| library_name: kernels | |
| license: apache-2.0 | |
| This is the repository card of kernels-community/finegrained-fp8 that has been pushed on the Hub. It was built to be used with the [`kernels` library](https://github.com/huggingface/kernels). This card was automatically generated. | |
| ## How to use | |
| ```python | |
| # make sure `kernels` is installed: `pip install -U kernels` | |
| from kernels import get_kernel | |
| kernel_module = get_kernel("kernels-community/finegrained-fp8") | |
| fp8_act_quant = kernel_module.fp8_act_quant | |
| fp8_act_quant(...) | |
| ``` | |
| ## Available functions | |
| - `fp8_act_quant` | |
| - `w8a8_fp8_matmul` | |
| - `w8a8_block_fp8_matmul` | |
| - `w8a8_tensor_fp8_matmul` | |
| - `w8a8_fp8_matmul_batched` | |
| - `w8a8_block_fp8_matmul_batched` | |
| - `w8a8_tensor_fp8_matmul_batched` | |
| - `w8a8_fp8_matmul_grouped` | |
| - `w8a8_block_fp8_matmul_grouped` | |
| - `w8a8_tensor_fp8_matmul_grouped` | |
| ## Benchmarks | |
| No benchmark available yet. | |