llama.cpp
2c516611 - CUDA: mul_mat_vec_q for batch sizes > 1 (#5351)

Commit

1 year ago

CUDA: mul_mat_vec_q for batch sizes > 1 (#5351)

References

#5351 - CUDA: mul_mat_vec_q for batch sizes > 1

Author

JohannesGaessler

JohannesGaessler

Parents

Loading