segformer-b0-finetuned-segments-sidewalk-test

This model is a fine-tuned version of nvidia/mit-b0 on the sungile/custom-object-masking dataset. It achieves the following results on the evaluation set:

• Loss: 0.1913
• Mean Iou: nan
• Mean Accuracy: nan
• Overall Accuracy: nan
• Accuracy Background: nan
• Accuracy Object: nan
• Iou Background: nan
• Iou Object: nan

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: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• 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: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Background	Accuracy Object	Iou Background	Iou Object
0.5399	0.25	20	0.5706	nan	nan	nan	nan	nan	nan	nan
0.4193	0.5	40	0.4288	nan	nan	nan	nan	nan	nan	nan
0.3546	0.75	60	0.3178	nan	nan	nan	nan	nan	nan	nan
0.2735	1.0	80	0.2525	nan	nan	nan	nan	nan	nan	nan
0.2771	1.25	100	0.2201	nan	nan	nan	nan	nan	nan	nan
0.2285	1.5	120	0.2078	nan	nan	nan	nan	nan	nan	nan
0.2012	1.75	140	0.1966	nan	nan	nan	nan	nan	nan	nan
0.1894	2.0	160	0.1913	nan	nan	nan	nan	nan	nan	nan

Framework versions

• Transformers 4.47.1
• Pytorch 2.1.0+cu118
• Datasets 3.2.0
• Tokenizers 0.21.0