Text Generation
Transformers
Safetensors
English
Japanese
llama
conversational
text-generation-inference
Inference Endpoints

Is the license "gemma"?

#1
by torotoki - opened

The license of swallow-70b and swallow-8b is set as "gemma". Is that correct?
image.png

see here

tokyotech-llm org

@torotoki

Instructions-tuned models have the Gemma license:

  • tokyotech-llm/Llama-3.1-Swallow-70B-Instruct-v0.1
  • tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.1
  • tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.2

Base models not have the Gemma license:

  • tokyotech-llm/Llama-3.1-Swallow-70B-v0.1
  • tokyotech-llm/Llama-3.1-Swallow-8B-v0.1

The reason our instruction-tuned models have the Gemma license is that, as @kaisugi introduced, we performed instruction tuning using synthetic data generated with gemma2.

Taishi-N324 changed discussion status to closed

Sign up or log in to comment