| # 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. |