Please ad finetune / chat model tag
Is this not a finetune? Why does it have the "pretrained" tag?
Please consider changing the tag as people might get the wrong idea. If I understand SPIN right, then this is NOT a "pretrained" model as defined by the leaderboard with "new, base models, trained on a given text corpora using masked modelling", as SPIN involves a reward function derived from a finetuning dataset, which in this case is the ultrachat_200k dataset. Instead, the chat model tag might fit better.
Is this not a finetune? Why does it have the "pretrained" tag?
Please consider changing the tag as people might get the wrong idea. If I understand SPIN right, then this is NOT a "pretrained" model as defined by the leaderboard with "new, base models, trained on a given text corpora using masked modelling", as SPIN involves a reward function derived from a finetuning dataset, which in this case is the ultrachat_200k dataset. Instead, the chat model tag might fit better.
Hi, thanks for your advice. Do you know how how to change the tag? I try to resubmit to the llm leaderboard to change the tag. But it said that 'This model has been already submitted.'.
You can open a pull-request to your models submitted to the leaderboard by modifying the request file via the "contribute" button: https://huggingface.co/datasets/open-llm-leaderboard/requests/tree/main/amu
You can open a pull-request to your models submitted to the leaderboard by modifying the request file via the "contribute" button: https://huggingface.co/datasets/open-llm-leaderboard/requests/tree/main/amu
https://huggingface.co/datasets/open-llm-leaderboard/requests/discussions/84 I has opened a pr for it. It seems ok. Thank you again!
Thank you and you are welcome :-)