Instructions to use unaibai/spat-instance with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use unaibai/spat-instance with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-segmentation", model="unaibai/spat-instance")# pip install -U transformers accelerate # Load model directly from transformers import AutoImageProcessor, Mask2FormerForUniversalSegmentation processor = AutoImageProcessor.from_pretrained("unaibai/spat-instance") model = Mask2FormerForUniversalSegmentation.from_pretrained("unaibai/spat-instance", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download deploy_config.json from unaibai/spat-instance: direct link, hf CLI and curl.
- Browser
- Download file 472 Bytes
-
https://huggingface.co/unaibai/spat-instance/resolve/main/deploy_config.json
- Command line
-
hf download hf://unaibai/spat-instance/deploy_config.json
-
curl -L -o deploy_config.json https://huggingface.co/unaibai/spat-instance/resolve/main/deploy_config.json
472 Bytes
| { | |
| "title": "BOOST", | |
| "description": "", | |
| "classes": [ | |
| "booster", | |
| "spat" | |
| ], | |
| "scale": 1.0, | |
| "tiling": { | |
| "tile": 1024, | |
| "overlap": 256, | |
| "edge_margin": 2 | |
| }, | |
| "inference": { | |
| "topk_per_tile": 150, | |
| "mask_threshold": 0.5, | |
| "nms_iou": 0.5, | |
| "nms_iomin": 0.8, | |
| "score_threshold_eval": 0.05 | |
| }, | |
| "score_threshold": 0.95, | |
| "tta_chosen_on_val": [ | |
| "hflip" | |
| ], | |
| "base_checkpoint": "facebook/mask2former-swin-large-coco-instance" | |
| } |