Re: Packages that work with LLMs - OpenCoder seems problematic too

On 8/14/26 7:32 AM, Jean Louis wrote:
> I will verify more of the LLMs which claim to be trully free, such as:
> 
> - Moxin github.com/moxin-org/Moxin-LLM -- but trained with the non-free 
> LLM?
> - OLMo (from Allen AI) allenai.org/olmo
> - Marin github.com/dlwh/marin
> - DaVinci github.com/GAIR-NLP/daVinci-LLM
> - SmolLM github.com/huggingface/smollm (good candidate)
> - OpenCoder opencoder-llm.github.io
> 
> ...
> 
> Jean Louis


OpenCode already mentions "The Stack" immediately near the top of the 
page: https://opencoder-llm.github.io/ I again can't find any indication 
on the entire page that there was an attempt to filter that data set for 
licensing. Their entire OpenCoder paper doesn't contain the words 
"license" or "licensing" anywhere in the entire document, see for 
yourself here: https://arxiv.org/html/2411.04905v3

So I don't think OpenCode is limited to training data with open license 
without attribution requirements either. But I could be wrong, check it 
for yourself, as usual this isn't legal advice.

Regards,

Ellie