tilelang

tilelang copied to clipboard

[FFI] Rebase tvm to v0.22.0 to utilize tvm-ffi

17

This pull request introduces significant refactoring and modernization of the FFI (Foreign Function Interface) and object system usage in the codebase, particularly in the layout and IR (Intermediate Representation) modules....

LeiWang1999

[Question] `func_or_mod.__name__` in MetalKernelAdapter

2

### Required prerequisites - [x] I have read the documentation . - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.)...

kurisu6912

### Required prerequisites - [x] I have read the documentation . - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.)...

PannenetsF

[BUG] A simple T.copy fails

1

### Required prerequisites - [x] I have read the documentation . - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.)...

zhengqigao

[EXAMPLE] In the flash attention example keep the max of all blocks seen in scores_max numerical stability

9

In the flash attention example, keep the max of previous scores_max and max(acc_s) in scores_max for numerical stability From [Flash Attention 2 paper](https://arxiv.org/pdf/2205.14135), Algorithm 1 $$m_i^{\text{new}} = \max(m_i, \tilde{m}_{ij})$$ ##...

vpj

[Question] Why is the version of tilelang superior to the Cutlass version of deepseek?

4

### Required prerequisites - [x] I have read the documentation . - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.)...

zzyplaybasketball

[Enhancement] Enhance async proxy analyzing logic and GEMM dispatching

3

- Added a mapping for GEMM instruction prefixes in `gemm.h`. - Renamed GEMM functions to include `mma_` prefix for clarity in `gemm_mma.h`, `gemm_sm70.h`, `gemm_sm90.h`, and `gemm_sp_sm80.h`. - Updated function signatures...

Rachmanino

[Feature Request][CI] Enable `ccache` or `sccache` for `cibuildwheel` for `manylinux`

1

### Required prerequisites - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.) ### Motivation CIBW runs natively on macOS, allowing...

XuehaiPan

[Language] support `T.gemm_sp_v2` on sm80 and sm89

2

Roadmap: - [ ] Clear TODOs # SM8x - [x] bf16/fp16 - [x] customized metadata layout - [x] tf32 - [ ] precision issue due to using fp32 as tf32...

botbw

[Feature Request] Integrate Z3 Prover

1

### Required prerequisites - [x] I have searched the [Issue Tracker](https://github.com/tile-ai/tilelang/issues) that this hasn't already been reported. (comment there if it has.) ### Motivation From some issues #1012 we can...

LeiWang1999