dit-base-finetuned-rvlcdip

This model is a fine-tuned version of microsoft/dit-base-finetuned-rvlcdip on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.5940
• Accuracy: 0.87
• Precision: 0.7623
• Recall: 0.87
• F1 Score: 0.8126

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1 Score
No log	1.0	4	0.6006	0.8708	0.7584	0.8708	0.8107
No log	2.0	8	0.5169	0.8708	0.7584	0.8708	0.8107
No log	3.0	12	0.4027	0.8708	0.7584	0.8708	0.8107
0.5427	4.0	16	0.3865	0.8708	0.7584	0.8708	0.8107
0.5427	5.0	20	0.3894	0.8708	0.7584	0.8708	0.8107
0.5427	6.0	24	0.3729	0.8708	0.7584	0.8708	0.8107
0.5427	7.0	28	0.3707	0.8708	0.7584	0.8708	0.8107
0.4458	8.0	32	0.3790	0.8708	0.7584	0.8708	0.8107
0.4458	9.0	36	0.3504	0.8708	0.7584	0.8708	0.8107
0.4458	10.0	40	0.3356	0.8708	0.7584	0.8708	0.8107
0.4458	11.0	44	0.4082	0.8708	0.7584	0.8708	0.8107
0.4369	12.0	48	0.3455	0.8708	0.7584	0.8708	0.8107
0.4369	13.0	52	0.3074	0.8708	0.7584	0.8708	0.8107
0.4369	14.0	56	0.3097	0.8708	0.7584	0.8708	0.8107
0.4109	15.0	60	0.3173	0.8708	0.7584	0.8708	0.8107

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3