How many non-trainable parameters exist in the Sudoku-solving CNN architecture?
How many non-trainable parameters exist in the Sudoku-solving CNN architecture?: a task in SmolDataEnvs-harbor-train-sandbox: 📊 SmolDataEnvs: Harbor (train) (Harbor dataset). Files (in /home/user/input, no subfolders): - sudoku.csv
The task
Files (in /home/user/input, no subfolders): - sudoku.csv