Hugging Face's logo

vuiseng9
/
swin-base-food101-int8-structured43-15eph

Image Classification Transformers PyTorch OpenVINO
swin vision generated_from_trainer Eval Results Inference Endpoints
Model card Files Community
swin-base-food101-int8-structured43-15eph / README.md
Vui Seng Chua
Add content
31bfa4b
preview code
|
raw history blame contribute delete
No virus
4.01 kB
metadata
license: apache-2.0
tags:
  - image-classification
  - vision
  - generated_from_trainer
datasets:
  - food101
metrics:
  - accuracy
model-index:
  - name: jpqd-swin-b-15eph-r1.00-s2e5-mock-main-merge-pr2
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: food101
          type: food101
          config: default
          split: validation
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.9144158415841585

jpqd-swin-b-15eph-r1.00-s2e5-mock-main-merge-pr2

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224 on the food101 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.2970
  • Accuracy: 0.9144

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: 16
  • eval_batch_size: 128
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 15.0

Training results

Training Loss Epoch Step Validation Loss Accuracy
3.8787 0.42 500 3.9971 0.7163
0.8429 0.84 1000 0.6450 0.8678
0.8561 1.27 1500 0.4160 0.8945
0.5777 1.69 2000 0.3664 0.9006
12.3601 2.11 2500 12.0328 0.9023
49.0606 2.54 3000 48.5000 0.8526
75.3173 2.96 3500 75.5341 0.6942
93.6153 3.38 4000 93.3091 0.5929
103.5744 3.8 4500 103.1211 0.5846
107.7701 4.23 5000 108.0755 0.5398
109.5736 4.65 5500 108.7624 0.5855
1.8028 5.07 6000 1.0960 0.8179
1.2549 5.49 6500 0.6560 0.8695
0.7199 5.92 7000 0.5619 0.8769
0.8874 6.34 7500 0.5151 0.8859
0.7429 6.76 8000 0.4830 0.8898
0.6759 7.19 8500 0.4681 0.8926
0.5352 7.61 9000 0.4360 0.8956
0.6021 8.03 9500 0.4202 0.8979
0.5617 8.45 10000 0.3940 0.9003
0.7235 8.88 10500 0.3915 0.9000
0.5323 9.3 11000 0.3793 0.9017
0.589 9.72 11500 0.3670 0.9051
0.425 10.14 12000 0.3615 0.9059
0.7103 10.57 12500 0.3479 0.9070
0.6251 10.99 13000 0.3472 0.9073
0.623 11.41 13500 0.3353 0.9088
0.6012 11.83 14000 0.3292 0.9098
0.4984 12.26 14500 0.3230 0.9112
0.4763 12.68 15000 0.3158 0.9109
0.3209 13.1 15500 0.3120 0.9123
0.4854 13.52 16000 0.3057 0.9126
0.5472 13.95 16500 0.3032 0.9134
0.3264 14.37 17000 0.3013 0.9134
0.4136 14.79 17500 0.2977 0.9141

Framework versions

  • Transformers 4.26.1
  • Pytorch 1.13.1+cu117
  • Datasets 2.10.1
  • Tokenizers 0.13.2