Instructions to use replicate/flash-mla with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Kernels
How to use replicate/flash-mla with Kernels:
# !pip install kernels from kernels import get_kernel # a version (or an explicit revision) is required; see the "Files and versions" tab for the available ones kernel = get_kernel("replicate/flash-mla", version=1) - Notebooks
- Google Colab
- Kaggle
Download flash_mla/named_barrier.h from replicate/flash-mla: direct link, hf CLI and curl.
- Browser
- Download file 316 Bytes
-
https://huggingface.co/replicate/flash-mla/resolve/main/flash_mla/named_barrier.h
- Command line
-
hf download hf://replicate/flash-mla/flash_mla/named_barrier.h
-
curl -L -o named_barrier.h https://huggingface.co/replicate/flash-mla/resolve/main/flash_mla/named_barrier.h
316 Bytes
| namespace flash { | |
| //////////////////////////////////////////////////////////////////////////////////////////////////// | |
| // Enumerates the reserved named barriers to avoid potential conflicts | |
| enum class NamedBarriers { | |
| SReady = 1, | |
| SoftmaxReady = 2, | |
| }; | |
| } // flash | |