Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit history

Commits on Jan 6, 2026

• feat: Add Granite-Docling model support with 'GraniteDoclingChatHandler'

  Show description for 80279c4
  JamePeng

  committed
  80279c4View commit details

  Copy full SHA for 80279c4

  View code at this pointBrowse repository at this point

• Update llama.cpp API 20260106

  JamePeng

  committed
  5ea1285View commit details

  Copy full SHA for 5ea1285

  View code at this pointBrowse repository at this point

Commits on Jan 5, 2026

• Update llama.cpp API 20260105

  JamePeng

  committed
  bf09ae5View commit details

  Copy full SHA for bf09ae5

  View code at this pointBrowse repository at this point

• Synchronize Pillow documents with supported image MIME types for conversion, and remove image formats that are not supported for reading and conversion.

  JamePeng

  committed
  6b31e8dView commit details

  Copy full SHA for 6b31e8d

  View code at this pointBrowse repository at this point

Commits on Jan 4, 2026

• Bump version to 0.3.20

  JamePeng

  committed
  c7d53ceView commit details

  Copy full SHA for c7d53ce

  View code at this pointBrowse repository at this point

Commits on Jan 3, 2026

• Small fix for Llava15ChatHandler class

  JamePeng

  committed
  c7721b4View commit details

  Copy full SHA for c7721b4

  View code at this pointBrowse repository at this point

Commits on Jan 1, 2026

• Update llama_vocab_pre_type varriable

  JamePeng

  committed
  6f7fa11View commit details

  Copy full SHA for 6f7fa11

  View code at this pointBrowse repository at this point

• Fixed `embeddings` typos

  JamePeng

  committed
  8a96113View commit details

  Copy full SHA for 8a96113

  View code at this pointBrowse repository at this point

• Update llama_context_params and fix the `embeddings` typo

  JamePeng

  committed
  b6bf564View commit details

  Copy full SHA for b6bf564

  View code at this pointBrowse repository at this point

Commits on Dec 31, 2025

• fix(Llama): implement fallback to full cache clear in eval

  Show description for 644a8a2
  JamePeng

  committed
  644a8a2View commit details

  Copy full SHA for 644a8a2

  View code at this pointBrowse repository at this point

• Bump version to 0.3.19

  Show description for 2ff40cc
  JamePeng

  committed
  2ff40ccView commit details

  Copy full SHA for 2ff40cc

  View code at this pointBrowse repository at this point

• refactor(Llama): enhance error handling and cleanup in `eval` method

  Show description for 06f64ae
  JamePeng

  committed
  06f64aeView commit details

  Copy full SHA for 06f64ae

  View code at this pointBrowse repository at this point

Commits on Dec 30, 2025

• Update Submodule vendor/llama.cpp d77d7c5..4849661

  JamePeng

  committed
  6d31ab0View commit details

  Copy full SHA for 6d31ab0

  View code at this pointBrowse repository at this point

• Correct incorrect assignments in llama.py

  JamePeng

  committed
  8f7211cView commit details

  Copy full SHA for 8f7211c

  View code at this pointBrowse repository at this point

• Remove unnecessary checks in LlamaContext Class

  JamePeng

  committed
  3e00e8eView commit details

  Copy full SHA for 3e00e8e

  View code at this pointBrowse repository at this point

• refactor(LlamaBatch): enhance safety checks and fix indexing logic

  Show description for 93c9f7e
  JamePeng

  committed
  93c9f7eView commit details

  Copy full SHA for 93c9f7e

  View code at this pointBrowse repository at this point

• Update llama.cpp API 20251230

  JamePeng

  committed
  3716d0aView commit details

  Copy full SHA for 3716d0a

  View code at this pointBrowse repository at this point

Commits on Dec 27, 2025

• Fix typos and zero-pad process

  JamePeng

  committed
  eff3b2bView commit details

  Copy full SHA for eff3b2b

  View code at this pointBrowse repository at this point

• Improve Extract Embeddings Branch Code

  JamePeng

  committed
  71281d0View commit details

  Copy full SHA for 71281d0

  View code at this pointBrowse repository at this point

• Added n_ctx,n_batch,n_ubatch parameter into LlamaEmbedding

  JamePeng

  committed
  fcdf9acView commit details

  Copy full SHA for fcdf9ac

  View code at this pointBrowse repository at this point

• Update LlamaModel api-call into _internals.py

  JamePeng

  committed
  fb08475View commit details

  Copy full SHA for fb08475

  View code at this pointBrowse repository at this point

• Added n_gpu_layers parameter into LlamaEmbedding

  JamePeng

  committed
  573dba7View commit details

  Copy full SHA for 573dba7

  View code at this pointBrowse repository at this point

• Docs: Update README to feature LlamaEmbedding and Reranking workflows

  Show description for e100501
  JamePeng

  committed
  e100501View commit details

  Copy full SHA for e100501

  View code at this pointBrowse repository at this point

• Refactor: Extract embedding logic to `LlamaEmbedding` class, Rerank support and fix parallel batching

  Show description for 5c424ab
  JamePeng

  committed
  5c424abView commit details

  Copy full SHA for 5c424ab

  View code at this pointBrowse repository at this point

• Update llama.cpp API 20251227

  JamePeng

  committed
  a11d97aView commit details

  Copy full SHA for a11d97a

  View code at this pointBrowse repository at this point

Commits on Dec 26, 2025

• Fixed a small bug in the Qwen3-VL chat template.

  alcoftTAO

  committed
  eb4749cView commit details

  Copy full SHA for eb4749c

  View code at this pointBrowse repository at this point

Commits on Dec 19, 2025

• Bump version to 0.3.18

  Show description for 171bd19
  JamePeng

  committed
  171bd19View commit details

  Copy full SHA for 171bd19

  View code at this pointBrowse repository at this point

• feat: implement GLM46VChatHandler for GLM-4.6V Series Model

  Show description for d5131e2
  JamePeng

  committed
  d5131e2View commit details

  Copy full SHA for d5131e2

  View code at this pointBrowse repository at this point

• More Stop token patch for GLM41VChatHandler

  JamePeng

  committed
  70d5844View commit details

  Copy full SHA for 70d5844

  View code at this pointBrowse repository at this point

• feat: implement LFM2VLChatHandler for LFM2-VL series models

  JamePeng

  committed
  060f06dView commit details

  Copy full SHA for 060f06d

  View code at this pointBrowse repository at this point

• Optimization: Improved batch token processing logic in Llava15ChatHandler.

  JamePeng

  committed
  9b9710aView commit details

  Copy full SHA for 9b9710a

  View code at this pointBrowse repository at this point

Commits on Dec 17, 2025

• feat: implement `GLM41VChatHandler` for GLM-4.1V-9B-Thinking Model

  Show description for 236eff5
  JamePeng

  committed
  236eff5View commit details

  Copy full SHA for 236eff5

  View code at this pointBrowse repository at this point

Commits on Dec 15, 2025

• Sync llama.cpp API 20251215

  JamePeng

  committed
  c98ce3bView commit details

  Copy full SHA for c98ce3b

  View code at this pointBrowse repository at this point

Commits on Dec 9, 2025

• Fixed typo

  JamePeng

  committed
  b2aa656View commit details

  Copy full SHA for b2aa656

  View code at this pointBrowse repository at this point

Commits on Dec 7, 2025

• feat: Added the scan path for CUDA 13.0+ dynamic link libraries under Windows system ($env:CUDA_PATH\bin\x64)

  Show description for e5e19e8
  JamePeng

  committed
  e5e19e8View commit details

  Copy full SHA for e5e19e8

  View code at this pointBrowse repository at this point