BEiT_beit-base-finetuned-ade-640-640_Clean-Set1_RGB

This model is a fine-tuned version of microsoft/beit-base-finetuned-ade-640-640 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0603
• Mean Iou: 0.9672
• Mean Accuracy: 0.9774
• Overall Accuracy: 0.9930
• Accuracy Background: 0.9961
• Accuracy Melt: 0.9392
• Accuracy Substrate: 0.9971
• Iou Background: 0.9929
• Iou Melt: 0.9207
• Iou Substrate: 0.9879

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Melt	Accuracy Substrate	Iou Background	Iou Melt	Iou Substrate
0.3924	5.5556	50	0.3038	0.9022	0.9499	0.9809	0.9854	0.8738	0.9906	0.9853	0.7493	0.9719
0.0857	11.1111	100	0.0788	0.9656	0.9771	0.9931	0.9972	0.9377	0.9964	0.9939	0.9146	0.9883
0.0816	16.6667	150	0.0603	0.9672	0.9774	0.9930	0.9961	0.9392	0.9971	0.9929	0.9207	0.9879

Framework versions

• Transformers 4.41.2
• Pytorch 2.0.1+cu117
• Datasets 2.19.2
• Tokenizers 0.19.1