cgp-cnn
cgp-cnn copied to clipboard
RuntimeError: The size of tensor a (4) must match the size of tensor b (8) at non-singleton dimension 3 multiprocessing.pool.RemoteTraceback:
gpu_id: 0 , [['input', 0, 0], ['S_ResBlock_64_1', 0, 0], ['S_ConvBlock_128_5', 0, 0], ['S_ConvBlock_64_3', 0, 0], ['S_ResBlock_64_5', 2, 3], ['S_ConvBlock_128_3', 1, 1], ['S_ResBlock_32_1', 3, 1], ['Max_Pool', 4, 6], ['Avg_Pool', 7, 6], ['Sum', 5, 6], ['S_ResBlock_32_1', 8, 7], ['Sum', 9, 10], ['Max_Pool', 11, 10], ['S_ConvBlock_64_1', 12, 10], ['S_ConvBlock_128_3', 13, 13], ['S_ConvBlock_32_3', 14, 14], ['full', 15, 15]]
What should I do if there is a dimensional mismatch in the initialized network structure?