jcwang0602

/

MLLMSeg_InternVL2_5_1B_RES

like 0

This PR significantly enhances the model card for the MLLMSeg_InternVL2_5_1B_RES model by:

• Adding the pipeline_tag: image-segmentation, which makes the model discoverable on the Hugging Face Hub under this pipeline (https://huggingface.co/models?pipeline_tag=image-segmentation).
• Specifying library_name: transformers, enabling the "How to use" widget on the model page for easier integration.
• Linking directly to the paper: Unlocking the Potential of MLLMs in Referring Expression Segmentation via a Light-weight Mask Decoder.
• Including a direct link to the official GitHub repository for access to the code.
• Adding a comprehensive "How to Use" section with a runnable Python code snippet for inference, along with important notes regarding input/output processing.
• Incorporating performance metrics and visualization examples from the original GitHub repository to provide a clearer understanding of the model's capabilities.

This update aims to greatly improve the model's visibility and usability for the community.