Text Generation
Transformers
Safetensors
English
German
llama
goliath
deutsch
llama2
discoresearch
Inference Endpoints
text-generation-inference
No description provided.

Hey there

Here is the PR which fixes the stopping issues with GPT-Q and GGUF quants.
The fix is to set the eos / bos tokens to the config.

Kind regards,
Timon Käch

CyberTimon changed pull request status to open
Disco Research org

lgtm, ty!

jphme changed pull request status to merged

Sign up or log in to comment