My sample size is approximately 1500,these samples comes from 58 studies. I according to DA-YourOwnData-Tutorial.ipynb to train DeepAdapter,I use CPU rather than GPU to train it. In part 3.3. Train DeepAdapter,everything is OK. But when align the data,it takes so many many times,more than 10 hours and it not finished,the output is just an UMAP picture. What can I do to solve this problem. Thank you very much!