metadata
license: other
tags:
- vision
- image-segmentation
- generated_from_trainer
model-index:
- name: parking-utcustom-train-SF-RGBD-b5_1
results: []
parking-utcustom-train-SF-RGBD-b5_1
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: 0.0350
- Mean Iou: 0.4983
- Mean Accuracy: 0.9967
- Overall Accuracy: 0.9967
- Accuracy Unlabeled: nan
- Accuracy Parking: nan
- Accuracy Unparking: 0.9967
- Iou Unlabeled: nan
- Iou Parking: 0.0
- Iou Unparking: 0.9967
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: 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: 140
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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.4573 | 20.0 | 20 | 0.3024 | 0.3276 | 0.9829 | 0.9829 | nan | nan | 0.9829 | 0.0 | 0.0 | 0.9829 |
| 0.2183 | 40.0 | 40 | 0.2365 | 0.3318 | 0.9953 | 0.9953 | nan | nan | 0.9953 | 0.0 | 0.0 | 0.9953 |
| 0.1266 | 60.0 | 60 | 0.0999 | 1.0 | 1.0 | 1.0 | nan | nan | 1.0 | nan | nan | 1.0 |
| 0.0929 | 80.0 | 80 | 0.0590 | 0.4986 | 0.9972 | 0.9972 | nan | nan | 0.9972 | nan | 0.0 | 0.9972 |
| 0.0669 | 100.0 | 100 | 0.0375 | 0.4998 | 0.9996 | 0.9996 | nan | nan | 0.9996 | nan | 0.0 | 0.9996 |
| 0.0557 | 120.0 | 120 | 0.0443 | 0.4953 | 0.9907 | 0.9907 | nan | nan | 0.9907 | nan | 0.0 | 0.9907 |
| 0.0688 | 140.0 | 140 | 0.0350 | 0.4983 | 0.9967 | 0.9967 | nan | nan | 0.9967 | nan | 0.0 | 0.9967 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu117
- Datasets 2.13.1
- Tokenizers 0.13.3