Closed VoyageWang closed 1 week ago
Thank you for your attention to our work. The code is being sorted out.
Pre-training requires 8 RTX3090 to train for about 1 day. A single dataset such as refcoco/+/g only requires one RTX3090 to train for about half a day.
Hi, when will the code be released? The training computational resources you used are a single 3090 GPU or 8 GPUs, which is not clearly stated in your paper. Thanks for your great work!