dahara1 commited on
Commit
d71921c
β€’
1 Parent(s): 9e095ee

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -4,8 +4,8 @@ language:
4
  ---
5
 
6
  ## ζœ¬γƒ’γƒ‡γƒ«γ«γ€γ„γ¦ about this model.
7
- [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)γ‚’[ζ—₯本θͺžγŒε€šγε«γΎγ‚Œγ‚‹ι‡θ¦εΊ¦θ‘Œεˆ—(iMatrix)](dahara1/imatrix-jpn-test)γ‚’δ½Ώγ£γ¦ι‡ε­εŒ–γ—γ€θΆ…ι•·ζ–‡(32Kδ»₯上)θ¦η΄„γ‚’ε―θƒ½γ«γ—γŸggufη‰ˆγ§γ™γ€‚ζ—₯本θͺžε―ΎεΏœθƒ½εŠ›γŒε€šγ‚γ«δΏζŒγ•γ‚Œγ¦γ„γ‚‹δΊ‹γ‚’ζœŸεΎ…γ—γ¦γ„γΎγ™γ€‚
8
- This is a gguf version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) that has been quantized using [importance matrix (iMatrix) that contains a lot of Japanese](dahara1/imatrix-jpn-test) to enable summarization of long texts (over 32K). We hope that it retains a large amount of Japanese support.
9
 
10
  ε°‘γͺくともQwen2.5-3B-Instruct-gguf-japanese-imatrix-128K/Qwen2.5-3B-Instruct-Q8_0-f16.ggufが32KγƒˆγƒΌγ‚―γƒ³γ‚’θΆ…γˆγ‚‹θΆ…ι•·ζ–‡γ‚’ζ­£γ—γθ¦η΄„γ§γγ‚‹δΊ‹γ‚’η’ΊθͺζΈˆγ§γ™γ€‚
11
  It has been confirmed that at least Qwen2.5-3B-Instruct-gguf-japanese-imatrix-128K/Qwen2.5-3B-Instruct-Q8_0-f16.gguf can correctly summarize extremely long texts exceeding 32K tokens.
 
4
  ---
5
 
6
  ## ζœ¬γƒ’γƒ‡γƒ«γ«γ€γ„γ¦ about this model.
7
+ [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct)γ‚’[ζ—₯本θͺžγŒε€šγε«γΎγ‚Œγ‚‹ι‡θ¦εΊ¦θ‘Œεˆ—(iMatrix)](https://huggingface.co/dahara1/imatrix-jpn-test)γ‚’δ½Ώγ£γ¦ι‡ε­εŒ–γ—γ€θΆ…ι•·ζ–‡(32Kδ»₯上)θ¦η΄„γ‚’ε―θƒ½γ«γ—γŸggufη‰ˆγ§γ™γ€‚ζ—₯本θͺžε―ΎεΏœθƒ½εŠ›γŒε€šγ‚γ«δΏζŒγ•γ‚Œγ¦γ„γ‚‹δΊ‹γ‚’ζœŸεΎ…γ—γ¦γ„γΎγ™γ€‚
8
+ This is a gguf version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) that has been quantized using [importance matrix (iMatrix) that contains a lot of Japanese](https://huggingface.co/dahara1/imatrix-jpn-test) to enable summarization of long texts (over 32K). We hope that it retains a large amount of Japanese support.
9
 
10
  ε°‘γͺくともQwen2.5-3B-Instruct-gguf-japanese-imatrix-128K/Qwen2.5-3B-Instruct-Q8_0-f16.ggufが32KγƒˆγƒΌγ‚―γƒ³γ‚’θΆ…γˆγ‚‹θΆ…ι•·ζ–‡γ‚’ζ­£γ—γθ¦η΄„γ§γγ‚‹δΊ‹γ‚’η’ΊθͺζΈˆγ§γ™γ€‚
11
  It has been confirmed that at least Qwen2.5-3B-Instruct-gguf-japanese-imatrix-128K/Qwen2.5-3B-Instruct-Q8_0-f16.gguf can correctly summarize extremely long texts exceeding 32K tokens.