parking-utcustom-train-SF-RGBD-b0_2

This model is a fine-tuned version of nvidia/mit-b0 on the sam1120/parking-utcustom-train dataset. It achieves the following results on the evaluation set:

• Loss: 0.1376
• Mean Iou: 0.4738
• Mean Accuracy: 0.9476
• Overall Accuracy: 0.9476
• Accuracy Unlabeled: nan
• Accuracy Parking: nan
• Accuracy Unparking: 0.9476
• Iou Unlabeled: nan
• Iou Parking: 0.0
• Iou Unparking: 0.9476

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 150

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Parking	Accuracy Unparking	Iou Unlabeled	Iou Parking	Iou Unparking
0.5832	20.0	20	0.5894	1.0	1.0	1.0	nan	nan	1.0	nan	nan	1.0
0.3162	40.0	40	0.2686	1.0	1.0	1.0	nan	nan	1.0	nan	nan	1.0
0.2152	60.0	60	0.1349	1.0	1.0	1.0	nan	nan	1.0	nan	nan	1.0
0.1517	80.0	80	0.0822	1.0	1.0	1.0	nan	nan	1.0	nan	nan	1.0
0.1293	100.0	100	0.0609	1.0	1.0	1.0	nan	nan	1.0	nan	nan	1.0
0.0935	120.0	120	0.0711	0.4978	0.9956	0.9956	nan	nan	0.9956	nan	0.0	0.9956
0.0835	140.0	140	0.1376	0.4738	0.9476	0.9476	nan	nan	0.9476	nan	0.0	0.9476

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3