Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 22, 2025

• Install libstdcxx-ng from conda-forge for Linux

  Uminosachi

  committed
  ad0afe9

  Copy full SHA for ad0afe9

• Add google/gemma-3-4b-it to multimodal list

  Uminosachi

  committed
  00a3e94

  Copy full SHA for 00a3e94

Commits on Feb 15, 2025

• Update README for macOS installation

  Uminosachi

  committed
  b0fd56a

  Copy full SHA for b0fd56a

Commits on Feb 12, 2025

• Add regular expression state to is_match function

  Uminosachi

  committed
  aec888d

  Copy full SHA for aec888d

Commits on Jan 29, 2025

• Remove space after @ in decorator

  Uminosachi

  committed
  abc57dc

  Copy full SHA for abc57dc

• Add paligemma2-3b-pt-224,448 to multimodal list

  Uminosachi

  committed
  8806882

  Copy full SHA for 8806882

Commits on Dec 15, 2024

• Add continuous processing of multiple prompts

  Uminosachi

  committed
  82d866f

  Copy full SHA for 82d866f

Commits on Dec 14, 2024

• Add instruction to wget and install builded wheel

  Uminosachi

  committed
  df2c7db

  Copy full SHA for df2c7db

Commits on Oct 15, 2024

• Set gradio to less than 5.0.0 in requirements.txt

  Uminosachi

  committed
  d58c337

  Copy full SHA for d58c337

Commits on Oct 8, 2024

• Fix Llama-3.2 Vision chat template for instruction

  Uminosachi

  committed
  87a525e

  Copy full SHA for 87a525e

Commits on Oct 7, 2024

• Move text processing bar reset out of model

  Uminosachi

  committed
  a2aae8e

  Copy full SHA for a2aae8e

Commits on Oct 5, 2024

• Update README for Llama-3.2 Vision model

  Uminosachi

  committed
  e2051d3

  Copy full SHA for e2051d3

• Set MllamaProcessor for Llama-3.2 Vision model

  Uminosachi

  committed
  50dc2f5

  Copy full SHA for 50dc2f5

• Add limited trial support for Llama-3.2-11B-Vision

  Uminosachi

  committed
  1ad393d

  Copy full SHA for 1ad393d

Commits on Oct 3, 2024

• Remove trust_remote_code=True for local module

  Uminosachi

  committed
  9736736

  Copy full SHA for 9736736

• Support inference of Phi-3.5-vision-instruct

  Uminosachi

  committed
  2accbf4

  Copy full SHA for 2accbf4

Commits on Sep 30, 2024

• Remove model to be incompatible with transformers

  Uminosachi

  committed
  8546418

  Copy full SHA for 8546418

• Add patch_size and feature_select_strategy args

  Uminosachi

  committed
  d509b3c

  Copy full SHA for d509b3c

• Use image first input when transformers is newer

  Uminosachi

  committed
  d10beb6

  Copy full SHA for d10beb6

Commits on Sep 29, 2024

• Set attn_impl to eager if flash-attn doesn't exist

  Uminosachi

  committed
  04f85d6

  Copy full SHA for 04f85d6

• Remove underperforming quantized model

  Uminosachi

  committed
  3301975

  Copy full SHA for 3301975

• Modify ImageProcessor inputs order to image first

  Uminosachi

  committed
  9fc5e7e

  Copy full SHA for 9fc5e7e

• Remove auto-gptq --extra-index-url for CUDA 12.1

  Uminosachi

  committed
  729698c

  Copy full SHA for 729698c

• Fix Visual Studio installation link and correct typo

  Uminosachi

  committed
  09ceae3

  Copy full SHA for 09ceae3

Commits on Sep 28, 2024

• Revert CUDA version of auto-gptq URL to v11.8

  Uminosachi

  committed
  441a173

  Copy full SHA for 441a173

• Change default CUDA version from v11.8 to v12.1

  Uminosachi

  committed
  577dbbf

  Copy full SHA for 577dbbf

• Add code to test Llama-3.2-Vision-Instruct-4bit

  Uminosachi

  committed
  b156762

  Copy full SHA for b156762

Commits on Aug 18, 2024

• Support inference of MiniCPM-V-2_6 and int4 models

  Uminosachi

  committed
  c994a35

  Copy full SHA for c994a35

Commits on Aug 16, 2024

• Set CMAKE_BUILD_PARALLEL_LEVEL to speed up build

  Uminosachi

  committed
  04c95b9

  Copy full SHA for 04c95b9

• Change -DLLAMA_CUDA to -DGGML_CUDA in CMAKE_ARGS

  Uminosachi

  committed
  81b75ba

  Copy full SHA for 81b75ba

• Add /utf-8 flag to CMAKE_CXX_FLAGS for Windows

  Uminosachi

  committed
  946570c

  Copy full SHA for 946570c

• Add installation instructions for OpenMP libraries

  Uminosachi

  committed
  9b7a307

  Copy full SHA for 9b7a307

Commits on Aug 15, 2024

• Fix jinja2 template escape chars from two to one

  Uminosachi

  committed
  0db1ca8

  Copy full SHA for 0db1ca8

Commits on Aug 6, 2024

• Support inference of EvoVLM-JP-v1-7B model

  Uminosachi

  committed
  7f6e95f

  Copy full SHA for 7f6e95f

Commits on Jul 16, 2024

• Remove Phi-3-small due to quantization issues

  Uminosachi

  committed
  051fc46

  Copy full SHA for 051fc46

Pagination

Previous

Next