1 Release published by 1 person

• v1.77 koboldcpp-1.77

  published Nov 1, 2024

1 Pull request opened by 1 person

• debug: display tokens that were dropped by XTC sampler when debugmode is enabled

  #1201 opened Nov 5, 2024

5 Issues closed by 5 people

• Multiuser issue

  #1193 closed Nov 4, 2024

• 1.77 initialization too slow

  #1196 closed Nov 3, 2024

• Whisper errors on KoboldAI Lite frontend (SillyTavern with same KoboldCpp Whisper works well)

  #1189 closed Nov 1, 2024

• Support ComfyUI as image generation backend

  #1017 closed Nov 1, 2024

• Loop

  #1194 closed Nov 1, 2024

5 Issues opened by 5 people

• Oute TTS integration

  #1202 opened Nov 5, 2024

• Windows 10 AMD ROCm not loading model when it did a week ago

  #1200 opened Nov 4, 2024

• Runpod template is broken

  #1199 opened Nov 4, 2024

• [Question] Any plans to support models other than GGUF and modalities other than QnA ?

  #1197 opened Nov 2, 2024

• Logprobs view – show even when aborted

  #1195 opened Nov 1, 2024

7 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• [Enhancement] Adding support for YaRN scaling

  #868 commented on Nov 1, 2024 • 0 new comments

• trycloudflare CURLEWRITEERROR

  #1155 commented on Nov 1, 2024 • 0 new comments

• Odd repetition and other issues recently - Incorrect ftype?

  #1192 commented on Nov 2, 2024 • 0 new comments

• KoboldCPP crashes after Arch system update when loading GGUF model: ggml_cuda_host_malloc ... invalid argument

  #1158 commented on Nov 4, 2024 • 0 new comments

• Multi-task concurrency function is required, which is very important for API users. Similar to ollama, this function has been implemented

  #866 commented on Nov 5, 2024 • 0 new comments

• Add a way to use Control Vectors

  #1002 commented on Nov 5, 2024 • 0 new comments

• support compiling with intel MKL for with LLAMA_MKL=1

  #1063 commented on Nov 5, 2024 • 0 new comments