parking-utcustom-train-SF-RGB-b5_2
This model is a fine-tuned version of nvidia/mit-b5 on the sam1120/parking-utcustom-train dataset. It achieves the following results on the evaluation set:
- Loss: 0.1577
- Mean Iou: 0.4996
- Mean Accuracy: 0.9992
- Overall Accuracy: 0.9992
- Accuracy Unlabeled: nan
- Accuracy Parking: nan
- Accuracy Unparking: 0.9992
- Iou Unlabeled: nan
- Iou Parking: 0.0
- Iou Unparking: 0.9992
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: 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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.767 | 20.0 | 20 | 0.6634 | 0.3299 | 0.9898 | 0.9898 | nan | nan | 0.9898 | 0.0 | 0.0 | 0.9898 |
| 0.5036 | 40.0 | 40 | 0.3752 | 0.3316 | 0.9949 | 0.9949 | nan | nan | 0.9949 | 0.0 | 0.0 | 0.9949 |
| 0.3486 | 60.0 | 60 | 0.2976 | 0.3319 | 0.9958 | 0.9958 | nan | nan | 0.9958 | 0.0 | 0.0 | 0.9958 |
| 0.2729 | 80.0 | 80 | 0.2355 | 0.3326 | 0.9978 | 0.9978 | nan | nan | 0.9978 | 0.0 | 0.0 | 0.9978 |
| 0.2246 | 100.0 | 100 | 0.1822 | 0.4983 | 0.9966 | 0.9966 | nan | nan | 0.9966 | nan | 0.0 | 0.9966 |
| 0.2131 | 120.0 | 120 | 0.1577 | 0.4996 | 0.9992 | 0.9992 | nan | nan | 0.9992 | nan | 0.0 | 0.9992 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3
- Downloads last month
- 0
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.