bibidentuhanoi

/

BMO-8B-Instruct-Llama3

like 0

Hi bibidentuhanoi,

Thank you for sharing the BMO-8B-Instruct-Llama3 model; the integration with Unsloth makes it very efficient for my current project.

I am currently setting up a workflow based on your model, but I noticed a slight discrepancy regarding the base model information which I hope you could clarify:

The Model Card mentions the base is cognitivecomputations/dolphin-2.9-llama3-8b, but some metadata in the Hugging Face model tree points towards dphn/dolphin-2.9-llama3-8b. Could you confirm which specific repository was used as the direct source for your fine-tuning?

I want to ensure I’m using the exact corresponding base weights to maintain consistency during further development:)

Thanks again for your contribution to the community!