Huggingface models for sillytavern
First you need a huggingface id go here and create one if not have yet: https://huggingface.co/chat/
git clone https://github.com/kainatquaderee/Silly-Hug.git
cd ./Silly-Hug
pip install -r requerements.txt
Fill all .env parameters email password and modelid provided bellow
python server.py
go to sillytavern --> connection-profile --> chat completion --> custom
in custom endpoint put http://127.0.0.1:5000/v1
make sure to turn off streaming in settings.
note: put id in .env for using
| Model ID | Model Name |
|---|---|
| 0 | meta-llama/Llama-3.3-70B-Instruct |
| 1 | Qwen/Qwen2.5-72B-Instruct |
| 2 | CohereForAI/c4ai-command-r-plus-08-2024 |
| 3 | deepseek-ai/DeepSeek-R1-Distill-Qwen-32B |
| 4 | nvidia/Llama-3.1-Nemotron-70B-Instruct-HF |
| 5 | Qwen/QwQ-32B |
| 6 | Qwen/Qwen2.5-Coder-32B-Instruct |
| 7 | google/gemma-3-27b-it |
| 8 | meta-llama/Llama-3.2-11B-Vision-Instruct |
| 9 | NousResearch/Hermes-3-Llama-3.1-8B |
| 10 | mistralai/Mistral-Nemo-Instruct-2407 |
| 11 | microsoft/Phi-3.5-mini-instruct |
I have my pc and I can use 12B models. But it is always great if we have atleast 1 free way to get cloud model service. As Hugging chat is free and opensource, I search A way to get It working. First I searched For tools And then I Foud this wonderful tool https://github.com/Soulter/hugging-chat-api then I created A discussion On sillytavern Github SillyTavern/SillyTavern#3813 to request this feature.
@Cohee1207 said I can fork And add it myself If I want. so here we are now we have this tool. Please help me more to make it even better.
I would appreciate your support and contribution to this project. No need to do extensive work; even small contributions can help:
- Share this project with others to reach more people who may benefit from a free alternative.
- Star and like this repository to show your support.
- Create pull requests and contribute code if you can.
- Contact me on Discord or GitHub for any questions or further collaboration.