New
google/xnnpack — SIMD Inference Kernels for ARM·x86·WebAssembly·RISC-V
google/xnnpack is a low-level C library that optimizes floating-point neural network inference operations on ARM, x86, WebAssembly, and RISC-V, accelerating higher-level frameworks such as TensorFlow Lite and PyTorch.