etsy-room-classifier

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5172
• Accuracy: 0.8614

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 2025
• distributed_type: multi-GPU
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.9286	1.0	90	1.8959	0.4785
1.3979	2.0	180	1.3772	0.7393
1.0135	3.0	270	1.0012	0.7987
0.8192	4.0	360	0.8217	0.8284
0.7478	5.0	450	0.7445	0.8251
0.6672	6.0	540	0.7037	0.8383
0.6584	7.0	630	0.6666	0.8449
0.5567	8.0	720	0.6131	0.8548
0.617	9.0	810	0.5836	0.8515
0.4916	10.0	900	0.5588	0.8713
0.5411	11.0	990	0.5741	0.8383
0.5194	12.0	1080	0.5511	0.8548
0.484	13.0	1170	0.5406	0.8614
0.4906	14.0	1260	0.5187	0.8482
0.4461	15.0	1350	0.4862	0.8812
0.408	16.0	1440	0.5168	0.8746
0.465	17.0	1530	0.4795	0.8746
0.4229	18.0	1620	0.5314	0.8482
0.4614	19.0	1710	0.5498	0.8383
0.4443	20.0	1800	0.5172	0.8614

Framework versions

• Transformers 4.54.1
• Pytorch 2.3.1+cu121
• Datasets 2.19.1
• Tokenizers 0.21.4