swinv2-tiny-patch4-window8-256-finetuned-THFOOD-50

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on the THFOOD-50 dataset. It achieves the following results on the:
Train set

• Loss: 0.1669
• Accuracy: 0.9557

Validation set

• Loss: 0.2535
• Accuracy: 0.9344

Test set

• Loss: 0.2669
• Accuracy: 0.9292

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: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.6558	0.99	47	3.1956	0.28
1.705	1.99	94	1.1701	0.6787
0.9805	2.98	141	0.6492	0.8125
0.7925	4.0	189	0.4724	0.8644
0.6169	4.99	236	0.4129	0.8738
0.5343	5.99	283	0.3717	0.8825
0.5196	6.98	330	0.3654	0.8906
0.5059	8.0	378	0.3267	0.8969
0.4432	8.99	425	0.2996	0.9081
0.3819	9.99	472	0.3056	0.9087
0.3627	10.98	519	0.2796	0.9213
0.3505	12.0	567	0.2753	0.915
0.3224	12.99	614	0.2830	0.9206
0.3206	13.99	661	0.2797	0.9231
0.3141	14.98	708	0.2569	0.9287
0.2946	16.0	756	0.2582	0.9319
0.3008	16.99	803	0.2583	0.9337
0.2356	17.99	850	0.2567	0.9281
0.2954	18.98	897	0.2581	0.9319
0.2628	19.89	940	0.2535	0.9344

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3