Open-source implementations, research code, and tools for the machine learning community
This Jupyter Notebook implements a graph neural network (GNN) pipeline for binary survival classification (alive vs. deceased) using multi-omics breast cancer data.
Design novel protein sequences with unique binding sites using GPT-2.
DINOv2 Nuclei Segmentation.
A Gradio app for Microsoft's SOTA grounded radiology report generation model.
Fine tuned SmolLM-135M-Instruct using Group Relative Policy Optimization (with and without Flash Attenton). Made a Gradio Space for the model.
Fine tuned Llama 2 on a custom product review sentiment dataset with instructions.
A Gradio app for TotalSegmentator.
Fine-tuned the model.
A DICOM annotation tool for 3D segmentation.
Speech emotion classification with multi-head self-attention.
A Gradio app for SigLIP 2 Zero-shot image classification.
Space