Closed WhenMelancholy closed 7 months ago
Thank you for your reminder. I will work on this issue and will keep you posted.
Hi,
I did double check that the detailed dataset list was provided in the supplementary file in Table 6 in our ICLR submission here on openreview. This was added during the rebuttal. Meanwhile, in the BioRXiv version we haven't updated this information, and we will replace it with our camera-ready version soon. Sorry for causing the confusion.
Note that in Septemeber (right after the submission) we have trained a new version of the model with CellxGene Census LTS 2023-05-15. Therefore, any checkpoints after September was based on this data source.
Thank you for your help! I will check the ICLR version.
Hello, I noticed that the paper mentioned updates on the details of the training dataset would be made available in the GitHub repository. May I ask if there are plans to update this part? For instance, which parts of the HTCA HCA GEO dataset were used? Thank you!