Pinned Loading

1. Parallel-BPE-Tokenizer Parallel-BPE-Tokenizer Public

   High-performance GPT-2-style BPE tokenizer in C++ with parallel batch encoding, thread-pool execution, thread-local caching, and benchmark-driven comparison against tiktoken and GPT2TokenizerFast

   C++ 2

2. CUDAkernels CUDAkernels Public

   CUDA kernels implemented from scratch for GPU programming, reductions, shared memory, warp-level primitives, and ML/attention kernels.

   Python

3. cutlass cutlass Public

   Forked from NVIDIA/cutlass

   CUDA Templates and Python DSLs for High-Performance Linear Algebra

   C++

4. executorch executorch Public

   Forked from pytorch/executorch

   On-device AI across mobile, embedded and edge for PyTorch

   Python