KawrakowandGitHub d5507e33ae Split mode graph for dense Gemma4 assistant (#2022) ...

* WIP: Split mode graph for Gemma4 assistant

Something is not right - acceptance drops to nearly zero.

* Per model CUDA contexts

Still not working!?

* This works

The issue was that I was not correctly calculating the number
of KV heads for the split KV cache.

* Compiler warnings

* It is better to use llama_context pointers as keys

2026-06-24 18:29:32 +02:00

..

ggml-alloc.h

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

ggml-backend.h

Fix DFlash oerformance with split mode graph (#1980)

2026-06-17 18:40:02 +02:00

ggml-cann.h

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

ggml-cpp.h

Port mdmd from mainline + Qwen2/2.5-VL support (#798)

2025-09-27 08:45:29 +02:00

ggml-cuda.h

Split mode graph for dense Gemma4 assistant (#2022)

2026-06-24 18:29:32 +02:00

ggml-kompute.h

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

ggml-metal.h

Merge mainline - Aug 12 2024 (#17)

2024-08-12 15:14:32 +02:00

ggml-rpc.h

server: improve speed of speculative decoding (#1119)

2026-01-10 08:01:22 +02:00

ggml-sycl.h

Merge mainline llama.cpp (#3)

2024-07-27 07:55:01 +02:00

ggml-vulkan.h

Vulkan: a fresh start (#608)

2025-07-15 08:03:13 +02:00

ggml.h

on-demand tensor reload (#1989)

2026-06-22 16:36:34 +02:00