Hugging Face's logo

amaye15
/
microsoft-resnet-50-batch32-lr0.0005-standford-dogs

Image Classification
Transformers
Safetensors
resnet
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
microsoft-resnet-50-batch32-lr0.0005-standford-dogs / README.md
amaye15's picture
amaye15
End of training
692e598 verified
preview code
|
raw
history blame
11.4 kB
metadata
license: apache-2.0
base_model: microsoft/resnet-50
tags:
  - generated_from_trainer
datasets:
  - stanford-dogs
metrics:
  - accuracy
  - f1
  - precision
  - recall
model-index:
  - name: microsoft-resnet-50-batch32-lr0.0005-standford-dogs
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: stanford-dogs
          type: stanford-dogs
          config: default
          split: full
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8386783284742468
          - name: F1
            type: f1
            value: 0.8259546998355447
          - name: Precision
            type: precision
            value: 0.8457483127517197
          - name: Recall
            type: recall
            value: 0.8314858626273427

microsoft-resnet-50-batch32-lr0.0005-standford-dogs

This model is a fine-tuned version of microsoft/resnet-50 on the stanford-dogs dataset. It achieves the following results on the evaluation set:

  • Loss: 1.1545
  • Accuracy: 0.8387
  • F1: 0.8260
  • Precision: 0.8457
  • Recall: 0.8315

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: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • training_steps: 1000

Training results

Training Loss Epoch Step Validation Loss Accuracy F1 Precision Recall
4.7829 0.0777 10 4.7747 0.2119 0.1874 0.3919 0.1982
4.7714 0.1553 20 4.7572 0.2038 0.1842 0.4262 0.1836
4.7606 0.2330 30 4.7367 0.3586 0.3433 0.6517 0.3307
4.747 0.3107 40 4.7149 0.4303 0.4272 0.7734 0.4039
4.7253 0.3883 50 4.6846 0.4361 0.4678 0.7906 0.4160
4.7069 0.4660 60 4.6534 0.5330 0.5397 0.8048 0.5093
4.6857 0.5437 70 4.6177 0.5500 0.5511 0.7998 0.5264
4.6569 0.6214 80 4.5764 0.5739 0.5800 0.8208 0.5517
4.6293 0.6990 90 4.5359 0.6142 0.6149 0.8075 0.5926
4.5953 0.7767 100 4.4828 0.6207 0.6233 0.8109 0.6000
4.5651 0.8544 110 4.4257 0.6591 0.6585 0.8148 0.6393
4.5296 0.9320 120 4.3647 0.7063 0.7012 0.8284 0.6882
4.4911 1.0097 130 4.2998 0.7089 0.7074 0.8326 0.6924
4.4442 1.0874 140 4.2288 0.6939 0.6890 0.8302 0.6759
4.3912 1.1650 150 4.1527 0.6873 0.6863 0.8262 0.6703
4.3393 1.2427 160 4.0884 0.7250 0.7127 0.8251 0.7082
4.3019 1.3204 170 3.9946 0.7262 0.7152 0.8234 0.7098
4.2366 1.3981 180 3.9314 0.7301 0.7177 0.8230 0.7143
4.1966 1.4757 190 3.8398 0.7325 0.7196 0.8169 0.7175
4.1402 1.5534 200 3.7587 0.7381 0.7217 0.8149 0.7221
4.0771 1.6311 210 3.6745 0.7310 0.7149 0.8125 0.7160
4.0436 1.7087 220 3.5729 0.7364 0.7189 0.8121 0.7214
3.9697 1.7864 230 3.5030 0.7490 0.7339 0.8172 0.7358
3.9181 1.8641 240 3.4505 0.7541 0.7379 0.8123 0.7408
3.8573 1.9417 250 3.3529 0.7646 0.7453 0.8136 0.7521
3.8077 2.0194 260 3.2566 0.7660 0.7482 0.8093 0.7540
3.7449 2.0971 270 3.1869 0.7709 0.7510 0.8144 0.7588
3.682 2.1748 280 3.0898 0.7668 0.7440 0.8097 0.7548
3.6461 2.2524 290 3.0377 0.7641 0.7381 0.8100 0.7511
3.6004 2.3301 300 2.9001 0.7648 0.7384 0.8061 0.7522
3.5478 2.4078 310 2.8623 0.7653 0.7410 0.8060 0.7529
3.4971 2.4854 320 2.7961 0.7675 0.7447 0.8068 0.7558
3.4446 2.5631 330 2.6960 0.7690 0.7486 0.8128 0.7582
3.4093 2.6408 340 2.6480 0.7821 0.7652 0.8151 0.7718
3.3994 2.7184 350 2.5330 0.7847 0.7676 0.8156 0.7742
3.2963 2.7961 360 2.4866 0.7855 0.7681 0.8154 0.7752
3.2615 2.8738 370 2.4344 0.7891 0.7740 0.8172 0.7792
3.2024 2.9515 380 2.4011 0.7794 0.7638 0.8126 0.7694
3.1641 3.0291 390 2.3039 0.7835 0.7659 0.8100 0.7736
3.0719 3.1068 400 2.2471 0.7796 0.7608 0.8072 0.7691
3.0808 3.1845 410 2.2130 0.7896 0.7717 0.8137 0.7795
2.9916 3.2621 420 2.1387 0.7823 0.7652 0.8104 0.7718
2.9898 3.3398 430 2.0905 0.7981 0.7821 0.8250 0.7886
2.9597 3.4175 440 2.0260 0.7923 0.7769 0.8192 0.7826
2.9068 3.4951 450 1.9944 0.7976 0.7816 0.8233 0.7877
2.8423 3.5728 460 1.9643 0.7976 0.7805 0.8185 0.7876
2.8323 3.6505 470 1.8926 0.7935 0.7754 0.8136 0.7837
2.7814 3.7282 480 1.8676 0.8017 0.7856 0.8208 0.7917
2.7337 3.8058 490 1.8320 0.8052 0.7905 0.8246 0.7957
2.7215 3.8835 500 1.8003 0.7986 0.7834 0.8208 0.7890
2.6456 3.9612 510 1.7754 0.8005 0.7848 0.8230 0.7914
2.6494 4.0388 520 1.7083 0.8054 0.7895 0.8252 0.7967
2.5878 4.1165 530 1.6836 0.8054 0.7878 0.8239 0.7967
2.592 4.1942 540 1.6770 0.8005 0.7826 0.8220 0.7912
2.5698 4.2718 550 1.6184 0.8056 0.7881 0.8268 0.7970
2.52 4.3495 560 1.6368 0.8064 0.7898 0.8267 0.7975
2.5317 4.4272 570 1.5952 0.8059 0.7891 0.8289 0.7972
2.4199 4.5049 580 1.5518 0.8163 0.8002 0.8337 0.8082
2.4357 4.5825 590 1.5375 0.8095 0.7933 0.8263 0.8012
2.4217 4.6602 600 1.4994 0.8127 0.7964 0.8297 0.8042
2.428 4.7379 610 1.4671 0.8156 0.8003 0.8309 0.8074
2.3725 4.8155 620 1.4402 0.8141 0.7973 0.8295 0.8054
2.3594 4.8932 630 1.4566 0.8134 0.7976 0.8287 0.8049
2.3279 4.9709 640 1.4359 0.8183 0.8034 0.8314 0.8100
2.3166 5.0485 650 1.4067 0.8226 0.8086 0.8343 0.8149
2.3062 5.1262 660 1.3913 0.8212 0.8072 0.8340 0.8131
2.3096 5.2039 670 1.3577 0.8241 0.8107 0.8373 0.8159
2.2514 5.2816 680 1.3574 0.8270 0.8136 0.8371 0.8193
2.2053 5.3592 690 1.3450 0.8239 0.8101 0.8370 0.8164
2.2347 5.4369 700 1.3331 0.8270 0.8137 0.8388 0.8194
2.215 5.5146 710 1.2902 0.8294 0.8154 0.8419 0.8219
2.175 5.5922 720 1.2861 0.8256 0.8114 0.8388 0.8181
2.2212 5.6699 730 1.2637 0.8321 0.8180 0.8440 0.8241
2.1459 5.7476 740 1.2827 0.8302 0.8166 0.8396 0.8227
2.1615 5.8252 750 1.2800 0.8311 0.8184 0.8496 0.8239
2.0966 5.9029 760 1.2742 0.8326 0.8195 0.8418 0.8251
2.1314 5.9806 770 1.2464 0.8316 0.8184 0.8407 0.8238
2.0846 6.0583 780 1.2409 0.8326 0.8189 0.8414 0.8250
2.0522 6.1359 790 1.2023 0.8365 0.8233 0.8455 0.8292
2.0724 6.2136 800 1.2252 0.8309 0.8174 0.8396 0.8235
2.0848 6.2913 810 1.2025 0.8321 0.8186 0.8424 0.8248
2.0402 6.3689 820 1.2130 0.8333 0.8189 0.8428 0.8255
2.0778 6.4466 830 1.1809 0.8375 0.8249 0.8532 0.8302
2.0963 6.5243 840 1.1696 0.8365 0.8231 0.8527 0.8289
2.0576 6.6019 850 1.1866 0.8321 0.8181 0.8411 0.8245
2.0386 6.6796 860 1.1882 0.8302 0.8160 0.8389 0.8227
2.0084 6.7573 870 1.1696 0.8372 0.8244 0.8446 0.8301
2.0571 6.8350 880 1.1622 0.8353 0.8217 0.8437 0.8280
2.0264 6.9126 890 1.1640 0.8336 0.8204 0.8429 0.8263
2.0077 6.9903 900 1.1673 0.8367 0.8241 0.8447 0.8295
2.0492 7.0680 910 1.1455 0.8404 0.8269 0.8462 0.8330
1.9973 7.1456 920 1.1538 0.8379 0.8250 0.8455 0.8307
1.9961 7.2233 930 1.1502 0.8367 0.8236 0.8415 0.8295
1.9681 7.3010 940 1.1657 0.8384 0.8254 0.8463 0.8311
2.0188 7.3786 950 1.1309 0.8379 0.8252 0.8445 0.8310
2.0225 7.4563 960 1.1547 0.8367 0.8231 0.8446 0.8294
1.9562 7.5340 970 1.1474 0.8377 0.8243 0.8457 0.8305
2.0247 7.6117 980 1.1251 0.8365 0.8241 0.8449 0.8294
1.9355 7.6893 990 1.1349 0.8397 0.8276 0.8532 0.8329
1.9804 7.7670 1000 1.1545 0.8387 0.8260 0.8457 0.8315

Framework versions

  • Transformers 4.40.2
  • Pytorch 2.3.0
  • Datasets 2.19.1
  • Tokenizers 0.19.1