Xkev

/

Llama-3.2V-11B-cot

like 158

Discussion

bdytx5
2 days ago

Could you release the base fine tune model without the CoT training? I am writing an article on this. Thanks

Xkev
Owner
2 days ago
•
edited 2 days ago

Hi, the base model is Llama-3.2-11B-Vision-Instruct: https://huggingface.co/meta-llama/Llama-3.2-11B-Vision-Instruct

Xkev changed discussion status to closed 2 days ago

bdytx5
2 days ago

Sorry, I mean the "Direct Training" model on the LLaVA 100k dataset (eg just trained on the answers rather than the entire CoT data). from the paper "Here, LLaVA-o1 (with Direct Training) refers to the model trained directly on the original VQA dataset’s Q&A pairs"

Leave a comment