367 |
}
|
| 368 |
},
|
| 369 |
+
"total_flos": 7.814523952420126e+18,
|
| 370 |
+
"train_batch_size": 16,
|
| 371 |
"trial_name": null,
|
| 372 |
"trial_params": null
|
| 373 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5bb4f73efab8b12b0f0b7fd78089aeb5800da77a2e0861cb0fac114d0f5be367
|
| 3 |
size 5176
|