rpg_timelens
rpg_timelens copied to clipboard
question about training data
The paper mentioned "We finetune our network on real data module-by-module in the order of training" in section 3, so can the real data for fine-tuning be released for our to train the model to reproduce the results.