forked from flashinfer-ai/flashinfer
-
Notifications
You must be signed in to change notification settings - Fork 9
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#217 In ROCm/flashinfer;
- Status: Open.#134 In ROCm/flashinfer;
- Status: Open.#101 In ROCm/flashinfer;
Make it possible to support different HEAD_DIM in prefill kernels outside of those specified in dispatch.cuh::DISPATCH_HEAD_DIM
enhancementNew feature or requestNew feature or requestStatus: Open.#85 In ROCm/flashinfer;Batch decode FP8 tests fail on mi300a
bugSomething isn't workingSomething isn't workingStatus: Open.Enable FP8 WFMA support for CDNA3
enhancementNew feature or requestNew feature or requestStatus: Open.Support manylinux wheel builds for amd-flashinfer
infrastructureInfrastructure upgradeInfrastructure upgradeStatus: Open.Create a feature matrix listing the supported features from flashinfer-ai/flashinfer that have been enabled for rocm-flashinfer
documentationImprovements or additions to documentationImprovements or additions to documentationStatus: Open.Fix the FLASHINFER_ENABLE_FP8 flag in pyproject.toml and jit.cpp_ext
enhancementNew feature or requestNew feature or requestStatus: Open.Redesign of attention/generic
questionFurther information is requestedFurther information is requestedrefactoringRequires refactoring of codebase.Requires refactoring of codebase.Status: Open.Does rocm/flashinfer have something similar to flashinfer.green_ctx?
enhancementNew feature or requestNew feature or requestquestionFurther information is requestedFurther information is requestedStatus: Open.#29 In ROCm/flashinfer;