Pinned Loading

1. retriforever retriforever Public

   KISS! A local retrieval layer for AI code assistants. No cloud, no embedding magic, no subscription for what ctags used to do.

   Python

2. mlops mlops Public

   Collection of CUDA C++ custom and standart operations for training and inference on NVIDIA GPUs, from Maxwell to Hopper (sm_50 .. sm_90; support varies by operation). Use them directly from C / C++…

   Cuda 1

3. nvfp4effma nvfp4effma Public

   NVFP4 GEMM, GEMV, linear layers and quantization for NVIDIA GPUs without native FP4 (sm_50–sm_90): every NVIDIA NVFP4 variant (ModelOpt, Transformer Engine 1x16 / 16x16, 4over6, swizzled scales) bi…

   Cuda 1

4. e5m2effma e5m2effma Public

   FP8 (E5M2 and E4M3, any pairing) GEMM, GEMV and fused AdamW for NVIDIA GPUs without native FP8 (sm_50–sm_87, Maxwell to Ampere): exact products, FP32 accumulation, TE HYBRID backward, C API and PyT…

   Cuda 1

5. e4m3effma e4m3effma Public

   FP8 (E4M3) GEMM, GEMV and fused AdamW for NVIDIA GPUs without native FP8 (sm_50–sm_87, Maxwell to Ampere): exact products, FP32 accumulation, C API and PyTorch bindings incl. torch._scaled_mm and F…

   Cuda 1 1

6. x.AGENTS.md x.AGENTS.md Public

   A comprehensive coding style and operational protocol for AI agents — mandatory rules, best practices, and language-specific guidance to produce maintainable, secure, and reliable code.

   2