Mingkun Huang

GCC 5 or lower version does work for me. By the way, here is a better setup: https://github.com/HawkAaron/warp-ctc/blob/pytorch_bindings/tensorflow_binding/setup.py#L60

Here is my setup: https://github.com/HawkAaron/warp-ctc/blob/pytorch_bindings/tensorflow_binding/setup.py By the way, it seems that you didn't set `CUDA_HOME`.

I have a better solution here: https://github.com/HawkAaron/warp-ctc/blob/pytorch_bindings/tensorflow_binding/setup.py Both compile tf from source or binary install can use it.

Since the Compute Capability of 2080Ti is 7.5 We need to add `set(CUDA_NVCC_FLAGS "${CUDA_NVCC_FLAGS} -gencode arch=compute_75,code=sm_75")` in CMakeList.txt For more details, please refer to [my configs](https://github.com/HawkAaron/warp-ctc/blob/pytorch_bindings/CMakeLists.txt#L54).

Please follow the instructions here: https://github.com/HawkAaron/warp-ctc/blob/pytorch_bindings/tensorflow_binding/setup.py#L60

Did you compile with GPU ?

Yeah, the hyparameters should be the same.

Sorry for late reply. Maybe you need to create 'data-npy' dir first.