Skip to content

GitLab

Switch branch/tag
  1. 16 Nov, 2023 1 commit
    • Zhuohan Li's avatar
      TP/quantization/weight loading refactor part 2 - Refactor quantized linear... · 7076fa1c
      Zhuohan Li authored 
      TP/quantization/weight loading refactor part 2 - Refactor quantized linear logic and extend quantization support to all models (#1622)

Refactor the tensor parallelism, quantization, and weight-loading codes.

Summary of the new features enabled by this PR:
- **All models** are able to be quantized with AWQ and SqueezeLLM, and [soon GPTQ](https://github.com/vllm-project/vllm/pull/1580).
- Model loading code became much simpler.
- Support model parallelism for all MQA/GQA models when the number of key/value heads is smaller than the tensor parallel size.
      7076fa1c
  3. 02 Oct, 2023 1 commit
  5. 13 Sep, 2023 1 commit
  7. 07 Sep, 2023 1 commit
  9. 05 Sep, 2023 1 commit
  11. 03 Jul, 2023 1 commit
  13. 17 Jun, 2023 1 commit
  15. 25 May, 2023 1 commit
  17. 24 May, 2023 1 commit
  19. 19 May, 2023 1 commit
  21. 15 May, 2023 3 commits
  23. 09 May, 2023 1 commit
  25. 05 May, 2023 1 commit
  27. 04 May, 2023 1 commit
  29. 03 May, 2023 1 commit
  31. 28 Apr, 2023 1 commit
  33. 09 Apr, 2023 1 commit
  35. 02 Apr, 2023 2 commits
  37. 30 Mar, 2023 2 commits
  39. 21 Mar, 2023 1 commit
  41. 10 Mar, 2023 1 commit
  43. 25 Feb, 2023 1 commit
  45. 23 Feb, 2023 3 commits
  47. 22 Feb, 2023 1 commit
  49. 09 Feb, 2023 1 commit