Hugging Face's logo

bdpc
/
resnet101-base_tobacco-cnn_tobacco3482_kd_CEKD_t5.0_a0.9

Image Classification
Transformers
PyTorch
resnet
Generated from Trainer
Inference Endpoints
Model card Files Community
1
resnet101-base_tobacco-cnn_tobacco3482_kd_CEKD_t5.0_a0.9 / README.md
bdpc's picture
bdpc
Saving best model to hub
64f09f3
preview code
|
raw
history blame
7.84 kB
metadata
license: apache-2.0
base_model: microsoft/resnet-50
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: resnet101-base_tobacco-cnn_tobacco3482_kd_CEKD_t5.0_a0.9
    results: []

resnet101-base_tobacco-cnn_tobacco3482_kd_CEKD_t5.0_a0.9

This model is a fine-tuned version of microsoft/resnet-50 on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.8862
  • Accuracy: 0.675
  • Brier Loss: 0.4233
  • Nll: 2.4267
  • F1 Micro: 0.675
  • F1 Macro: 0.6266
  • Ece: 0.2528
  • Aurc: 0.1205

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 0.0001
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Accuracy Brier Loss Nll F1 Micro F1 Macro Ece Aurc
No log 1.0 13 2.1186 0.165 0.8967 8.5414 0.165 0.1128 0.2087 0.8330
No log 2.0 26 2.1139 0.14 0.8960 8.0889 0.14 0.0907 0.1924 0.8318
No log 3.0 39 2.0743 0.195 0.8880 6.6316 0.195 0.1098 0.2224 0.7879
No log 4.0 52 2.0101 0.205 0.8741 6.0411 0.205 0.0851 0.2448 0.7302
No log 5.0 65 1.9697 0.22 0.8650 5.8808 0.22 0.1090 0.2441 0.7307
No log 6.0 78 1.8642 0.27 0.8396 6.0693 0.27 0.1370 0.2742 0.6623
No log 7.0 91 1.7716 0.35 0.8100 5.7342 0.35 0.1964 0.3131 0.4496
No log 8.0 104 1.7580 0.33 0.8084 5.8902 0.33 0.1762 0.3185 0.5663
No log 9.0 117 1.7346 0.425 0.8000 5.5871 0.425 0.2645 0.3466 0.3888
No log 10.0 130 1.6557 0.365 0.7744 5.2246 0.3650 0.2256 0.2890 0.5081
No log 11.0 143 1.5067 0.46 0.7014 4.7492 0.46 0.3053 0.3024 0.2923
No log 12.0 156 1.5340 0.425 0.7212 4.4923 0.425 0.2746 0.2833 0.3650
No log 13.0 169 1.5064 0.495 0.7111 4.1576 0.495 0.3443 0.3225 0.2907
No log 14.0 182 1.4767 0.54 0.6972 3.7984 0.54 0.3804 0.3381 0.2831
No log 15.0 195 1.3709 0.525 0.6453 3.7435 0.525 0.3771 0.3188 0.2541
No log 16.0 208 1.3204 0.535 0.6223 3.4971 0.535 0.3919 0.3115 0.2424
No log 17.0 221 1.4782 0.465 0.7008 3.4793 0.465 0.3731 0.3311 0.4138
No log 18.0 234 1.3456 0.49 0.6523 3.3409 0.49 0.3839 0.2832 0.3570
No log 19.0 247 1.2137 0.625 0.5708 3.4778 0.625 0.5087 0.3030 0.1904
No log 20.0 260 1.3527 0.565 0.6484 3.3840 0.565 0.4761 0.3402 0.3349
No log 21.0 273 1.1692 0.6 0.5633 2.9586 0.6 0.4932 0.3138 0.2299
No log 22.0 286 1.1144 0.65 0.5253 2.9930 0.65 0.5281 0.2768 0.1585
No log 23.0 299 1.0749 0.635 0.5048 2.8481 0.635 0.5404 0.2378 0.1642
No log 24.0 312 1.0619 0.665 0.5018 2.7665 0.665 0.5653 0.2741 0.1533
No log 25.0 325 1.0733 0.68 0.5036 2.6592 0.68 0.5960 0.2948 0.1633
No log 26.0 338 1.0319 0.655 0.4930 2.6467 0.655 0.5786 0.2598 0.1576
No log 27.0 351 1.0147 0.665 0.4805 2.6123 0.665 0.5877 0.2406 0.1405
No log 28.0 364 0.9862 0.675 0.4734 2.4990 0.675 0.5876 0.2512 0.1474
No log 29.0 377 0.9816 0.685 0.4696 2.5984 0.685 0.6131 0.2446 0.1428
No log 30.0 390 0.9755 0.66 0.4698 2.5609 0.66 0.6009 0.2562 0.1555
No log 31.0 403 0.9789 0.7 0.4601 2.6827 0.7 0.6374 0.2667 0.1271
No log 32.0 416 0.9426 0.695 0.4501 2.5256 0.695 0.6315 0.2560 0.1420
No log 33.0 429 0.9428 0.695 0.4461 2.6429 0.695 0.6298 0.2250 0.1243
No log 34.0 442 0.9370 0.675 0.4455 2.5812 0.675 0.6061 0.2523 0.1284
No log 35.0 455 0.9290 0.68 0.4391 2.3724 0.68 0.6174 0.2459 0.1361
No log 36.0 468 0.9190 0.66 0.4393 2.3838 0.66 0.6140 0.2201 0.1327
No log 37.0 481 0.9061 0.685 0.4310 2.3683 0.685 0.6390 0.2222 0.1196
No log 38.0 494 0.9184 0.705 0.4387 2.5054 0.705 0.6444 0.2479 0.1191
1.0876 39.0 507 0.9185 0.685 0.4425 2.4429 0.685 0.6327 0.2450 0.1337
1.0876 40.0 520 0.9002 0.66 0.4289 2.4439 0.66 0.6226 0.2152 0.1302
1.0876 41.0 533 0.9027 0.68 0.4319 2.3802 0.68 0.6179 0.2247 0.1200
1.0876 42.0 546 0.8977 0.68 0.4321 2.3577 0.68 0.6195 0.2296 0.1250
1.0876 43.0 559 0.8861 0.685 0.4215 2.3150 0.685 0.6324 0.1870 0.1198
1.0876 44.0 572 0.8913 0.68 0.4235 2.4228 0.68 0.6328 0.2260 0.1193
1.0876 45.0 585 0.8895 0.675 0.4251 2.4104 0.675 0.6264 0.2409 0.1208
1.0876 46.0 598 0.8901 0.665 0.4223 2.3598 0.665 0.6146 0.2235 0.1208
1.0876 47.0 611 0.8809 0.68 0.4206 2.3528 0.68 0.6233 0.2306 0.1222
1.0876 48.0 624 0.8845 0.69 0.4243 2.4251 0.69 0.6362 0.2232 0.1219
1.0876 49.0 637 0.8849 0.675 0.4243 2.4261 0.675 0.6207 0.2192 0.1242
1.0876 50.0 650 0.8862 0.675 0.4233 2.4267 0.675 0.6266 0.2528 0.1205

Framework versions

  • Transformers 4.33.3
  • Pytorch 2.2.0.dev20231002
  • Datasets 2.7.1
  • Tokenizers 0.13.3