marella/ctransformers
1
0
Fork 0
Code Issues 107 Pull requests 5 Projects Releases 30 Packages Wiki Activity Actions
30 releases 30 tags
  • v0.2.17 9041ba5aa0

    0.2.17 Stable

    marella released this 2023-08-01 09:55:14 -04:00 | 44 commits to main since this release

    Changes

    • Fix CPU feature detection for ARM processors
  • v0.2.16 953748a586

    0.2.16 Stable

    marella released this 2023-07-30 14:26:42 -04:00 | 46 commits to main since this release

    Changes

    • Auto detect CPU features AVX2, AVX
  • v0.2.15 71f2e50071

    0.2.15 Stable

    marella released this 2023-07-28 19:46:52 -04:00 | 50 commits to main since this release

    Changes

    • Add support for LLaMA 2 70B models
    • Update ggml and llama.cpp
  • v0.2.14 5a92fabd44

    0.2.14 Stable

    marella released this 2023-07-20 14:50:03 -04:00 | 56 commits to main since this release

    Changes

    • Add Metal support for LLaMA models
  • v0.2.13 12efc56196

    0.2.13 Stable

    marella released this 2023-07-17 14:57:49 -04:00 | 61 commits to main since this release

    Changes

    • Add precompiled libraries for CUDA 12
  • v0.2.12 81d2a8c185

    0.2.12 Stable

    marella released this 2023-07-15 11:11:53 -04:00 | 66 commits to main since this release

    Changes

    • Add support for GGCC model format for Falcon models
  • v0.2.11 9937314b08

    0.2.11 Stable

    marella released this 2023-07-03 15:32:04 -04:00 | 69 commits to main since this release

    Changes

    • Add support for Replit models
    • Increase default number of threads based on hardware threads
  • v0.2.10 20d601b025

    0.2.10 Stable

    marella released this 2023-06-22 16:15:56 -04:00 | 75 commits to main since this release

    Changes

    • Add experimental support for Falcon models
  • v0.2.9 f585491d28

    Stable

    marella released this 2023-06-18 14:23:39 -04:00 | 79 commits to main since this release

    Changes

    • Handle special tokens properly while tokenizing
    • Handle incomplete UTF-8 multi-byte characters while generating text
    • Increase buffer size in tokenize() for BOS token
    • Update ggml and llama.cpp
  • v0.2.8 653742a2f1

    0.2.8 Stable

    marella released this 2023-06-12 11:14:20 -04:00 | 86 commits to main since this release

    Changes

    • Add support for the new k-quantization formats
    • Update GGML