Specify Python, Keras, and Tensorflow versions

[Darelbi]

I have been able to install the Keras.NET nuget package, guided by runtime exceptions I also installed

Python 3.8.10 (which is the last 3.8.x version that comes with a installer)
added python to PATH.

I also installed PIP by running "python get-pip.py" script.

Here comes the problems:

I can use pip to install tensorflow and keras, HOWEVER the installed versions mismatch with Keras.NET, consequently

I get this error when I try to train a deep neural network:

: 'ValueError : decay is deprecated in the new Keras optimizer, please check the docstring for valid arguments, or use the legacy optimizer, e.g., tf.keras.optimizers.legacy.Adam.'

Then I try as optimizer the string

"tf.keras.optimizers.legacy.Adam"

and I go a step on => dude, still missing tensorflow.keras...

'ModuleNotFoundError : No module named 'tensorflow.keras''

AND THIS ONE IS NOT FOUND ON PIP

Installed also Tensorflow.Keras nuget package but then it keeps insisting saying

ImportError : No module named tensorflow

but if I do "pip show tensorflow" it shows tensorflow is installed, I can even run pure python example scripts with tensorflow

without effect.. that means i need to use even older versions?

SINCE NEWEST VERSIONS GIVES PROBLEMS I TRIED VERSIONS OF 2020:
pip install keras==2.6.0
pip install tensorflow==2.3.2

but then i have this error:
Python.Runtime.PythonException: 'TypeError : Descriptors cannot be created directly.
If this call came from a _pb2.py file, your generated code is out of date and must be regenerated with protoc >= 3.19.0.
If you cannot immediately regenerate your protos, some other possible workarounds are:

1. Downgrade the protobuf package to 3.20.x or lower.
2. Set PROTOCOL_BUFFERS_PYTHON_IMPLEMENTATION=python (but this will use pure-Python parsing and will be much slower).

More information: https://developers.google.com/protocol-buffers/docs/news/2022-05-06#python-updates'

SOMEHOW I MANAGED playing with numpy and protobuf libraries: here are my current packages:

absl-py 2.1.0
astunparse 1.6.3
cachetools 5.3.3
certifi 2024.2.2
charset-normalizer 3.3.2
flatbuffers 24.3.7
gast 0.3.3
google-auth 2.28.2
google-auth-oauthlib 1.0.0
google-pasta 0.2.0
grpcio 1.62.1
h5py 2.10.0
idna 3.6
importlib_metadata 7.0.2
keras 2.6.0
Keras-Applications 1.0.8
Keras-Preprocessing 1.1.2
libclang 16.0.6
Markdown 3.6
MarkupSafe 2.1.5
numpy 1.18.0
oauthlib 3.2.2
opt-einsum 3.3.0
packaging 24.0
pip 24.0
protobuf 3.20.0
pyasn1 0.5.1
pyasn1-modules 0.3.0
PyYAML 6.0.1
requests 2.31.0
requests-oauthlib 1.4.0
rsa 4.9
scipy 1.10.1
setuptools 69.2.0
six 1.16.0
tensorboard 2.14.0
tensorboard-data-server 0.7.2
tensorflow 2.3.2
tensorflow-estimator 2.3.0
tensorflow-io-gcs-filesystem 0.31.0
termcolor 2.4.0
typing_extensions 4.5.0
urllib3 2.2.1
Werkzeug 3.0.1
wheel 0.43.0
wrapt 1.16.0
zipp 3.18.1

[KudaChan]

I am also having the same problem. The issue is with the Pythonnet Nuget package. version is mismatched. Don't know how to solve this.