Pinned Loading

1. 100-Days-Of-CUDA 100-Days-Of-CUDA Public

   Jupyter Notebook 2

2. Flash_Attention_v2_GPU_Optimized Flash_Attention_v2_GPU_Optimized Public

   Custom implementation of FlashAttention v2 from scratch using CUDA and Triton. Optimized for high-performance memory-efficient attention in Transformer models.

   Jupyter Notebook

3. PipeEdge PipeEdge Public

4. LLM_Based_CKD_Prediction LLM_Based_CKD_Prediction Public

   Finetuning Llama-3.2 3B model on MIMIC-IV dataset for next visit diagnosis of chronic kidney disease

   Python

5. multiminddev multiminddev Public

   Multi-turn rl based agentic code generation tool using verl and langgraph

   Python

6. Gemma3-270M-From_Scratch Gemma3-270M-From_Scratch Public

   Custom implementation and pre-training of Gemma3-270M model using sliding window attention

   Jupyter Notebook