FAQ 🙋¶

Q. Is there a limitation on the computational resources?
: Based on the GC documentation, it has 16GB of GPU memory limitation (NVIDIA-T4). Also, CPU time is limited to 2 hours. Note that we use batch size 1 for the inference, so there will be less chance of having a CUDA OOM.

Q. Is the limitation on the number of submissions applied to a team or participant?
:  It is applied to a team, i.e. one team can submit up to 10 times and 1 time for validation and test phase, respectively.

Q. I got an error with "Algorithm failed on one or more cases", what is the reason?
: Detailed log is hidden from participants to prevent information leakage. For example, someone can print(image, label) to get the information of the hidden test sets. To access the detailed log, please send an email to oncology-ai-research@lunit.io. Before sending an email, make sure to do and pass a local test explained here. 

Q. Docker image is not getting activated.
: It can happen when the Grand Challenge server is unstable. It is usually solved when you try a few hours later. If it is not solved, please send an email to support@grand-challenge.org.