Files
ik_llama.cpp/ggml
Iwan Kawrakow 79cd04b2a6 iq2_k: better CUDA dot product
Almost on par with iq2_xs (168 t/s vs 172 t/s).
2024-07-30 12:49:55 +03:00
..
2024-07-27 07:55:01 +02:00
2024-07-29 12:38:46 +03:00
2024-07-30 12:49:55 +03:00
2024-07-27 07:55:01 +02:00
2024-07-27 07:55:01 +02:00