Closed Sino0904 closed 9 months ago
Sure, we will release the training data soon! The training pipeline is in OFA, based on Fairseq
The training codes are in this repo: https://github.com/OFA-Sys/OFASys. Please follow their fine-tuning instructions. We find eventually it is really hard to incorporate this whole repo into ours so we give up on that plan.
Will the code for training the model be provided? thanks!