llama.cpp
5fa07c2f - CUDA: optimize FA for GQA + large batches (#12014)

Commit

1 year ago

CUDA: optimize FA for GQA + large batches (#12014)

References

#12014 - CUDA: optimize FA for GQA + large batches

Author

JohannesGaessler

JohannesGaessler

Parents

Loading