Trainable parameters in each stage

Thanks for great work!

I have a question about trainable parameters in each stage.

Here's what I think, but is it right?
- Stage 1: Vision Encoder + Projector
- Stage 2: Vision Encoder + Projector + LLM
- Stage 3: Vision Encoder + Video Compressor + Projector + LLM (**I'm curious about this part.**)
- Stage 4: Vision Encoder + Video Compressor + Projector + LLM

Thank you.

Start by reviewing the repository's training-stage documentation and Jupyter notebooks to identify which components are trainable in stages 1–4. Done means documenting or confirming the parameter status for every listed component, with an explicit answer for the stage 3 video compressor.

Written by the indexing model from the issue text.