Text Generation
Transformers
NeMo
Safetensors
mistral
text-generation-inference
Inference Endpoints

How to efficiently use this model?

#3
by Varkoyote - opened

Hello! I just wanted to know if there was a specific way to use this model please, because for me it's just generating nonsense and repeating itself very quickly... thank you!

NVIDIA org
β€’
edited Aug 22

Try using top_p = 0.95, temperature = 0.7, and frequency_penalty = 0.5 for less repetitive completions. However, typically, the best way to use this model is to fine-tune it/align it to follow instructions.

Hello! I just wanted to know if there was a specific way to use this model please, because for me it's just generating nonsense and repeating itself very quickly... thank you!

This isn't instruction tuned. It's a base model which people will fine tune on top of.

Sign up or log in to comment