Pinned Loading

1. flash-attention-cuda flash-attention-cuda Public

   Custom FlashAttention CUDA kernel for Llama 3 8B inference. 2.26x speedup over naive attention, targeting RTX 4060 Ti → Jetson AGX Orin deployment.

   Python 1

2. flashattn-cuda-metal flashattn-cuda-metal Public

   FlashAttention CUDA kernel implementation and Metal port (RTX 4060 Ti, Apple M4 Pro)

   Python 1

3. sdpa-attention-benchmark sdpa-attention-benchmark Public

   Benchmark PyTorch SDPA backends (math vs flash) on RTX 4060 Ti with Nsight Systems profiling

   Python 1