This repository was archived by the owner on Nov 1, 2024. It is now read-only.

Description
🚀 Feature Request
Congratulations on this great work - Scaling Autoregressive Multi-Modal Models:Pretraining and Instruction Tuning.
The paper mentions that metaseq was used for the model training. Just wondering if there is a plan to release the code for these experiments and/or the model weights associated with the various models trained in the paper.
Thanks!