brucethemoose's picture
Update README.md
5a40fc8
|
raw
history blame
1.48 kB
metadata
license: other
license_name: yi-license
license_link: https://huggingface.co/01-ai/Yi-34B/blob/main/LICENSE
language:
  - en
library_name: transformers
pipeline_tag: text-generation

NousResearch/Nous-Capybara-34B and migtissera/Tess-M-Creative-v1.0 ties merged with mergekit.

I would suggest an exllama version for local inference with 40K+ context: https://huggingface.co/brucethemoose/Capybara-Tess-Yi-34B-200K-exl2-4bpw-fiction https://huggingface.co/brucethemoose/Capybara-Tess-Yi-34B-200K-exl2-31bpw-fiction

he following config:

models:
  - model: /home/alpha/Storage/Models/Raw/larryvrh_Yi-34B-200K-Llamafied
    # no parameters necessary for base model
  - model: /home/alpha/Storage/Models/Raw/migtissera_Tess-M-v1.0
    parameters:
      density: 0.6
      weight: 1.0
  - model: /home/alpha/Storage/Models/Raw/Nous-Capybara-34B
    parameters:
      density: 0.6
      weight: 1.0
merge_method: ties
base_model: //home/alpha/Storage/Models/Raw/larryvrh_Yi-34B-200K-Llamafied
parameters:
  normalize: true
  int8_mask: true
dtype: float16

Both are 200K context models with Vicuna syntax, so:

Prompt Format:

SYSTEM: ...
USER: ...
ASSISTANT: ...

Stop token: </s>


Credits:

https://github.com/cg123/mergekit

https://huggingface.co/NousResearch/Nous-Capybara-34B/discussions

https://huggingface.co/migtissera/Tess-M-Creative-v1.0

https://huggingface.co/larryvrh/Yi-34B-200K-Llamafied

https://huggingface.co/01-ai/Yi-34B-200K