README.md

metadata

base_model: MBZUAI/swiftformer-xs
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: swiftformer-xs-DMAE
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: imagefolder
          type: imagefolder
          config: default
          split: validation
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.717391304347826

swiftformer-xs-DMAE

This model is a fine-tuned version of MBZUAI/swiftformer-xs on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.8492
• Accuracy: 0.7174

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: 0.0003
• train_batch_size: 16
• eval_batch_size: 16
• 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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.86	3	1.3859	0.3261
No log	2.0	7	1.3806	0.5217
1.3829	2.86	10	1.3677	0.5870
1.3829	4.0	14	1.3037	0.6739
1.3829	4.86	17	1.2540	0.6522
1.3074	6.0	21	1.1509	0.6522
1.3074	6.86	24	1.0882	0.6739
1.3074	8.0	28	1.0569	0.6522
1.1437	8.86	31	1.0536	0.6304
1.1437	10.0	35	0.9993	0.6522
1.1437	10.86	38	0.9819	0.6739
1.0439	12.0	42	0.9593	0.6957
1.0439	12.86	45	0.9359	0.6304
1.0439	14.0	49	0.9467	0.6522
0.9813	14.86	52	0.9331	0.6522
0.9813	16.0	56	0.9386	0.6522
0.9813	16.86	59	0.9266	0.6739
0.8763	18.0	63	0.8833	0.6957
0.8763	18.86	66	0.8679	0.6739
0.9187	20.0	70	0.8639	0.6957
0.9187	20.86	73	0.8492	0.7174
0.9187	22.0	77	0.8846	0.6957
0.8067	22.86	80	0.9083	0.6522
0.8067	24.0	84	0.9269	0.6522
0.8067	24.86	87	0.8849	0.6739
0.7248	26.0	91	0.8935	0.6522
0.7248	26.86	94	0.8719	0.6957
0.7248	28.0	98	0.8759	0.6739
0.773	28.86	101	0.8887	0.6739
0.773	30.0	105	0.9288	0.6522
0.773	30.86	108	0.9041	0.6522
0.7467	32.0	112	0.9017	0.6522
0.7467	32.86	115	0.8767	0.6522
0.7467	34.0	119	0.8993	0.6739
0.7323	34.29	120	0.8946	0.6522

Framework versions

• Transformers 4.36.2
• Pytorch 2.1.2+cu118
• Datasets 2.16.1
• Tokenizers 0.15.0