Pinned Loading

1. GraphZero GraphZero Public

   Graph Zero is a novel AlphaZero-style algorithm incorporating Graph Attention Networks for improved game state representation, implemented in pure CUDA, distributed training

   Cuda 1

2. cuLLM cuLLM Public

   Large Language Model and Tokenizer written from scratch in CUDA, LLaMA3.1 architecture. Tokenizer written from scratch in C++, GPT-4o tokenizer.

   Cuda 1

3. vlm-o vlm-o Public

   VLM-o is a 3B parameter multi-modal vision model built for image analysis and object detection. This model is inspired by the PaliGemma architecture. Inference implementation written in PyTorch.

   Python 1

4. BitDB BitDB Public

   SQLite Database Clone written in one C file. Added LoRA fine-tuned Mistral-7B model written in MLX for natural language to database query support.

   C 2

5. paper_implementations paper_implementations Public

   Some ML paper implementations

   Python 1

6. agent-mu agent-mu Public

   Python