# Quantized Ops This folder contains kernels for quantization related ops, similar to `//executorch/kernels/portable/`. ## How to add quantized ops 1. Add a new operator definition in `quantized.yaml` 2. Implement the kernel for this operator. 3. Add unit test in `/test` directory. 4. Start to use `//executorch/kernels/quantized:generated_lib` in ExecuTorch or create your own generated lib if you only need a subset of the ops.