dvlab-research / LLaMA-VID

Official Implementation for LLaMA-VID: An Image is Worth 2 Tokens in Large Language Models
Apache License 2.0
622 stars 39 forks source link

Computation costs for each stage? #70

Closed Becomebright closed 3 months ago

Becomebright commented 3 months ago

Thank you for your inspiring work. Could you provide details on the computational costs, such as GPU hours, for each stage of training? I think this information would greatly assist in understanding and following your work.

Becomebright commented 3 months ago

Sorry, I've found the duplicated issue.