Search results

Bug: option -np is broken

Status: Open.

#1149 In ikawrakow/ik_llama.cpp;

· eiffel31 opened on Jan 14, 2026

Bug: -khad breaks model output

Status: Open.

#1148 In ikawrakow/ik_llama.cpp;

· milpster opened on Jan 14, 2026

Bug: using kv cache quant q4 or q5 breaks model output

Status: Open.

#1142 In ikawrakow/ik_llama.cpp;

· milpster opened on Jan 13, 2026

Repo history changed?

Status: Open.

#1133 In ikawrakow/ik_llama.cpp;

· firecoperana opened on Jan 11, 2026

Bug: Logit Bias failing

Status: Open.

#1127 In ikawrakow/ik_llama.cpp;

· sca255 opened on Jan 9, 2026

Feature Request: more granularized tensor repartition in graph split mode to acccomodate asymetrical GPU setups

enhancementNew feature or requestNew feature or request

Status: Open.

#1125 In ikawrakow/ik_llama.cpp;

· Nexesenex opened on Jan 9, 2026

Bug: Massive slowdown after updating from #1092

Status: Open.

#1117 In ikawrakow/ik_llama.cpp;

· Quairon-Nailo opened on Jan 7, 2026

Feature Request: AMD RCCL multi-gpu (pronounced Rickle lol)

enhancementNew feature or requestNew feature or request

Status: Open.

#1113 In ikawrakow/ik_llama.cpp;

· bennmann opened on Jan 6, 2026

Bug: Does token banning work?

Status: Open.

#1109 In ikawrakow/ik_llama.cpp;

· Ph0rk0z opened on Jan 5, 2026

Feature Request: support GLM4V vision encoder

enhancementNew feature or requestNew feature or request

Status: Open.

#1084 In ikawrakow/ik_llama.cpp;

· Quairon-Nailo opened on Dec 24, 2025

Feature Request: support for EAGLE3

enhancementNew feature or requestNew feature or request

Status: Open.

#1081 In ikawrakow/ik_llama.cpp;

· jagusztinl opened on Dec 22, 2025

Bug: ik_llama server enters infinite <code>NEXT_RESPONSE</code> loop on embedding requests (Qwen3-Embedding-4B)

Status: Open.

#1077 In ikawrakow/ik_llama.cpp;

· KeinNiemand opened on Dec 20, 2025