开放Eval bug: Gemma4 fails with NotImplemented: map: filter-mapping not implementedenhancementgood first issueggml-org/llama.cpp #21,547 创建于 2个月前 · C++ · 110,169 stars9 条评论0 个反应0 负责人
开放Feature Request: Graceful reasoning budget termination. Avoid mid-sentence cutoff.enhancementgood first issuehelp wantedggml-org/llama.cpp #20,632 创建于 2个月前 · C++ · 110,169 stars6 条评论0 个反应1 负责人
开放Eval bug: Model loading fails on Windows when path contains non-ASCII (Cyrillic) charactersbughelp wantedggml-org/llama.cpp #18,571 创建于 5个月前 · C++ · 110,169 stars5 条评论0 个反应0 负责人
开放Misc. bug: llama-finetune won't work even with 17M parameters arch:llamaReview Complexity : Highbughelp wantedggml-org/llama.cpp #18,499 创建于 5个月前 · C++ · 110,169 stars11 条评论0 个反应0 负责人
开放Feature Request: support the multi-modal Bagel modelReview Complexity : Highenhancementhelp wantedmodelmtmdggml-org/llama.cpp #17,861 创建于 6个月前 · C++ · 110,169 stars1 条评论1 个反应0 负责人
开放Feature Request: support Qwen3-omni, Qwen3-TTS and MiniCPM-oReview Complexity : Highenhancementgood first issuehelp wantedmtmdttsggml-org/llama.cpp #17,634 创建于 6个月前 · C++ · 110,169 stars13 条评论27 个反应0 负责人
开放Eval bug: wrong default threads in llama-benchgood first issuelow severityperformanceggml-org/llama.cpp #17,611 创建于 6个月前 · C++ · 110,169 stars6 条评论0 个反应0 负责人
开放GGUF convert support for Vibevoice3rd partyenhancementgood first issuehelp wantedmodelttsggml-org/llama.cpp #17,488 创建于 6个月前 · C++ · 110,169 stars3 条评论1 个反应0 负责人
开放Eval bug: llama server response hangs for /slots/0?action=erasebughelp wantedmedium severityserver/apiggml-org/llama.cpp #17,387 创建于 6个月前 · C++ · 110,169 stars5 条评论0 个反应0 负责人
开放Feature Request: Implement missing ops from backendsenhancementgood first issueggml-org/llama.cpp #14,909 创建于 10个月前 · C++ · 110,169 stars33 条评论6 个反应0 负责人
开放tutorials : list for llama.cppgood first issuehelp wantedroadmapggml-org/llama.cpp #13,523 创建于 去年 · C++ · 110,169 stars20 条评论20 个反应0 负责人
开放tts : add support for Orpheusgood first issuettsggml-org/llama.cpp #12,476 创建于 去年 · C++ · 110,169 stars7 条评论26 个反应0 负责人
开放Feature Request: Mapping model name to LoRA configenhancementgood first issueserverggml-org/llama.cpp #11,031 创建于 去年 · C++ · 110,169 stars6 条评论2 个反应0 负责人
开放examples : add configuration presetsdocumentationenhancementexamplesgood first issuehelp wantedggml-org/llama.cpp #10,932 创建于 去年 · C++ · 110,169 stars9 条评论3 个反应0 负责人
开放Feature Request: Allow Filtering LLama Server Response Fieldsenhancementgood first issueggml-org/llama.cpp #10,819 创建于 去年 · C++ · 110,169 stars10 条评论0 个反应0 负责人
开放Compile bug: ios swift xcode build error when upgrade to llama : use cmake for swift build buildgood first issuehelp wantedggml-org/llama.cpp #10,747 创建于 去年 · C++ · 110,169 stars41 条评论3 个反应0 负责人
开放Misc. bug: server provides strutured output for response_format: json_object, but not for response_format: json_schemaenhancementgood first issueggml-org/llama.cpp #10,732 创建于 去年 · C++ · 110,169 stars5 条评论3 个反应0 负责人
开放llama : add CLI assistantenhancementgood first issueroadmapggml-org/llama.cpp #10,688 创建于 去年 · C++ · 110,169 stars4 条评论3 个反应0 负责人
开放Feature Request: llama-server hot swapping cvectors via API like we can do with LoRA adapters nowenhancementgood first issueggml-org/llama.cpp #10,685 创建于 去年 · C++ · 110,169 stars2 条评论4 个反应0 负责人
开放ggml : add ANE backendhelp wantedresearch 🔬roadmapggml-org/llama.cpp #10,453 创建于 去年 · C++ · 110,169 stars21 条评论84 个反应0 负责人
开放Eval bug: Gemma4 fails with NotImplemented: map: filter-mapping not implementedenhancementgood first issueggml-org/llama.cpp #21,547 创建于 2个月前 · C++ · 110,169 stars9 条评论0 个反应0 负责人
开放Feature Request: Graceful reasoning budget termination. Avoid mid-sentence cutoff.enhancementgood first issuehelp wantedggml-org/llama.cpp #20,632 创建于 2个月前 · C++ · 110,169 stars6 条评论0 个反应1 负责人
开放Eval bug: Model loading fails on Windows when path contains non-ASCII (Cyrillic) charactersbughelp wantedggml-org/llama.cpp #18,571 创建于 5个月前 · C++ · 110,169 stars5 条评论0 个反应0 负责人
开放Misc. bug: llama-finetune won't work even with 17M parameters arch:llamaReview Complexity : Highbughelp wantedggml-org/llama.cpp #18,499 创建于 5个月前 · C++ · 110,169 stars11 条评论0 个反应0 负责人
开放Feature Request: support the multi-modal Bagel modelReview Complexity : Highenhancementhelp wantedmodelmtmdggml-org/llama.cpp #17,861 创建于 6个月前 · C++ · 110,169 stars1 条评论1 个反应0 负责人
开放Feature Request: support Qwen3-omni, Qwen3-TTS and MiniCPM-oReview Complexity : Highenhancementgood first issuehelp wantedmtmdttsggml-org/llama.cpp #17,634 创建于 6个月前 · C++ · 110,169 stars13 条评论27 个反应0 负责人
开放Eval bug: wrong default threads in llama-benchgood first issuelow severityperformanceggml-org/llama.cpp #17,611 创建于 6个月前 · C++ · 110,169 stars6 条评论0 个反应0 负责人
开放GGUF convert support for Vibevoice3rd partyenhancementgood first issuehelp wantedmodelttsggml-org/llama.cpp #17,488 创建于 6个月前 · C++ · 110,169 stars3 条评论1 个反应0 负责人
开放Eval bug: llama server response hangs for /slots/0?action=erasebughelp wantedmedium severityserver/apiggml-org/llama.cpp #17,387 创建于 6个月前 · C++ · 110,169 stars5 条评论0 个反应0 负责人
开放Feature Request: Implement missing ops from backendsenhancementgood first issueggml-org/llama.cpp #14,909 创建于 10个月前 · C++ · 110,169 stars33 条评论6 个反应0 负责人
开放tutorials : list for llama.cppgood first issuehelp wantedroadmapggml-org/llama.cpp #13,523 创建于 去年 · C++ · 110,169 stars20 条评论20 个反应0 负责人
开放tts : add support for Orpheusgood first issuettsggml-org/llama.cpp #12,476 创建于 去年 · C++ · 110,169 stars7 条评论26 个反应0 负责人
开放Feature Request: Mapping model name to LoRA configenhancementgood first issueserverggml-org/llama.cpp #11,031 创建于 去年 · C++ · 110,169 stars6 条评论2 个反应0 负责人
开放examples : add configuration presetsdocumentationenhancementexamplesgood first issuehelp wantedggml-org/llama.cpp #10,932 创建于 去年 · C++ · 110,169 stars9 条评论3 个反应0 负责人
开放Feature Request: Allow Filtering LLama Server Response Fieldsenhancementgood first issueggml-org/llama.cpp #10,819 创建于 去年 · C++ · 110,169 stars10 条评论0 个反应0 负责人
开放Compile bug: ios swift xcode build error when upgrade to llama : use cmake for swift build buildgood first issuehelp wantedggml-org/llama.cpp #10,747 创建于 去年 · C++ · 110,169 stars41 条评论3 个反应0 负责人
开放Misc. bug: server provides strutured output for response_format: json_object, but not for response_format: json_schemaenhancementgood first issueggml-org/llama.cpp #10,732 创建于 去年 · C++ · 110,169 stars5 条评论3 个反应0 负责人
开放llama : add CLI assistantenhancementgood first issueroadmapggml-org/llama.cpp #10,688 创建于 去年 · C++ · 110,169 stars4 条评论3 个反应0 负责人
开放Feature Request: llama-server hot swapping cvectors via API like we can do with LoRA adapters nowenhancementgood first issueggml-org/llama.cpp #10,685 创建于 去年 · C++ · 110,169 stars2 条评论4 个反应0 负责人
开放ggml : add ANE backendhelp wantedresearch 🔬roadmapggml-org/llama.cpp #10,453 创建于 去年 · C++ · 110,169 stars21 条评论84 个反应0 负责人