beans_ViT

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

• Loss: 1.2997
• Accuracy: 0.7969
• F1: 0.7991

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
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.0	17	0.8693	0.6090	0.5953
No log	2.0	34	0.9652	0.6015	0.5977
No log	3.0	51	0.7178	0.6992	0.6927
No log	4.0	68	0.7488	0.6992	0.6955
No log	5.0	85	0.6517	0.7068	0.7070
No log	6.0	102	0.7816	0.6842	0.6541
No log	7.0	119	0.5014	0.7744	0.7733
No log	8.0	136	0.5321	0.7669	0.7680
No log	9.0	153	0.5985	0.7444	0.7457
No log	10.0	170	0.4675	0.8271	0.8274
No log	11.0	187	0.5750	0.7744	0.7576
No log	12.0	204	0.6617	0.7293	0.7066
No log	13.0	221	0.6396	0.7594	0.7577
No log	14.0	238	0.4302	0.8346	0.8352
No log	15.0	255	0.4018	0.8421	0.8427
No log	16.0	272	0.5673	0.7895	0.7883
No log	17.0	289	0.5037	0.8120	0.8097
No log	18.0	306	0.5939	0.8496	0.8487
No log	19.0	323	0.6590	0.8120	0.8111
No log	20.0	340	0.6060	0.8571	0.8559
No log	21.0	357	0.5806	0.8421	0.8418
No log	22.0	374	0.6180	0.8421	0.8414
No log	23.0	391	0.7707	0.7669	0.7633
No log	24.0	408	0.5440	0.8421	0.8418
No log	25.0	425	0.6596	0.8496	0.8497
No log	26.0	442	0.5393	0.8346	0.8342
No log	27.0	459	0.6320	0.8797	0.8795
No log	28.0	476	0.5903	0.8496	0.8507
No log	29.0	493	0.6826	0.8647	0.8644
0.3346	30.0	510	0.6493	0.8571	0.8567

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.3.1
• Tokenizers 0.21.0