parking-utcustom-train-SF-RGB-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: 0.7834
• Mean Iou: 0.3040
• Mean Accuracy: 0.9120
• Overall Accuracy: 0.9120
• Accuracy Unlabeled: nan
• Accuracy Parking: nan
• Accuracy Unparking: 0.9120
• Iou Unlabeled: 0.0
• Iou Parking: 0.0
• Iou Unparking: 0.9120

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: 3e-05
• 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: 120

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
1.0965	20.0	20	1.0974	0.1433	0.4299	0.4299	nan	nan	0.4299	0.0	0.0	0.4299
0.9563	40.0	40	1.0286	0.2412	0.7235	0.7235	nan	nan	0.7235	0.0	0.0	0.7235
0.8707	60.0	60	0.9260	0.2870	0.8609	0.8609	nan	nan	0.8609	0.0	0.0	0.8609
0.7662	80.0	80	0.8392	0.2951	0.8853	0.8853	nan	nan	0.8853	0.0	0.0	0.8853
0.7385	100.0	100	0.7800	0.3058	0.9173	0.9173	nan	nan	0.9173	0.0	0.0	0.9173
0.7107	120.0	120	0.7834	0.3040	0.9120	0.9120	nan	nan	0.9120	0.0	0.0	0.9120

Framework versions

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