TEAL

TEAL copied to clipboard

milktea888

Hi author, Thanks for publishing such high-quality training-free sparisty research work. I saw that you have plans to integrate sparse computing and speculative decoding, and can you share me when...

CarlHuangNuc

Hello, I really enjoy this paper and the code. I want to reproduce the part below(speedup)  So I tested with 1 A100(with following explanation in README), and obtained the...

quaternior

it's strange for the MMLU result

1

Hello, I test the MMLU, using different sparsity from 0 to 90%。But the mmlu score forcus on 50%, which didn't show significant decline as the sparsity becoming large. Since MMLU...

milktea888

Fix the typo in Readme

shaginhekvs

Incomplete implementation of SparseGEMV

3

https://github.com/FasterDecoding/TEAL/blob/fb7373c93ac3594817c9ee64d4e08b47430a1822/kernels/sparse_gemv.py#L271 Hi, I notice that the SparseGEMV kernel only manage the case when `batch_size=1 & seqlen=1`. Beyond that case, the kernel outputs wrong answer. Is it expected that this kernel...

kyang-06