?????? ?????? How to configure the TensorFlow environment so that it can be used on gnuradio?

[????????]

Hi

Thank you for your reply. I have tried this operation method, but the problem still exists. I will re-describe my operation process:

First, I used the gr_modtool tool to create an OOT module

Then, I used the gr_modtool tool again to add a .py file (this is my block file)

Since the .py file in the previous step needs to use the "util_model.py" file, I will copy this file directly to the /python directory

Then execute, cmake, make, make install, make ldconfig, create a simple flow diagram to test this module, but an error "no module named util_model" is reported

So according to your method, change the CMaketxt file, delete the build directory and create a new one, re-execute cmake..., but "no module named util_model" still exists

In addition, I found that two warnings appeared when I executed the make command??Attached picture??

I don??t know if the reason for this error is related to it

Regards,

linge93

------------------ ???????? ------------------

??????: "Vasil Velichkov" <vvvelichkov@gmail.com>;

????????: 2021??11??24??(??????) ????10:20

??????: "????????"<2127629883@qq.com>;

????: "discuss-gnuradio"<discuss-gnuradio@gnu.org>;

????: Re: ?????? How to configure the TensorFlow environment so that it can be used on gnuradio?

Hi linge93,

On 24/11/2021 15.24, ???????? wrote:
> I copied the "util_model.py" file directly to the /python directory

>  I now know that I can??t just put it under the gr_oot/python folder

Open gr_oot/python/CMakeLists.txt and add util_model.py in the GR_PYTHON_INSTALL files, something similar to

GR_PYTHON_INSTALL(
    FILES
    __init__.py
    util_model.py
    DESTINATION ${GR_PYTHON_DIR}/oot
)

Add in case you are using git for source control don't forget to add it to the repository.

git add python/util_model.py

Regards,
Vasil

error.png
Description: Binary data