Issues

• [BUG]: Android inference is 15-35x slower than a Release build of the same llama.cpp commit on the same phone

  SciSharp/LLamaSharp#1456 · Filter by author chrisgull opened on Oct 12, 2026

• [BUG]: The model muse-glimmer is not loading

  SciSharp/LLamaSharp#1432 · Filter by author LyohaLyoha opened on Aug 27, 2026
  2

• Getting a Illegal instruction (core dumped) when trying to run

  SciSharp/LLamaSharp#1422 · Filter by author blackdow1 opened on Aug 21, 2026
  5

  

• [BUG]: The model Qwen3.8-27B is not loading

  SciSharp/LLamaSharp#1416 · Filter by author aropb opened on Aug 14, 2026
  13

• [Feature]: Speculative Decoding

  do not closefeature requestUpstream
  SciSharp/LLamaSharp#1388 · Filter by author aropb opened on May 18, 2026
  3

• TODO: implementing the manual history tracking fix in StatefulExecutorBase

  do not closeenhancement
  SciSharp/LLamaSharp#1376 · Filter by author zsogitbe opened on Apr 26, 2026
  1

• [Feature]: Cuda 13 support

  do not closefeature requestgood first issue
  SciSharp/LLamaSharp#1360 · Filter by author aropb opened on Mar 21, 2026
  7

  

• [BUG]: 0.24.0 CUDA detection bug

  SciSharp/LLamaSharp#1221 · Filter by author BifrostTitan opened on Jul 17, 2025
  10

• [BUG]: Not loading cuda backend on laptop

  do not closegood first issuehelp wantedstale
  SciSharp/LLamaSharp#990 · Filter by author wased89 opened on Nov 19, 2024
  10

• [BUG]:RUN LLama.Examples =>KernelMemory.cs System.AccessViolationException:“Attempted to read or write protected memory. This is often an indication that other memory is corrupt.”

  do not closegood first issuehelp wantedstale
  SciSharp/LLamaSharp#980 · Filter by author freefer opened on Nov 13, 2024
  5

• [Feature]: NativeLibraryConfig.WithLibrary availablility in Dotnet Standard 2.1 (For compatibility with Unity3D)

  do not closeenhancementfeature requeststale
  SciSharp/LLamaSharp#960 · Filter by author ltmx opened on Oct 22, 2024
  6

• Namespace should be consistent

  do not closestale
  SciSharp/LLamaSharp#693 · Filter by author SanftMonster opened on Apr 24, 2024
  3

  

• [Proposal] Refactor the mid-level and high-level implementations of LLamaSharp

  break changediscussiondo not closefeature requeststale
  SciSharp/LLamaSharp#684 · Filter by author SanftMonster opened on Apr 21, 2024
  7

• Consider adding Windows on ARM build of llama.dll to LLamaSharp.Backend.Cpu

  backenddo not closefeature requestnative lib
  SciSharp/LLamaSharp#600 · Filter by author david-nordvall opened on Mar 14, 2024
  9

• Examples don't run with CUDA12

  do not closeenhancementnative lib
  SciSharp/LLamaSharp#599 · Filter by author EtienneT opened on Mar 13, 2024
  19

• Thread Safety in llama.cpp

  do not closeUpstream
  SciSharp/LLamaSharp#596 · Filter by author martindevans opened on Mar 12, 2024
  4

• Support cublas computation without requiring CUDA installed

  do not closeenhancementgood first issuehelp wanted
  SciSharp/LLamaSharp#350 · Filter by author SanftMonster opened on Dec 9, 2023
  16