Dual-Explain Stage Full — SAM2
Final LoRA adapter and saved task modules for the SAM2 ablation.
Dependencies:
- Stage-1 base:
tkhangg0910/Merged-Dual-Explain-Llava-Stage_1-7b - Segmentation backbone:
facebook/sam2-hiera-large - Segmentation type:
sam2 - Classes: GCA, Koedam, MTA
- Evaluation mask threshold: 0.5
See PEFT_README.md for the generated PEFT metadata.
- Downloads last month
- 24
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support