Pinned Loading

1. flash-attention-minimal flash-attention-minimal Public

   Flash Attention in ~100 lines of CUDA (forward pass only)

   Cuda 1.2k 114

2. cuda-1brc cuda-1brc Public

   My CUDA solution to the 1BRC

   Cuda 11 3

3. mixed-precision-from-scratch mixed-precision-from-scratch Public

   Mixed precision training from scratch with Tensors and CUDA

   Python 30 4

4. paged-attention-minimal paged-attention-minimal Public

   a minimal cache manager for PagedAttention, on top of llama3.

   Python 146 12

5. insta-chat insta-chat Public

   DIY Instagram Chat Automation with Google Sheets

   HTML 240 29

6. HazyResearch/ThunderKittens HazyResearch/ThunderKittens Public

   Tile primitives for speedy kernels

   Cuda 3.5k 300