dgovil / mldeform Goto Github PK

Hi @dgovil,

I tried your Cylinder_setup scene in Maya 2022.
Just as you mentioned I wrote the data and trained it. I used Collab for training.
Then I loaded and created a deformer and set up the joints.
I am connecting joints in a similar way you are doing in test_deformer function
I can see the vertices are moved but they are moving in the wrong way.

I am attaching herewith the training input and output data that I am loading in the deformer.
Test1.zip

Kindly advise if this is expected? or what/if I have missed.

I didn't see the training data in the repo, can I have the training data?

maya python version is 2.7.11
but pandas need numpy 1.3 or above...
tensorflow need python 3.5 or above..
so, i can't use maya running ‘’from MLDeform import train ‘’
happen a error...
because maya can't import pandas and tensorflow

I got one joint to write out training data. If it's multiple joints than I get this error.

Train on 0 samples, validate on 1 samples
Epoch 1/200
# Error: Empty training data.
# Traceback (most recent call last):
#   File "<maya console>", line 3, in <module>
#   File "C:/Users/Edward Conry/Documents/maya/2022/scripts\MLDeform2\_training\train.py", line 179, in train
#     batch_size=batch_size)
#   File "C:\Program Files\Autodesk\Maya2022\Python37\lib\site-packages\tensorflow\python\keras\engine\training.py", line 780, in fit
#     steps_name='steps_per_epoch')
#   File "C:\Program Files\Autodesk\Maya2022\Python37\lib\site-packages\tensorflow\python\keras\engine\training_arrays.py", line 380, in model_iteration
#     aggregator.finalize()
#   File "C:\Program Files\Autodesk\Maya2022\Python37\lib\site-packages\tensorflow\python\keras\engine\training_utils.py", line 111, in finalize
#     raise ValueError('Empty training data.')
# ValueError: Empty training data. # 

Hi Dhruv, I am trying to run your repo but kind of stuck at bringing TensorFlow to Maya.

I first tried to install TensorFlow using mayapy, but using that I was getting attached error

It complains when imp.load_module('_pywrap_tensorflow_internal' but _pywrap_tensorflow_internal is present.

Anyway, I followed your advice and install tf2 using anaconda, but now I am not sure how to add anaconda library path to Maya library paths. I am guessing it should be conda/pkgs that I need to reference in Maya, but not sure where. Kindly advice.

OS: Linux/CentOS7
Maya Version: 2020