PyTorch bindings for CUDA-Warp Recurrent Neural Aligner. ... Size) – The batch shape of the inputs (asssuming input tensors of shape batch_shape x n x d ).
確定! 回上一頁