README.md

metadata

license: other
base_model: nvidia/mit-b0
tags:
  - vision
  - image-segmentation
  - generated_from_trainer
model-index:
  - name: segformer-b0-finetuned-segments-SixrayKnife8-19-2024
    results: []

segformer-b0-finetuned-segments-SixrayKnife8-19-2024

This model is a fine-tuned version of nvidia/mit-b0 on the saad7489/SixraygunTest dataset. It achieves the following results on the evaluation set:

• Loss: 0.1817
• Mean Iou: 0.8160
• Mean Accuracy: 0.8823
• Overall Accuracy: 0.9881
• Accuracy Bkg: 0.9954
• Accuracy Gun: 0.7759
• Accuracy Knife: 0.8755
• Iou Bkg: 0.9890
• Iou Gun: 0.7014
• Iou Knife: 0.7574

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: 20
• eval_batch_size: 20
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Bkg	Accuracy Gun	Accuracy Knife	Iou Bkg	Iou Gun	Iou Knife
0.4406	5.0	20	0.4093	0.7210	0.7883	0.9804	0.9938	0.6719	0.6991	0.9807	0.5730	0.6092
0.3699	10.0	40	0.3327	0.7327	0.7880	0.9819	0.9954	0.6559	0.7128	0.9824	0.5724	0.6432
0.31	15.0	60	0.3035	0.7698	0.8614	0.9842	0.9926	0.7207	0.8709	0.9853	0.6217	0.7023
0.2852	20.0	80	0.2649	0.7817	0.8711	0.9850	0.9928	0.7453	0.8752	0.9860	0.6423	0.7168
0.2583	25.0	100	0.2329	0.7936	0.8693	0.9863	0.9943	0.7497	0.8639	0.9873	0.6628	0.7307
0.2521	30.0	120	0.2194	0.7975	0.8778	0.9867	0.9942	0.7530	0.8862	0.9879	0.6731	0.7316
0.2357	35.0	140	0.2044	0.8042	0.8804	0.9871	0.9944	0.7635	0.8833	0.9881	0.6789	0.7456
0.2198	40.0	160	0.1929	0.8126	0.8789	0.9878	0.9953	0.7685	0.8728	0.9888	0.6937	0.7552
0.1909	45.0	180	0.1837	0.8151	0.8810	0.9880	0.9954	0.7726	0.8750	0.9890	0.6997	0.7568
0.1908	50.0	200	0.1817	0.8160	0.8823	0.9881	0.9954	0.7759	0.8755	0.9890	0.7014	0.7574

Framework versions

• Transformers 4.42.4
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1