metadata

license: apache-2.0
base_model: microsoft/swinv2-tiny-patch4-window8-256
tags:
  - generated_from_trainer
datasets:
  - imagefolder
metrics:
  - accuracy
model-index:
  - name: swinv2-tiny-patch4-window8-256-RH
    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: 0.6448598130841121

swinv2-tiny-patch4-window8-256-RH

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 0.6681
Accuracy: 0.6449

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: 0.00015
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	1.0	8	4.5659	0.4112
4.5175	2.0	16	3.6362	0.4112
3.9284	3.0	24	1.6019	0.4112
1.6086	4.0	32	0.7110	0.4112
0.7392	5.0	40	0.6825	0.5888
0.7392	6.0	48	0.6795	0.5888
0.7073	7.0	56	0.6814	0.5888
0.6956	8.0	64	0.7061	0.5888
0.6898	9.0	72	0.7014	0.5888
0.7026	10.0	80	0.7214	0.4112
0.7026	11.0	88	0.7186	0.5888
0.7696	12.0	96	0.6837	0.5888
0.6909	13.0	104	0.6823	0.5888
0.6799	14.0	112	0.6781	0.5888
0.6782	15.0	120	0.6938	0.5888
0.6782	16.0	128	0.6766	0.5888
0.6952	17.0	136	0.7123	0.5888
0.6875	18.0	144	0.6891	0.5607
0.6919	19.0	152	0.7076	0.5888
0.6751	20.0	160	0.7011	0.4953
0.6751	21.0	168	0.6962	0.5888
0.689	22.0	176	0.6857	0.5701
0.6826	23.0	184	0.6935	0.5888
0.6841	24.0	192	0.7219	0.5888
0.6657	25.0	200	0.6610	0.5888
0.6657	26.0	208	0.6681	0.6449
0.6524	27.0	216	0.7225	0.5888
0.6567	28.0	224	0.7117	0.5888
0.6402	29.0	232	0.6999	0.6262
0.66	30.0	240	0.6799	0.6075
0.66	31.0	248	0.6677	0.6075
0.6469	32.0	256	0.6735	0.5981
0.6355	33.0	264	0.6853	0.6168
0.6245	34.0	272	0.7008	0.6262
0.6306	35.0	280	0.6990	0.5981
0.6306	36.0	288	0.6981	0.6355
0.6208	37.0	296	0.7103	0.6262
0.6339	38.0	304	0.7050	0.6355
0.5959	39.0	312	0.6989	0.6355
0.6059	40.0	320	0.6990	0.6355

Framework versions

Transformers 4.36.2
Pytorch 2.1.2+cu118
Datasets 2.16.1
Tokenizers 0.15.0