clip-vit-base-patch32-finetuned-eurosat

This model is a fine-tuned version of openai/clip-vit-base-patch32 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0987
• Accuracy: 0.9716

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.4295	0.9979	351	0.2629	0.915
0.4167	1.9986	703	0.2365	0.9222
0.4104	2.9993	1055	0.2205	0.9252
0.3847	4.0	1407	0.1917	0.9338
0.3928	4.9979	1758	0.1803	0.9414
0.311	5.9986	2110	0.1429	0.9524
0.2614	6.9993	2462	0.1137	0.961
0.2579	8.0	2814	0.1102	0.9638
0.1993	8.9979	3165	0.1037	0.9688
0.1921	9.9787	3510	0.0987	0.9716

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1