UTEL-UIUC/ft-aug-mask2former-swin-small-ade-semantic
Browse files- README.md +45 -27
- all_results.json +6 -6
- model.safetensors +1 -1
- train_results.json +6 -6
- trainer_state.json +474 -189
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [facebook/mask2former-swin-tiny-ade-semantic](https://huggingface.co/facebook/mask2former-swin-tiny-ade-semantic) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
- Loss:
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -43,37 +43,55 @@ The following hyperparameters were used during training:
|
|
| 43 |
- lr_scheduler_type: linear
|
| 44 |
- lr_scheduler_warmup_steps: 4
|
| 45 |
- num_epochs: 100
|
| 46 |
-
- mixed_precision_training: Native AMP
|
| 47 |
|
| 48 |
### Training results
|
| 49 |
|
| 50 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 51 |
|:-------------:|:------:|:----:|:---------------:|
|
| 52 |
-
|
|
| 53 |
-
|
|
| 54 |
-
| 37.
|
| 55 |
-
| 35.
|
| 56 |
-
| 34.
|
| 57 |
-
|
|
| 58 |
-
| 33.
|
| 59 |
-
| 32.
|
| 60 |
-
| 32.
|
| 61 |
-
| 30.
|
| 62 |
-
| 31.
|
| 63 |
-
|
|
| 64 |
-
| 30.
|
| 65 |
-
| 30.
|
| 66 |
-
|
|
| 67 |
-
| 29.
|
| 68 |
-
| 28.
|
| 69 |
-
| 29.
|
| 70 |
-
|
|
| 71 |
-
|
|
| 72 |
-
|
|
| 73 |
-
|
|
| 74 |
-
|
|
| 75 |
-
|
|
| 76 |
-
| 27.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 77 |
|
| 78 |
|
| 79 |
### Framework versions
|
|
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [facebook/mask2former-swin-tiny-ade-semantic](https://huggingface.co/facebook/mask2former-swin-tiny-ade-semantic) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 29.1112
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 43 |
- lr_scheduler_type: linear
|
| 44 |
- lr_scheduler_warmup_steps: 4
|
| 45 |
- num_epochs: 100
|
|
|
|
| 46 |
|
| 47 |
### Training results
|
| 48 |
|
| 49 |
| Training Loss | Epoch | Step | Validation Loss |
|
| 50 |
|:-------------:|:------:|:----:|:---------------:|
|
| 51 |
+
| 50.7018 | 0.1408 | 50 | 44.2435 |
|
| 52 |
+
| 40.5877 | 0.2817 | 100 | 39.6465 |
|
| 53 |
+
| 37.4102 | 0.4225 | 150 | 37.2471 |
|
| 54 |
+
| 35.7502 | 0.5634 | 200 | 36.3455 |
|
| 55 |
+
| 34.7067 | 0.7042 | 250 | 34.8824 |
|
| 56 |
+
| 34.0798 | 0.8451 | 300 | 34.8520 |
|
| 57 |
+
| 33.3503 | 0.9859 | 350 | 33.7321 |
|
| 58 |
+
| 32.3436 | 1.1268 | 400 | 33.1560 |
|
| 59 |
+
| 32.3845 | 1.2676 | 450 | 33.0411 |
|
| 60 |
+
| 30.8809 | 1.4085 | 500 | 32.7852 |
|
| 61 |
+
| 31.689 | 1.5493 | 550 | 31.9914 |
|
| 62 |
+
| 31.036 | 1.6901 | 600 | 32.7297 |
|
| 63 |
+
| 30.9795 | 1.8310 | 650 | 31.8848 |
|
| 64 |
+
| 30.7918 | 1.9718 | 700 | 31.5285 |
|
| 65 |
+
| 30.1432 | 2.1127 | 750 | 32.0634 |
|
| 66 |
+
| 29.7082 | 2.2535 | 800 | 31.1849 |
|
| 67 |
+
| 28.7869 | 2.3944 | 850 | 30.9022 |
|
| 68 |
+
| 29.4227 | 2.5352 | 900 | 30.5902 |
|
| 69 |
+
| 29.1865 | 2.6761 | 950 | 30.3818 |
|
| 70 |
+
| 29.2715 | 2.8169 | 1000 | 30.9196 |
|
| 71 |
+
| 29.1941 | 2.9577 | 1050 | 30.8163 |
|
| 72 |
+
| 28.5256 | 3.0986 | 1100 | 30.4730 |
|
| 73 |
+
| 28.0419 | 3.2394 | 1150 | 30.6531 |
|
| 74 |
+
| 28.0538 | 3.3803 | 1200 | 30.0779 |
|
| 75 |
+
| 27.9463 | 3.5211 | 1250 | 30.6114 |
|
| 76 |
+
| 27.4152 | 3.6620 | 1300 | 30.5519 |
|
| 77 |
+
| 27.7461 | 3.8028 | 1350 | 29.5641 |
|
| 78 |
+
| 27.5604 | 3.9437 | 1400 | 30.1296 |
|
| 79 |
+
| 27.381 | 4.0845 | 1450 | 30.5017 |
|
| 80 |
+
| 26.3816 | 4.2254 | 1500 | 29.6898 |
|
| 81 |
+
| 26.5218 | 4.3662 | 1550 | 29.9475 |
|
| 82 |
+
| 26.9798 | 4.5070 | 1600 | 29.3323 |
|
| 83 |
+
| 26.8186 | 4.6479 | 1650 | 29.5755 |
|
| 84 |
+
| 27.5111 | 4.7887 | 1700 | 30.7945 |
|
| 85 |
+
| 27.0839 | 4.9296 | 1750 | 29.4147 |
|
| 86 |
+
| 26.6393 | 5.0704 | 1800 | 28.7983 |
|
| 87 |
+
| 26.3564 | 5.2113 | 1850 | 29.2245 |
|
| 88 |
+
| 25.6174 | 5.3521 | 1900 | 28.9337 |
|
| 89 |
+
| 25.8777 | 5.4930 | 1950 | 29.4778 |
|
| 90 |
+
| 25.6848 | 5.6338 | 2000 | 28.4992 |
|
| 91 |
+
| 26.4625 | 5.7746 | 2050 | 29.6182 |
|
| 92 |
+
| 26.8448 | 5.9155 | 2100 | 29.5377 |
|
| 93 |
+
| 26.0681 | 6.0563 | 2150 | 29.2390 |
|
| 94 |
+
| 25.628 | 6.1972 | 2200 | 29.1112 |
|
| 95 |
|
| 96 |
|
| 97 |
### Framework versions
|
all_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 8.
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 6.197183098591549,
|
| 3 |
+
"total_flos": 8.850907754333798e+18,
|
| 4 |
+
"train_loss": 29.80798134543679,
|
| 5 |
+
"train_runtime": 17574.1061,
|
| 6 |
+
"train_samples_per_second": 64.612,
|
| 7 |
+
"train_steps_per_second": 2.02
|
| 8 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 190070416
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14b34ad757d964baed8cb111ffe1387710bae463d2b89500a89e153820b9681b
|
| 3 |
size 190070416
|
train_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos": 8.
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 6.197183098591549,
|
| 3 |
+
"total_flos": 8.850907754333798e+18,
|
| 4 |
+
"train_loss": 29.80798134543679,
|
| 5 |
+
"train_runtime": 17574.1061,
|
| 6 |
+
"train_samples_per_second": 64.612,
|
| 7 |
+
"train_steps_per_second": 2.02
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,396 +1,681 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric":
|
| 3 |
-
"best_model_checkpoint": "mask2former/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 50,
|
| 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": 0.14084507042253522,
|
| 13 |
-
"grad_norm":
|
| 14 |
-
"learning_rate": 4.
|
| 15 |
-
"loss":
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.14084507042253522,
|
| 20 |
-
"eval_loss":
|
| 21 |
-
"eval_runtime":
|
| 22 |
-
"eval_samples_per_second":
|
| 23 |
-
"eval_steps_per_second": 1.
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.28169014084507044,
|
| 28 |
-
"grad_norm":
|
| 29 |
-
"learning_rate": 4.
|
| 30 |
-
"loss":
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.28169014084507044,
|
| 35 |
-
"eval_loss":
|
| 36 |
-
"eval_runtime":
|
| 37 |
-
"eval_samples_per_second":
|
| 38 |
-
"eval_steps_per_second": 1.
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.4225352112676056,
|
| 43 |
-
"grad_norm":
|
| 44 |
-
"learning_rate": 4.
|
| 45 |
-
"loss": 37.
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.4225352112676056,
|
| 50 |
-
"eval_loss":
|
| 51 |
-
"eval_runtime":
|
| 52 |
-
"eval_samples_per_second": 10.
|
| 53 |
-
"eval_steps_per_second": 1.
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.5633802816901409,
|
| 58 |
-
"grad_norm":
|
| 59 |
-
"learning_rate": 4.
|
| 60 |
-
"loss": 35.
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 0.5633802816901409,
|
| 65 |
-
"eval_loss":
|
| 66 |
-
"eval_runtime":
|
| 67 |
-
"eval_samples_per_second":
|
| 68 |
-
"eval_steps_per_second": 1.
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.704225352112676,
|
| 73 |
-
"grad_norm":
|
| 74 |
-
"learning_rate": 4.
|
| 75 |
-
"loss": 34.
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
"epoch": 0.704225352112676,
|
| 80 |
-
"eval_loss":
|
| 81 |
-
"eval_runtime":
|
| 82 |
-
"eval_samples_per_second": 10.
|
| 83 |
-
"eval_steps_per_second": 1.
|
| 84 |
"step": 250
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.8450704225352113,
|
| 88 |
-
"grad_norm":
|
| 89 |
-
"learning_rate": 4.
|
| 90 |
-
"loss":
|
| 91 |
"step": 300
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.8450704225352113,
|
| 95 |
-
"eval_loss":
|
| 96 |
-
"eval_runtime":
|
| 97 |
-
"eval_samples_per_second":
|
| 98 |
-
"eval_steps_per_second": 1.
|
| 99 |
"step": 300
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.9859154929577465,
|
| 103 |
-
"grad_norm":
|
| 104 |
-
"learning_rate": 4.
|
| 105 |
-
"loss": 33.
|
| 106 |
"step": 350
|
| 107 |
},
|
| 108 |
{
|
| 109 |
"epoch": 0.9859154929577465,
|
| 110 |
-
"eval_loss":
|
| 111 |
-
"eval_runtime":
|
| 112 |
-
"eval_samples_per_second":
|
| 113 |
-
"eval_steps_per_second": 1.
|
| 114 |
"step": 350
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 1.1267605633802817,
|
| 118 |
-
"grad_norm":
|
| 119 |
-
"learning_rate": 4.
|
| 120 |
-
"loss": 32.
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 1.1267605633802817,
|
| 125 |
-
"eval_loss":
|
| 126 |
-
"eval_runtime":
|
| 127 |
-
"eval_samples_per_second":
|
| 128 |
-
"eval_steps_per_second": 1.
|
| 129 |
"step": 400
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.267605633802817,
|
| 133 |
-
"grad_norm":
|
| 134 |
-
"learning_rate": 4.
|
| 135 |
-
"loss": 32.
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 1.267605633802817,
|
| 140 |
-
"eval_loss":
|
| 141 |
-
"eval_runtime":
|
| 142 |
-
"eval_samples_per_second":
|
| 143 |
-
"eval_steps_per_second": 1.
|
| 144 |
"step": 450
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.408450704225352,
|
| 148 |
-
"grad_norm":
|
| 149 |
-
"learning_rate": 4.
|
| 150 |
-
"loss": 30.
|
| 151 |
"step": 500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.408450704225352,
|
| 155 |
-
"eval_loss":
|
| 156 |
-
"eval_runtime":
|
| 157 |
-
"eval_samples_per_second":
|
| 158 |
-
"eval_steps_per_second": 1.
|
| 159 |
"step": 500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 1.5492957746478875,
|
| 163 |
-
"grad_norm":
|
| 164 |
-
"learning_rate": 4.
|
| 165 |
-
"loss": 31.
|
| 166 |
"step": 550
|
| 167 |
},
|
| 168 |
{
|
| 169 |
"epoch": 1.5492957746478875,
|
| 170 |
-
"eval_loss":
|
| 171 |
-
"eval_runtime":
|
| 172 |
-
"eval_samples_per_second":
|
| 173 |
-
"eval_steps_per_second": 1.
|
| 174 |
"step": 550
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 1.6901408450704225,
|
| 178 |
-
"grad_norm":
|
| 179 |
-
"learning_rate": 4.
|
| 180 |
-
"loss":
|
| 181 |
"step": 600
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 1.6901408450704225,
|
| 185 |
-
"eval_loss":
|
| 186 |
-
"eval_runtime":
|
| 187 |
-
"eval_samples_per_second":
|
| 188 |
-
"eval_steps_per_second": 1.
|
| 189 |
"step": 600
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 1.8309859154929577,
|
| 193 |
-
"grad_norm":
|
| 194 |
-
"learning_rate": 4.
|
| 195 |
-
"loss": 30.
|
| 196 |
"step": 650
|
| 197 |
},
|
| 198 |
{
|
| 199 |
"epoch": 1.8309859154929577,
|
| 200 |
-
"eval_loss":
|
| 201 |
-
"eval_runtime":
|
| 202 |
-
"eval_samples_per_second":
|
| 203 |
-
"eval_steps_per_second": 1.
|
| 204 |
"step": 650
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 1.971830985915493,
|
| 208 |
-
"grad_norm":
|
| 209 |
-
"learning_rate": 4.
|
| 210 |
-
"loss": 30.
|
| 211 |
"step": 700
|
| 212 |
},
|
| 213 |
{
|
| 214 |
"epoch": 1.971830985915493,
|
| 215 |
-
"eval_loss":
|
| 216 |
-
"eval_runtime":
|
| 217 |
-
"eval_samples_per_second":
|
| 218 |
-
"eval_steps_per_second": 1.
|
| 219 |
"step": 700
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 2.112676056338028,
|
| 223 |
-
"grad_norm":
|
| 224 |
-
"learning_rate": 4.
|
| 225 |
-
"loss":
|
| 226 |
"step": 750
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 2.112676056338028,
|
| 230 |
-
"eval_loss":
|
| 231 |
-
"eval_runtime":
|
| 232 |
-
"eval_samples_per_second":
|
| 233 |
-
"eval_steps_per_second": 1.
|
| 234 |
"step": 750
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 2.2535211267605635,
|
| 238 |
-
"grad_norm":
|
| 239 |
-
"learning_rate": 4.
|
| 240 |
-
"loss": 29.
|
| 241 |
"step": 800
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 2.2535211267605635,
|
| 245 |
-
"eval_loss":
|
| 246 |
-
"eval_runtime":
|
| 247 |
-
"eval_samples_per_second":
|
| 248 |
-
"eval_steps_per_second": 1.
|
| 249 |
"step": 800
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 2.3943661971830985,
|
| 253 |
-
"grad_norm":
|
| 254 |
-
"learning_rate": 4.
|
| 255 |
-
"loss": 28.
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 2.3943661971830985,
|
| 260 |
-
"eval_loss":
|
| 261 |
-
"eval_runtime":
|
| 262 |
-
"eval_samples_per_second": 9.
|
| 263 |
-
"eval_steps_per_second": 1.
|
| 264 |
"step": 850
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 2.535211267605634,
|
| 268 |
-
"grad_norm":
|
| 269 |
-
"learning_rate": 4.
|
| 270 |
-
"loss": 29.
|
| 271 |
"step": 900
|
| 272 |
},
|
| 273 |
{
|
| 274 |
"epoch": 2.535211267605634,
|
| 275 |
-
"eval_loss":
|
| 276 |
-
"eval_runtime":
|
| 277 |
-
"eval_samples_per_second":
|
| 278 |
-
"eval_steps_per_second": 1.
|
| 279 |
"step": 900
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 2.676056338028169,
|
| 283 |
-
"grad_norm":
|
| 284 |
-
"learning_rate": 4.
|
| 285 |
-
"loss":
|
| 286 |
"step": 950
|
| 287 |
},
|
| 288 |
{
|
| 289 |
"epoch": 2.676056338028169,
|
| 290 |
-
"eval_loss":
|
| 291 |
-
"eval_runtime":
|
| 292 |
-
"eval_samples_per_second": 10.
|
| 293 |
-
"eval_steps_per_second": 1.
|
| 294 |
"step": 950
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 2.816901408450704,
|
| 298 |
-
"grad_norm":
|
| 299 |
-
"learning_rate": 4.
|
| 300 |
-
"loss":
|
| 301 |
"step": 1000
|
| 302 |
},
|
| 303 |
{
|
| 304 |
"epoch": 2.816901408450704,
|
| 305 |
-
"eval_loss":
|
| 306 |
-
"eval_runtime":
|
| 307 |
-
"eval_samples_per_second":
|
| 308 |
-
"eval_steps_per_second": 1.
|
| 309 |
"step": 1000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 2.9577464788732395,
|
| 313 |
-
"grad_norm":
|
| 314 |
-
"learning_rate": 4.
|
| 315 |
-
"loss":
|
| 316 |
"step": 1050
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 2.9577464788732395,
|
| 320 |
-
"eval_loss":
|
| 321 |
-
"eval_runtime":
|
| 322 |
-
"eval_samples_per_second":
|
| 323 |
-
"eval_steps_per_second": 1.
|
| 324 |
"step": 1050
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 3.0985915492957745,
|
| 328 |
-
"grad_norm":
|
| 329 |
-
"learning_rate": 4.
|
| 330 |
-
"loss":
|
| 331 |
"step": 1100
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 3.0985915492957745,
|
| 335 |
-
"eval_loss":
|
| 336 |
-
"eval_runtime":
|
| 337 |
-
"eval_samples_per_second":
|
| 338 |
-
"eval_steps_per_second": 1.
|
| 339 |
"step": 1100
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 3.23943661971831,
|
| 343 |
-
"grad_norm":
|
| 344 |
-
"learning_rate": 4.
|
| 345 |
-
"loss":
|
| 346 |
"step": 1150
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 3.23943661971831,
|
| 350 |
-
"eval_loss":
|
| 351 |
-
"eval_runtime":
|
| 352 |
-
"eval_samples_per_second":
|
| 353 |
-
"eval_steps_per_second": 1.
|
| 354 |
"step": 1150
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 3.380281690140845,
|
| 358 |
-
"grad_norm":
|
| 359 |
-
"learning_rate": 4.
|
| 360 |
-
"loss":
|
| 361 |
"step": 1200
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 3.380281690140845,
|
| 365 |
-
"eval_loss":
|
| 366 |
-
"eval_runtime":
|
| 367 |
-
"eval_samples_per_second":
|
| 368 |
-
"eval_steps_per_second": 1.
|
| 369 |
"step": 1200
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 3.52112676056338,
|
| 373 |
-
"grad_norm":
|
| 374 |
-
"learning_rate": 4.
|
| 375 |
-
"loss": 27.
|
| 376 |
"step": 1250
|
| 377 |
},
|
| 378 |
{
|
| 379 |
"epoch": 3.52112676056338,
|
| 380 |
-
"eval_loss":
|
| 381 |
-
"eval_runtime":
|
| 382 |
-
"eval_samples_per_second":
|
| 383 |
-
"eval_steps_per_second": 1.
|
| 384 |
"step": 1250
|
| 385 |
},
|
| 386 |
{
|
| 387 |
-
"epoch": 3.
|
| 388 |
-
"
|
| 389 |
-
"
|
| 390 |
-
"
|
| 391 |
-
"
|
| 392 |
-
|
| 393 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 394 |
}
|
| 395 |
],
|
| 396 |
"logging_steps": 50,
|
|
@@ -405,7 +690,7 @@
|
|
| 405 |
"early_stopping_threshold": 0.0
|
| 406 |
},
|
| 407 |
"attributes": {
|
| 408 |
-
"early_stopping_patience_counter":
|
| 409 |
}
|
| 410 |
},
|
| 411 |
"TrainerControl": {
|
|
@@ -414,12 +699,12 @@
|
|
| 414 |
"should_evaluate": false,
|
| 415 |
"should_log": false,
|
| 416 |
"should_save": true,
|
| 417 |
-
"should_training_stop":
|
| 418 |
},
|
| 419 |
"attributes": {}
|
| 420 |
}
|
| 421 |
},
|
| 422 |
-
"total_flos": 8.
|
| 423 |
"train_batch_size": 8,
|
| 424 |
"trial_name": null,
|
| 425 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 28.4991512298584,
|
| 3 |
+
"best_model_checkpoint": "mask2former/checkpoint-2000",
|
| 4 |
+
"epoch": 6.197183098591549,
|
| 5 |
"eval_steps": 50,
|
| 6 |
+
"global_step": 2200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.14084507042253522,
|
| 13 |
+
"grad_norm": 227.8921661376953,
|
| 14 |
+
"learning_rate": 4.9935203966644134e-05,
|
| 15 |
+
"loss": 50.7018,
|
| 16 |
"step": 50
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.14084507042253522,
|
| 20 |
+
"eval_loss": 44.24353790283203,
|
| 21 |
+
"eval_runtime": 135.6148,
|
| 22 |
+
"eval_samples_per_second": 9.306,
|
| 23 |
+
"eval_steps_per_second": 1.165,
|
| 24 |
"step": 50
|
| 25 |
},
|
| 26 |
{
|
| 27 |
"epoch": 0.28169014084507044,
|
| 28 |
+
"grad_norm": 232.9764404296875,
|
| 29 |
+
"learning_rate": 4.986477349560514e-05,
|
| 30 |
+
"loss": 40.5877,
|
| 31 |
"step": 100
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"epoch": 0.28169014084507044,
|
| 35 |
+
"eval_loss": 39.646488189697266,
|
| 36 |
+
"eval_runtime": 119.1324,
|
| 37 |
+
"eval_samples_per_second": 10.593,
|
| 38 |
+
"eval_steps_per_second": 1.326,
|
| 39 |
"step": 100
|
| 40 |
},
|
| 41 |
{
|
| 42 |
"epoch": 0.4225352112676056,
|
| 43 |
+
"grad_norm": 375.6308898925781,
|
| 44 |
+
"learning_rate": 4.979434302456615e-05,
|
| 45 |
+
"loss": 37.4102,
|
| 46 |
"step": 150
|
| 47 |
},
|
| 48 |
{
|
| 49 |
"epoch": 0.4225352112676056,
|
| 50 |
+
"eval_loss": 37.247108459472656,
|
| 51 |
+
"eval_runtime": 119.6132,
|
| 52 |
+
"eval_samples_per_second": 10.551,
|
| 53 |
+
"eval_steps_per_second": 1.321,
|
| 54 |
"step": 150
|
| 55 |
},
|
| 56 |
{
|
| 57 |
"epoch": 0.5633802816901409,
|
| 58 |
+
"grad_norm": 189.58392333984375,
|
| 59 |
+
"learning_rate": 4.972391255352716e-05,
|
| 60 |
+
"loss": 35.7502,
|
| 61 |
"step": 200
|
| 62 |
},
|
| 63 |
{
|
| 64 |
"epoch": 0.5633802816901409,
|
| 65 |
+
"eval_loss": 36.34551239013672,
|
| 66 |
+
"eval_runtime": 132.8881,
|
| 67 |
+
"eval_samples_per_second": 9.497,
|
| 68 |
+
"eval_steps_per_second": 1.189,
|
| 69 |
"step": 200
|
| 70 |
},
|
| 71 |
{
|
| 72 |
"epoch": 0.704225352112676,
|
| 73 |
+
"grad_norm": 163.74330139160156,
|
| 74 |
+
"learning_rate": 4.965348208248817e-05,
|
| 75 |
+
"loss": 34.7067,
|
| 76 |
"step": 250
|
| 77 |
},
|
| 78 |
{
|
| 79 |
"epoch": 0.704225352112676,
|
| 80 |
+
"eval_loss": 34.88238525390625,
|
| 81 |
+
"eval_runtime": 120.2665,
|
| 82 |
+
"eval_samples_per_second": 10.493,
|
| 83 |
+
"eval_steps_per_second": 1.314,
|
| 84 |
"step": 250
|
| 85 |
},
|
| 86 |
{
|
| 87 |
"epoch": 0.8450704225352113,
|
| 88 |
+
"grad_norm": 173.85494995117188,
|
| 89 |
+
"learning_rate": 4.958305161144918e-05,
|
| 90 |
+
"loss": 34.0798,
|
| 91 |
"step": 300
|
| 92 |
},
|
| 93 |
{
|
| 94 |
"epoch": 0.8450704225352113,
|
| 95 |
+
"eval_loss": 34.85204315185547,
|
| 96 |
+
"eval_runtime": 129.3353,
|
| 97 |
+
"eval_samples_per_second": 9.758,
|
| 98 |
+
"eval_steps_per_second": 1.222,
|
| 99 |
"step": 300
|
| 100 |
},
|
| 101 |
{
|
| 102 |
"epoch": 0.9859154929577465,
|
| 103 |
+
"grad_norm": 167.4436492919922,
|
| 104 |
+
"learning_rate": 4.951262114041019e-05,
|
| 105 |
+
"loss": 33.3503,
|
| 106 |
"step": 350
|
| 107 |
},
|
| 108 |
{
|
| 109 |
"epoch": 0.9859154929577465,
|
| 110 |
+
"eval_loss": 33.73210144042969,
|
| 111 |
+
"eval_runtime": 143.3502,
|
| 112 |
+
"eval_samples_per_second": 8.804,
|
| 113 |
+
"eval_steps_per_second": 1.102,
|
| 114 |
"step": 350
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 1.1267605633802817,
|
| 118 |
+
"grad_norm": 202.66012573242188,
|
| 119 |
+
"learning_rate": 4.944219066937119e-05,
|
| 120 |
+
"loss": 32.3436,
|
| 121 |
"step": 400
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 1.1267605633802817,
|
| 125 |
+
"eval_loss": 33.15604782104492,
|
| 126 |
+
"eval_runtime": 123.8826,
|
| 127 |
+
"eval_samples_per_second": 10.187,
|
| 128 |
+
"eval_steps_per_second": 1.275,
|
| 129 |
"step": 400
|
| 130 |
},
|
| 131 |
{
|
| 132 |
"epoch": 1.267605633802817,
|
| 133 |
+
"grad_norm": 145.71310424804688,
|
| 134 |
+
"learning_rate": 4.937176019833221e-05,
|
| 135 |
+
"loss": 32.3845,
|
| 136 |
"step": 450
|
| 137 |
},
|
| 138 |
{
|
| 139 |
"epoch": 1.267605633802817,
|
| 140 |
+
"eval_loss": 33.041107177734375,
|
| 141 |
+
"eval_runtime": 125.7421,
|
| 142 |
+
"eval_samples_per_second": 10.036,
|
| 143 |
+
"eval_steps_per_second": 1.257,
|
| 144 |
"step": 450
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"epoch": 1.408450704225352,
|
| 148 |
+
"grad_norm": 143.16452026367188,
|
| 149 |
+
"learning_rate": 4.930132972729321e-05,
|
| 150 |
+
"loss": 30.8809,
|
| 151 |
"step": 500
|
| 152 |
},
|
| 153 |
{
|
| 154 |
"epoch": 1.408450704225352,
|
| 155 |
+
"eval_loss": 32.785240173339844,
|
| 156 |
+
"eval_runtime": 137.1907,
|
| 157 |
+
"eval_samples_per_second": 9.199,
|
| 158 |
+
"eval_steps_per_second": 1.152,
|
| 159 |
"step": 500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
"epoch": 1.5492957746478875,
|
| 163 |
+
"grad_norm": 159.82777404785156,
|
| 164 |
+
"learning_rate": 4.9230899256254227e-05,
|
| 165 |
+
"loss": 31.689,
|
| 166 |
"step": 550
|
| 167 |
},
|
| 168 |
{
|
| 169 |
"epoch": 1.5492957746478875,
|
| 170 |
+
"eval_loss": 31.99137306213379,
|
| 171 |
+
"eval_runtime": 129.225,
|
| 172 |
+
"eval_samples_per_second": 9.766,
|
| 173 |
+
"eval_steps_per_second": 1.223,
|
| 174 |
"step": 550
|
| 175 |
},
|
| 176 |
{
|
| 177 |
"epoch": 1.6901408450704225,
|
| 178 |
+
"grad_norm": 160.78164672851562,
|
| 179 |
+
"learning_rate": 4.916046878521523e-05,
|
| 180 |
+
"loss": 31.036,
|
| 181 |
"step": 600
|
| 182 |
},
|
| 183 |
{
|
| 184 |
"epoch": 1.6901408450704225,
|
| 185 |
+
"eval_loss": 32.72974395751953,
|
| 186 |
+
"eval_runtime": 129.162,
|
| 187 |
+
"eval_samples_per_second": 9.771,
|
| 188 |
+
"eval_steps_per_second": 1.223,
|
| 189 |
"step": 600
|
| 190 |
},
|
| 191 |
{
|
| 192 |
"epoch": 1.8309859154929577,
|
| 193 |
+
"grad_norm": 206.84974670410156,
|
| 194 |
+
"learning_rate": 4.9090038314176246e-05,
|
| 195 |
+
"loss": 30.9795,
|
| 196 |
"step": 650
|
| 197 |
},
|
| 198 |
{
|
| 199 |
"epoch": 1.8309859154929577,
|
| 200 |
+
"eval_loss": 31.88483238220215,
|
| 201 |
+
"eval_runtime": 126.8688,
|
| 202 |
+
"eval_samples_per_second": 9.947,
|
| 203 |
+
"eval_steps_per_second": 1.245,
|
| 204 |
"step": 650
|
| 205 |
},
|
| 206 |
{
|
| 207 |
"epoch": 1.971830985915493,
|
| 208 |
+
"grad_norm": 128.7499237060547,
|
| 209 |
+
"learning_rate": 4.901960784313725e-05,
|
| 210 |
+
"loss": 30.7918,
|
| 211 |
"step": 700
|
| 212 |
},
|
| 213 |
{
|
| 214 |
"epoch": 1.971830985915493,
|
| 215 |
+
"eval_loss": 31.528514862060547,
|
| 216 |
+
"eval_runtime": 131.1893,
|
| 217 |
+
"eval_samples_per_second": 9.62,
|
| 218 |
+
"eval_steps_per_second": 1.204,
|
| 219 |
"step": 700
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 2.112676056338028,
|
| 223 |
+
"grad_norm": 155.95223999023438,
|
| 224 |
+
"learning_rate": 4.8949177372098266e-05,
|
| 225 |
+
"loss": 30.1432,
|
| 226 |
"step": 750
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 2.112676056338028,
|
| 230 |
+
"eval_loss": 32.06336212158203,
|
| 231 |
+
"eval_runtime": 130.1706,
|
| 232 |
+
"eval_samples_per_second": 9.695,
|
| 233 |
+
"eval_steps_per_second": 1.214,
|
| 234 |
"step": 750
|
| 235 |
},
|
| 236 |
{
|
| 237 |
"epoch": 2.2535211267605635,
|
| 238 |
+
"grad_norm": 122.61072540283203,
|
| 239 |
+
"learning_rate": 4.887874690105927e-05,
|
| 240 |
+
"loss": 29.7082,
|
| 241 |
"step": 800
|
| 242 |
},
|
| 243 |
{
|
| 244 |
"epoch": 2.2535211267605635,
|
| 245 |
+
"eval_loss": 31.184894561767578,
|
| 246 |
+
"eval_runtime": 121.0117,
|
| 247 |
+
"eval_samples_per_second": 10.429,
|
| 248 |
+
"eval_steps_per_second": 1.306,
|
| 249 |
"step": 800
|
| 250 |
},
|
| 251 |
{
|
| 252 |
"epoch": 2.3943661971830985,
|
| 253 |
+
"grad_norm": 139.07225036621094,
|
| 254 |
+
"learning_rate": 4.8808316430020286e-05,
|
| 255 |
+
"loss": 28.7869,
|
| 256 |
"step": 850
|
| 257 |
},
|
| 258 |
{
|
| 259 |
"epoch": 2.3943661971830985,
|
| 260 |
+
"eval_loss": 30.902196884155273,
|
| 261 |
+
"eval_runtime": 127.2634,
|
| 262 |
+
"eval_samples_per_second": 9.916,
|
| 263 |
+
"eval_steps_per_second": 1.242,
|
| 264 |
"step": 850
|
| 265 |
},
|
| 266 |
{
|
| 267 |
"epoch": 2.535211267605634,
|
| 268 |
+
"grad_norm": 188.21234130859375,
|
| 269 |
+
"learning_rate": 4.873788595898129e-05,
|
| 270 |
+
"loss": 29.4227,
|
| 271 |
"step": 900
|
| 272 |
},
|
| 273 |
{
|
| 274 |
"epoch": 2.535211267605634,
|
| 275 |
+
"eval_loss": 30.5902099609375,
|
| 276 |
+
"eval_runtime": 146.3048,
|
| 277 |
+
"eval_samples_per_second": 8.626,
|
| 278 |
+
"eval_steps_per_second": 1.08,
|
| 279 |
"step": 900
|
| 280 |
},
|
| 281 |
{
|
| 282 |
"epoch": 2.676056338028169,
|
| 283 |
+
"grad_norm": 302.049560546875,
|
| 284 |
+
"learning_rate": 4.8667455487942306e-05,
|
| 285 |
+
"loss": 29.1865,
|
| 286 |
"step": 950
|
| 287 |
},
|
| 288 |
{
|
| 289 |
"epoch": 2.676056338028169,
|
| 290 |
+
"eval_loss": 30.381799697875977,
|
| 291 |
+
"eval_runtime": 118.3226,
|
| 292 |
+
"eval_samples_per_second": 10.666,
|
| 293 |
+
"eval_steps_per_second": 1.335,
|
| 294 |
"step": 950
|
| 295 |
},
|
| 296 |
{
|
| 297 |
"epoch": 2.816901408450704,
|
| 298 |
+
"grad_norm": 151.5100860595703,
|
| 299 |
+
"learning_rate": 4.859702501690331e-05,
|
| 300 |
+
"loss": 29.2715,
|
| 301 |
"step": 1000
|
| 302 |
},
|
| 303 |
{
|
| 304 |
"epoch": 2.816901408450704,
|
| 305 |
+
"eval_loss": 30.919567108154297,
|
| 306 |
+
"eval_runtime": 128.7879,
|
| 307 |
+
"eval_samples_per_second": 9.799,
|
| 308 |
+
"eval_steps_per_second": 1.227,
|
| 309 |
"step": 1000
|
| 310 |
},
|
| 311 |
{
|
| 312 |
"epoch": 2.9577464788732395,
|
| 313 |
+
"grad_norm": 135.14910888671875,
|
| 314 |
+
"learning_rate": 4.8526594545864326e-05,
|
| 315 |
+
"loss": 29.1941,
|
| 316 |
"step": 1050
|
| 317 |
},
|
| 318 |
{
|
| 319 |
"epoch": 2.9577464788732395,
|
| 320 |
+
"eval_loss": 30.816268920898438,
|
| 321 |
+
"eval_runtime": 127.4007,
|
| 322 |
+
"eval_samples_per_second": 9.906,
|
| 323 |
+
"eval_steps_per_second": 1.24,
|
| 324 |
"step": 1050
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 3.0985915492957745,
|
| 328 |
+
"grad_norm": 121.41586303710938,
|
| 329 |
+
"learning_rate": 4.845616407482533e-05,
|
| 330 |
+
"loss": 28.5256,
|
| 331 |
"step": 1100
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 3.0985915492957745,
|
| 335 |
+
"eval_loss": 30.472957611083984,
|
| 336 |
+
"eval_runtime": 144.0632,
|
| 337 |
+
"eval_samples_per_second": 8.76,
|
| 338 |
+
"eval_steps_per_second": 1.097,
|
| 339 |
"step": 1100
|
| 340 |
},
|
| 341 |
{
|
| 342 |
"epoch": 3.23943661971831,
|
| 343 |
+
"grad_norm": 217.75247192382812,
|
| 344 |
+
"learning_rate": 4.8385733603786346e-05,
|
| 345 |
+
"loss": 28.0419,
|
| 346 |
"step": 1150
|
| 347 |
},
|
| 348 |
{
|
| 349 |
"epoch": 3.23943661971831,
|
| 350 |
+
"eval_loss": 30.653095245361328,
|
| 351 |
+
"eval_runtime": 129.0443,
|
| 352 |
+
"eval_samples_per_second": 9.78,
|
| 353 |
+
"eval_steps_per_second": 1.224,
|
| 354 |
"step": 1150
|
| 355 |
},
|
| 356 |
{
|
| 357 |
"epoch": 3.380281690140845,
|
| 358 |
+
"grad_norm": 129.21693420410156,
|
| 359 |
+
"learning_rate": 4.831530313274735e-05,
|
| 360 |
+
"loss": 28.0538,
|
| 361 |
"step": 1200
|
| 362 |
},
|
| 363 |
{
|
| 364 |
"epoch": 3.380281690140845,
|
| 365 |
+
"eval_loss": 30.077850341796875,
|
| 366 |
+
"eval_runtime": 119.4037,
|
| 367 |
+
"eval_samples_per_second": 10.569,
|
| 368 |
+
"eval_steps_per_second": 1.323,
|
| 369 |
"step": 1200
|
| 370 |
},
|
| 371 |
{
|
| 372 |
"epoch": 3.52112676056338,
|
| 373 |
+
"grad_norm": 116.27620697021484,
|
| 374 |
+
"learning_rate": 4.8244872661708365e-05,
|
| 375 |
+
"loss": 27.9463,
|
| 376 |
"step": 1250
|
| 377 |
},
|
| 378 |
{
|
| 379 |
"epoch": 3.52112676056338,
|
| 380 |
+
"eval_loss": 30.61139488220215,
|
| 381 |
+
"eval_runtime": 132.5429,
|
| 382 |
+
"eval_samples_per_second": 9.521,
|
| 383 |
+
"eval_steps_per_second": 1.192,
|
| 384 |
"step": 1250
|
| 385 |
},
|
| 386 |
{
|
| 387 |
+
"epoch": 3.6619718309859155,
|
| 388 |
+
"grad_norm": 149.399169921875,
|
| 389 |
+
"learning_rate": 4.817444219066937e-05,
|
| 390 |
+
"loss": 27.4152,
|
| 391 |
+
"step": 1300
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 3.6619718309859155,
|
| 395 |
+
"eval_loss": 30.551870346069336,
|
| 396 |
+
"eval_runtime": 130.4003,
|
| 397 |
+
"eval_samples_per_second": 9.678,
|
| 398 |
+
"eval_steps_per_second": 1.212,
|
| 399 |
+
"step": 1300
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 3.802816901408451,
|
| 403 |
+
"grad_norm": 175.60769653320312,
|
| 404 |
+
"learning_rate": 4.8104011719630385e-05,
|
| 405 |
+
"loss": 27.7461,
|
| 406 |
+
"step": 1350
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 3.802816901408451,
|
| 410 |
+
"eval_loss": 29.564067840576172,
|
| 411 |
+
"eval_runtime": 131.8055,
|
| 412 |
+
"eval_samples_per_second": 9.575,
|
| 413 |
+
"eval_steps_per_second": 1.199,
|
| 414 |
+
"step": 1350
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.943661971830986,
|
| 418 |
+
"grad_norm": 114.43487548828125,
|
| 419 |
+
"learning_rate": 4.803358124859139e-05,
|
| 420 |
+
"loss": 27.5604,
|
| 421 |
+
"step": 1400
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.943661971830986,
|
| 425 |
+
"eval_loss": 30.12961769104004,
|
| 426 |
+
"eval_runtime": 148.0714,
|
| 427 |
+
"eval_samples_per_second": 8.523,
|
| 428 |
+
"eval_steps_per_second": 1.067,
|
| 429 |
+
"step": 1400
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 4.084507042253521,
|
| 433 |
+
"grad_norm": 189.4749298095703,
|
| 434 |
+
"learning_rate": 4.79631507775524e-05,
|
| 435 |
+
"loss": 27.381,
|
| 436 |
+
"step": 1450
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 4.084507042253521,
|
| 440 |
+
"eval_loss": 30.50173568725586,
|
| 441 |
+
"eval_runtime": 124.4448,
|
| 442 |
+
"eval_samples_per_second": 10.141,
|
| 443 |
+
"eval_steps_per_second": 1.27,
|
| 444 |
+
"step": 1450
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 4.225352112676056,
|
| 448 |
+
"grad_norm": 129.88868713378906,
|
| 449 |
+
"learning_rate": 4.789272030651341e-05,
|
| 450 |
+
"loss": 26.3816,
|
| 451 |
+
"step": 1500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 4.225352112676056,
|
| 455 |
+
"eval_loss": 29.6898193359375,
|
| 456 |
+
"eval_runtime": 129.216,
|
| 457 |
+
"eval_samples_per_second": 9.767,
|
| 458 |
+
"eval_steps_per_second": 1.223,
|
| 459 |
+
"step": 1500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 4.366197183098592,
|
| 463 |
+
"grad_norm": 138.11952209472656,
|
| 464 |
+
"learning_rate": 4.782228983547442e-05,
|
| 465 |
+
"loss": 26.5218,
|
| 466 |
+
"step": 1550
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.366197183098592,
|
| 470 |
+
"eval_loss": 29.94746971130371,
|
| 471 |
+
"eval_runtime": 132.27,
|
| 472 |
+
"eval_samples_per_second": 9.541,
|
| 473 |
+
"eval_steps_per_second": 1.195,
|
| 474 |
+
"step": 1550
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 4.507042253521127,
|
| 478 |
+
"grad_norm": 185.40530395507812,
|
| 479 |
+
"learning_rate": 4.775185936443543e-05,
|
| 480 |
+
"loss": 26.9798,
|
| 481 |
+
"step": 1600
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 4.507042253521127,
|
| 485 |
+
"eval_loss": 29.332275390625,
|
| 486 |
+
"eval_runtime": 137.9492,
|
| 487 |
+
"eval_samples_per_second": 9.148,
|
| 488 |
+
"eval_steps_per_second": 1.145,
|
| 489 |
+
"step": 1600
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.647887323943662,
|
| 493 |
+
"grad_norm": 135.86349487304688,
|
| 494 |
+
"learning_rate": 4.768142889339644e-05,
|
| 495 |
+
"loss": 26.8186,
|
| 496 |
+
"step": 1650
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.647887323943662,
|
| 500 |
+
"eval_loss": 29.575531005859375,
|
| 501 |
+
"eval_runtime": 135.1713,
|
| 502 |
+
"eval_samples_per_second": 9.336,
|
| 503 |
+
"eval_steps_per_second": 1.169,
|
| 504 |
+
"step": 1650
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 4.788732394366197,
|
| 508 |
+
"grad_norm": 153.70196533203125,
|
| 509 |
+
"learning_rate": 4.761099842235745e-05,
|
| 510 |
+
"loss": 27.5111,
|
| 511 |
+
"step": 1700
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 4.788732394366197,
|
| 515 |
+
"eval_loss": 30.7945499420166,
|
| 516 |
+
"eval_runtime": 117.37,
|
| 517 |
+
"eval_samples_per_second": 10.752,
|
| 518 |
+
"eval_steps_per_second": 1.346,
|
| 519 |
+
"step": 1700
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 4.929577464788732,
|
| 523 |
+
"grad_norm": 150.87384033203125,
|
| 524 |
+
"learning_rate": 4.754056795131846e-05,
|
| 525 |
+
"loss": 27.0839,
|
| 526 |
+
"step": 1750
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.929577464788732,
|
| 530 |
+
"eval_loss": 29.414661407470703,
|
| 531 |
+
"eval_runtime": 120.7787,
|
| 532 |
+
"eval_samples_per_second": 10.449,
|
| 533 |
+
"eval_steps_per_second": 1.308,
|
| 534 |
+
"step": 1750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 5.070422535211268,
|
| 538 |
+
"grad_norm": 140.32867431640625,
|
| 539 |
+
"learning_rate": 4.747013748027947e-05,
|
| 540 |
+
"loss": 26.6393,
|
| 541 |
+
"step": 1800
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 5.070422535211268,
|
| 545 |
+
"eval_loss": 28.79827880859375,
|
| 546 |
+
"eval_runtime": 126.6565,
|
| 547 |
+
"eval_samples_per_second": 9.964,
|
| 548 |
+
"eval_steps_per_second": 1.247,
|
| 549 |
+
"step": 1800
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 5.211267605633803,
|
| 553 |
+
"grad_norm": 105.2396469116211,
|
| 554 |
+
"learning_rate": 4.739970700924048e-05,
|
| 555 |
+
"loss": 26.3564,
|
| 556 |
+
"step": 1850
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 5.211267605633803,
|
| 560 |
+
"eval_loss": 29.22454261779785,
|
| 561 |
+
"eval_runtime": 138.64,
|
| 562 |
+
"eval_samples_per_second": 9.103,
|
| 563 |
+
"eval_steps_per_second": 1.14,
|
| 564 |
+
"step": 1850
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 5.352112676056338,
|
| 568 |
+
"grad_norm": 158.32907104492188,
|
| 569 |
+
"learning_rate": 4.732927653820149e-05,
|
| 570 |
+
"loss": 25.6174,
|
| 571 |
+
"step": 1900
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 5.352112676056338,
|
| 575 |
+
"eval_loss": 28.933706283569336,
|
| 576 |
+
"eval_runtime": 124.4051,
|
| 577 |
+
"eval_samples_per_second": 10.144,
|
| 578 |
+
"eval_steps_per_second": 1.27,
|
| 579 |
+
"step": 1900
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 5.492957746478873,
|
| 583 |
+
"grad_norm": 209.70982360839844,
|
| 584 |
+
"learning_rate": 4.72588460671625e-05,
|
| 585 |
+
"loss": 25.8777,
|
| 586 |
+
"step": 1950
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 5.492957746478873,
|
| 590 |
+
"eval_loss": 29.477840423583984,
|
| 591 |
+
"eval_runtime": 138.0785,
|
| 592 |
+
"eval_samples_per_second": 9.14,
|
| 593 |
+
"eval_steps_per_second": 1.144,
|
| 594 |
+
"step": 1950
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 5.633802816901408,
|
| 598 |
+
"grad_norm": 172.13571166992188,
|
| 599 |
+
"learning_rate": 4.718841559612351e-05,
|
| 600 |
+
"loss": 25.6848,
|
| 601 |
+
"step": 2000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 5.633802816901408,
|
| 605 |
+
"eval_loss": 28.4991512298584,
|
| 606 |
+
"eval_runtime": 154.121,
|
| 607 |
+
"eval_samples_per_second": 8.188,
|
| 608 |
+
"eval_steps_per_second": 1.025,
|
| 609 |
+
"step": 2000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 5.774647887323944,
|
| 613 |
+
"grad_norm": 133.943115234375,
|
| 614 |
+
"learning_rate": 4.711798512508452e-05,
|
| 615 |
+
"loss": 26.4625,
|
| 616 |
+
"step": 2050
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 5.774647887323944,
|
| 620 |
+
"eval_loss": 29.618194580078125,
|
| 621 |
+
"eval_runtime": 118.4654,
|
| 622 |
+
"eval_samples_per_second": 10.653,
|
| 623 |
+
"eval_steps_per_second": 1.334,
|
| 624 |
+
"step": 2050
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 5.915492957746479,
|
| 628 |
+
"grad_norm": 143.0037078857422,
|
| 629 |
+
"learning_rate": 4.704755465404553e-05,
|
| 630 |
+
"loss": 26.8448,
|
| 631 |
+
"step": 2100
|
| 632 |
+
},
|
| 633 |
+
{
|
| 634 |
+
"epoch": 5.915492957746479,
|
| 635 |
+
"eval_loss": 29.537738800048828,
|
| 636 |
+
"eval_runtime": 144.9122,
|
| 637 |
+
"eval_samples_per_second": 8.709,
|
| 638 |
+
"eval_steps_per_second": 1.09,
|
| 639 |
+
"step": 2100
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 6.056338028169014,
|
| 643 |
+
"grad_norm": 127.10630798339844,
|
| 644 |
+
"learning_rate": 4.697712418300654e-05,
|
| 645 |
+
"loss": 26.0681,
|
| 646 |
+
"step": 2150
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 6.056338028169014,
|
| 650 |
+
"eval_loss": 29.239002227783203,
|
| 651 |
+
"eval_runtime": 131.0446,
|
| 652 |
+
"eval_samples_per_second": 9.63,
|
| 653 |
+
"eval_steps_per_second": 1.206,
|
| 654 |
+
"step": 2150
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 6.197183098591549,
|
| 658 |
+
"grad_norm": 142.51170349121094,
|
| 659 |
+
"learning_rate": 4.690669371196755e-05,
|
| 660 |
+
"loss": 25.628,
|
| 661 |
+
"step": 2200
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 6.197183098591549,
|
| 665 |
+
"eval_loss": 29.1112003326416,
|
| 666 |
+
"eval_runtime": 119.9659,
|
| 667 |
+
"eval_samples_per_second": 10.52,
|
| 668 |
+
"eval_steps_per_second": 1.317,
|
| 669 |
+
"step": 2200
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 6.197183098591549,
|
| 673 |
+
"step": 2200,
|
| 674 |
+
"total_flos": 8.850907754333798e+18,
|
| 675 |
+
"train_loss": 29.80798134543679,
|
| 676 |
+
"train_runtime": 17574.1061,
|
| 677 |
+
"train_samples_per_second": 64.612,
|
| 678 |
+
"train_steps_per_second": 2.02
|
| 679 |
}
|
| 680 |
],
|
| 681 |
"logging_steps": 50,
|
|
|
|
| 690 |
"early_stopping_threshold": 0.0
|
| 691 |
},
|
| 692 |
"attributes": {
|
| 693 |
+
"early_stopping_patience_counter": 4
|
| 694 |
}
|
| 695 |
},
|
| 696 |
"TrainerControl": {
|
|
|
|
| 699 |
"should_evaluate": false,
|
| 700 |
"should_log": false,
|
| 701 |
"should_save": true,
|
| 702 |
+
"should_training_stop": true
|
| 703 |
},
|
| 704 |
"attributes": {}
|
| 705 |
}
|
| 706 |
},
|
| 707 |
+
"total_flos": 8.850907754333798e+18,
|
| 708 |
"train_batch_size": 8,
|
| 709 |
"trial_name": null,
|
| 710 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5240
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e9da334fac8e5bddd30bc32e51ac79bfbfb26f83868f0b31c3f99b8c64a3f17
|
| 3 |
size 5240
|