Spaces:

Duplicated from natasa365/whisper.cpp

Xenobd
/

whisper.cpp

Running

App Files Files Community

whisper.cpp / ggml-cuda /fattn-common.cuh

Commit History

CUDA: deduplicate FlashAttention code (llama/7352)

65ab3e8

JohannesGaessler commited on May 18, 2024

CUDA: add FP32 FlashAttention vector kernel (llama/7188)

03d4b22
unverified

JohannesGaessler commited on May 12, 2024