Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on May 10, 2026

• Merge branch 'main' of github.com:lsdjesus/llama-cpp-python

  Show description for b30e414
  LSDJesus

  committed
  b30e414View commit details

  Copy full SHA for b30e414

  Browse repository at this point

Commits on May 9, 2026

• Add build_output.txt and update llama_cpp integration

  Show description for f7314d2
  LSDJesus

  committed
  f7314d2View commit details

  Copy full SHA for f7314d2

  Browse repository at this point

Commits on May 8, 2026

• fix image input

  LSDJesus

  committed
  3ba13b8View commit details

  Copy full SHA for 3ba13b8

  Browse repository at this point

• docs: add Luna fork reference cheat sheet

  LSDJesus

  committed
  cb8493cView commit details

  Copy full SHA for cb8493c

  Browse repository at this point

• feat: extend Luna hooks to Qwen3.5 and Qwen3.5-MoE model builders

  LSDJesus

  committed
  ef6ca35View commit details

  Copy full SHA for ef6ca35

  Browse repository at this point

Commits on May 3, 2026

• ci: upgrade cuda-toolkit to v0.2.35, add checkout to release job

  LSDJesus

  committed
  4d245cdView commit details

  Copy full SHA for 4d245cd

  Browse repository at this point

• docs: clarify penultimate capture as hidden_states[-2] per HF convention

  Show description for 6a2f12a
  LSDJesus

  committed
  6a2f12aView commit details

  Copy full SHA for 6a2f12a

  Browse repository at this point

• ci: pass CUDA architectures explicitly in CMAKE_ARGS

  LSDJesus

  committed
  195cad7View commit details

  Copy full SHA for 195cad7

  Browse repository at this point

• Update llama.py, llama_cpp.py, and vendor/llama.cpp

  LSDJesus

  committed
  60b68a9View commit details

  Copy full SHA for 60b68a9

  Browse repository at this point

• docs: add qwen3-vl hacking guide and penultimate layer extraction

  Show description for b07978a
  LSDJesus

  committed
  b07978aView commit details

  Copy full SHA for b07978a

  Browse repository at this point

• ci: fix release tag_name for manual dispatch

  Show description for 133fc93
  LSDJesus

  committed
  133fc93View commit details

  Copy full SHA for 133fc93

  Browse repository at this point

• ci: replace auditwheel with wheel retag for Linux CUDA builds

  Show description for 75c0193
  LSDJesus

  committed
  75c0193View commit details

  Copy full SHA for 75c0193

  Browse repository at this point

• ci: fix timeout + no-deps for wheel builds

  LSDJesus

  committed
  9d45154View commit details

  Copy full SHA for 9d45154

  Browse repository at this point

• ci: add CUDA wheel build workflow (abi3, sm_75-sm_120)

  LSDJesus

  committed
  bbe5f7dView commit details

  Copy full SHA for bbe5f7d

  Browse repository at this point

• feat: add semantic memory injection research and layer capture/skip C API extensions

  Show description for 1f20c99
  LSDJesus

  committed
  1f20c99View commit details

  Copy full SHA for 1f20c99

  Browse repository at this point

• Update submodule to my fork + my changes

  LSDJesus

  committed
  f79fa45View commit details

  Copy full SHA for f79fa45

  Browse repository at this point

• Update /docs/wiki/index.md

  Show description for de5a1e8
  JamePeng

  committed
  de5a1e8View commit details

  Copy full SHA for de5a1e8

  Browse repository at this point

• docs: add `LlamaGrammar` wiki page

  Show description for d7ed189
  JamePeng

  committed
  d7ed189View commit details

  Copy full SHA for d7ed189

  Browse repository at this point

• feat(grammar): sync JSON schema to GBNF converter with upstream

  Show description for 5336947
  JamePeng

  committed
  5336947View commit details

  Copy full SHA for 5336947

  Browse repository at this point

Commits on May 2, 2026

• Bump version to 0.3.37

  Show description for fe38cbf
  JamePeng

  committed
  fe38cbfView commit details

  Copy full SHA for fe38cbf

  Browse repository at this point

• Update Submodule vendor/llama.cpp b97ebdc..63d93d1

  JamePeng

  committed
  374c0d0View commit details

  Copy full SHA for 374c0d0

  Browse repository at this point

• docs: Update /docs/wiki/index.md

  Show description for 28f842c
  JamePeng

  committed
  28f842cView commit details

  Copy full SHA for 28f842c

  Browse repository at this point

• Update /docs/wiki/contributing-to-wiki.md

  JamePeng

  committed
  cbf15daView commit details

  Copy full SHA for cbf15da

  Browse repository at this point

• Upload /docs/wiki/modules/LlamaSpeculative.md for llama_speculative.py

  Show description for 50aafd4
  JamePeng

  committed
  50aafd4View commit details

  Copy full SHA for 50aafd4

  Browse repository at this point

• docs: update LLM wiki schema to v0.3

  Show description for 4e58a63
  JamePeng

  committed
  4e58a63View commit details

  Copy full SHA for 4e58a63

  Browse repository at this point

• Update docs/wiki/modules/LlamaCache.md and Separate the modules folder

  JamePeng

  committed
  ef618a6View commit details

  Copy full SHA for ef618a6

  Browse repository at this point

• Update Submodule vendor/llama.cpp 660b1b4..b97ebdc

  JamePeng

  committed
  b875c9aView commit details

  Copy full SHA for b875c9a

  Browse repository at this point

Commits on May 1, 2026

• Update /docs/wiki/Llama.md

  JamePeng

  committed
  120c5e2View commit details

  Copy full SHA for 120c5e2

  Browse repository at this point

• Update README.md

  JamePeng

  committed
  cfa0158View commit details

  Copy full SHA for cfa0158

  Browse repository at this point

Commits on Apr 30, 2026

• feat(llama): add fine-grained MoE CPU offloading controls

  Show description for b7064d7
  JamePeng

  committed
  b7064d7View commit details

  Copy full SHA for b7064d7

  Browse repository at this point

• feat(_ggml): implement ggml-backend API bindings and fix type hints

  Show description for 66dd88b
  JamePeng

  committed
  66dd88bView commit details

  Copy full SHA for 66dd88b

  Browse repository at this point

• Update Submodule vendor/llama.cpp 4414c04..660b1b4

  JamePeng

  committed
  db8bf45View commit details

  Copy full SHA for db8bf45

  Browse repository at this point

Commits on Apr 27, 2026

• feat(handler): Support `add_generation_prompt` parameter pass to MTMDChatHandler

  Show description for 969b45f
  JamePeng

  committed
  969b45fView commit details

  Copy full SHA for 969b45f

  Browse repository at this point

• Upload /docs/wiki/LlamaEmbedding.md for llama_embedding.py

  Show description for 5068a80
  JamePeng

  committed
  5068a80View commit details

  Copy full SHA for 5068a80

  Browse repository at this point

• Update Submodule vendor/llama.cpp b760272..4414c04

  JamePeng

  committed
  9d23199View commit details

  Copy full SHA for 9d23199

  Browse repository at this point