Files
llama.cpp/ggml
Aman Gupta a90eb94ca9 CUDA: fuse rope + set_rows (#16884)
* CUDA: add fused rope

* move k forward_expand up

* create helper function instead of re-using params

* make assert statement more in line with comment

* rope_norm: coalesced writes to global mem
2025-11-13 08:50:01 +08:00
..
2025-11-13 08:50:01 +08:00
2024-07-13 18:12:39 +02:00