Search results

Bug: Potential regression from #1398

Status: Open.

#1661 In ikawrakow/ik_llama.cpp;

· sayap opened on Apr 19, 2026

Feature Request: Backend optimization for speculative decoding

enhancementNew feature or requestNew feature or request

Status: Open.

#1630 In ikawrakow/ik_llama.cpp;

· SamuelOliveirads opened on Apr 13, 2026

Feature Request: merge_up_gate_shexp

enhancementNew feature or requestNew feature or request

Status: Open.

#1629 In ikawrakow/ik_llama.cpp;

· Nexesenex opened on Apr 13, 2026

MoE models are broken with CUDA + CPU and no layers offloaded to the GPU

Status: Open.

#1618 In ikawrakow/ik_llama.cpp;

· ZenSirzechs opened on Apr 11, 2026

Bug: Segfault on GLM 5.1 with spec-decoding

Status: Open.

#1612 In ikawrakow/ik_llama.cpp;

· ValeryNo opened on Apr 10, 2026

Bug: Missing ik-llama-builder

Status: Open.

#1608 In ikawrakow/ik_llama.cpp;

· DevilaN opened on Apr 10, 2026

Feature Request: SWA KV cache compression for Gemma 4 (and other SWA models)

Status: Open.

#1607 In ikawrakow/ik_llama.cpp;

· ormandj opened on Apr 9, 2026

Bug: Kimi K2.5 template: silent HTTP 400 when assistant message has content + tool_calls in multi-turn conversations

Status: Open.

#1605 In ikawrakow/ik_llama.cpp;

· SixPathsSolutions opened on Apr 9, 2026

Not sure is it a bug. - Failed with cuda error on GTX 1080

Status: Open.

#1580 In ikawrakow/ik_llama.cpp;

· bt-worm opened on Apr 4, 2026

Bug: Hybrid models revaluate the prompt invalidating kv

Status: Open.

#1576 In ikawrakow/ik_llama.cpp;

· nicman23 opened on Apr 3, 2026

Feature Request: Gemma4 model support

enhancementNew feature or requestNew feature or request

Status: Open.

#1572 In ikawrakow/ik_llama.cpp;

· VanceVagell opened on Apr 2, 2026

Bug: Inference slows down after parallel request processing

Status: Open.

#1541 In ikawrakow/ik_llama.cpp;

· JKratto opened on Mar 28, 2026