List view
Run an optimised version of the benchmark, which consists of: - Using datasets and dataloaders for embedding generation -- Add hdf5 and multi-hot labels datasets -- Add batch handling for embedders (CpG methylation and histone annotation tasks) -- Add handling of sequences of different lengths (Chromatin accessibility) -- Add chunking of sequences of length higher than max model length (gene finding) -- Add handling of cross validation evaluation method (enhancer annotation) - Use multipe GPUs
Overdue by 10 month(s)•Due by June 20, 2025•2/2 issues closed