search

huggingface / alignment-handbook

Robust recipes to align language models with human and AI preferences
https://huggingface.co/HuggingFaceH4
Apache License 2.0
4.2k stars 357 forks source link

Finetuned `zephyr-7b-beta` with internal data generates same reuslts as model `HuggingFaceH4/zephyr-7b-beta` #79

Open wxp16 opened 6 months ago

wxp16 commented 6 months ago

use huggingface pipleline to run inference task, but found finetuned HuggingFaceH4/zephyr-7b-beta and model HuggingFaceH4/zephyr-7b-beta generates exactly otuputs. Does anyone have any clue about this error?

timothylimyl commented 6 months ago

given you put the same model-id, it is most probably downloading the hugging face official repo weights to run the inference.