Did you train the whole LLM in the pretraining stage of share recipe？

[Yang-bug-star]

[baichuanzhou]

No, vision tower was not trained.

[Yang-bug-star]

oh，what i meant is that whether the language model was pre-trained in share training recipe, it's clear  that the  vision encoder was trained in pretraining of share recipe. 

---Original--- From: @.> Date: Wed, Mar 20, 2024 19:10 PM To: @.>; Cc: @.@.>; Subject: Re: [DLCV-BUAA/TinyLLaVABench] Did you train the whole LLM in the pretraining stage of share recipe？ (Issue #34)

No, vision tower was not trained.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[baichuanzhou]

Then, yes.