resnet-18-finetuned-eurosat

This model is a fine-tuned version of microsoft/resnet-18 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5506
• Accuracy: 0.8364

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3846	0.99	20	0.6352	0.7576
0.5009	1.98	40	0.6439	0.7485
0.5625	2.96	60	0.5722	0.7909
0.4928	4.0	81	0.5514	0.7970
0.4621	4.99	101	0.6104	0.7697
0.4367	5.98	121	0.5734	0.7939
0.4238	6.96	141	0.5558	0.8
0.4011	8.0	162	0.5549	0.8030
0.4129	8.99	182	0.5554	0.8061
0.384	9.98	202	0.5551	0.8152
0.3839	10.96	222	0.5742	0.8091
0.3496	12.0	243	0.5518	0.8303
0.3482	12.99	263	0.5390	0.8303
0.357	13.98	283	0.5544	0.8182
0.3341	14.96	303	0.5506	0.8364
0.3605	16.0	324	0.5546	0.8212
0.3041	16.99	344	0.5597	0.8212
0.3364	17.98	364	0.5730	0.8091
0.2976	18.96	384	0.5742	0.8091
0.3229	19.75	400	0.5653	0.8121

Framework versions

• Transformers 4.30.1
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3