Log into your account to keep track of your progress. You can work through the lessons without logging in but your progress will be lost when you refresh the page.

Have you ever thought about combining Houdini's procedural power with state of the art AI Algorithms without having to care about boilerplate scripts or neverending downloads? Look no further as Paul Ambrosiussen and Moritz Schwind cover their strategies and practice of doing exactly this with a modular Machine Learning Toolkit they developed for Houdini, allowing you to natively use Houdini's Data to drive, customize and process various AI Models.

You can download the plug-in using the link on this page. You can also join the Discord server to work with Paul and Moritz to expand and enhance this toolset.

Learn how to work with MLOPS, the free and Open Source Machine Learning Plugin for Houdini developed by Paul Ambrosiussen and Entagma, Licensed and Distributed by Bismuth Consultancy BV.



Still thinks “Space: 1999” is the coolest thing that ever happened on german TV. Be it pixels, hardware, code or cameras – if it’s interesting, Moritz is gonna take it apart. And sometimes even reassemble it. In his spare time he likes to dabble with code and create generative artwork. He claims his early exposure to QBasic is no help at all when working in Houdini, Cinema 4D, Processing or Arduino. But it might have been what started his fascination for the boundaries of code and art. When not wreaking havoc to any intriguing devices around him, he works as a freelance Art Director / Technical Director.

More from Moritz Schwind


  • Enivob 1 year ago  | 

    Thanks for the toolkit, it's a lot of fun. These tutorials got me up and running without any install issues.

  • GSO 1 year ago  | 


  • HubertPokrandt 11 months, 2 weeks ago  | 

    Hello ,

    Is there possibility to use MLOPs on AMD Radeon 6800xt ?

  • kamyab 11 months ago  | 


    Is it possible to do a tutorial for MLops installation on MacOS

  • Renzell 10 months, 2 weeks ago  | 

    I always wanted to make Art from my Videogame Screenshots, and MLOPS did that easy. I'm here for it and ill assist in any way I can to aid in the downfall of everything.

  • LukeP 10 months, 1 week ago  | 

    Where/how can we report issues with MLOPs?

  • sfalan 9 months, 1 week ago  | 

    When I try to install the dependencies, it pops up an error window essentially telling me that I don't have git installed.
    So, like LukeP suggested, where should we report issues?

    • JoeyBryarsFX 2 months ago  | 

      I had the same issue. Here is a link to the git installation guide:
      Once you have git installed the MLOP dependencies installation should run smoothly. I tested it on my machine, and everything ran smoothly after I installed git on my PC.

  • MAHMOUD ELRAGHEB 9 months ago  | 

    Hey guys, I have already a folder with tons of models and Loras, I share it between a1111 and comfyUI, I just need to change the yaml file to point it to the directory.
    But I'm not sure how to edit it to work with Houdini.

    For instance I have this path "A:\a1111\stable-diffusion-webui\models\Stable-diffusion\v1-5-pruned-emaonly.safetensors"

    but it's giving me an error,

    Invalid source /obj/geo1/sd_tokenizer1/tokenizer
    Error: Error generated by Python node.
    The specified model does not exist locally. Because the 'Local Cache' checkbox for this node is enabled, it also wont be downloaded automatically. Specify a valid local model or disable the 'Local Cache' parameter. a

    I want to be able to use the folders that I have already with models, loras and controlNet.
    Is this possible?

    "MLOPS_MODELS": "A:/a1111/stable-diffusion-webui/models/Stable-diffusion"
    "MLOPS_SD_MODEL": "A:/a1111/stable-diffusion-webui/models/Stable-diffusion/v1-5-pruned-emaonly.safetensors"

  • JIn_ 7 months, 1 week ago  | 

    hello, thanks for this amazing tool!!
    en..I have one question, is this tool working for headless houdini pipeline? like use python to load everything without GUI?

  • Beyon 5 months, 2 weeks ago  | 

    Ive got an installation issue, maybe someone here has resolved this :)

    `Traceback (most recent call last):
    File "mlops_install", line 5, in
    File "/Users/*****/Documents/MLOPs/scripts/python/mlops_utils.py", line 126, in install_mlops_dependencies
    pip_install(hou.text.expandString("$MLOPS/requirements.txt"), True, True)
    File "/Users/****/Documents/MLOPs/scripts/python/mlops_utils.py", line 364, in pip_install
    raise hou.Error(res[1].decode())
    hou.Error: error: subprocess-exited-with-error

    × python setup.py egg_info did not run successfully.
    │ exit code: 1
    ╰─> [10 lines of output]
    /Applications/Houdini/Houdini19.5.752/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/_distutils_hack/init.py:17: UserWarning: Distutils was imported before Setuptools, but importing Setuptools also replaces the distutils module in sys.modules. This may lead to undesirable behaviors or errors. To avoid these issues, avoid using distutils directly, ensure that setuptools is installed in the traditional way (e.g. not an editable install), and/or make sure that setuptools is always imported before distutils.
    /Applications/Houdini/Houdini19.5.752/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/_distutils_hack/init.py:30: UserWarning: Setuptools is replacing distutils.
    warnings.warn("Setuptools is replacing distutils.")
    Traceback (most recent call last):
    File "", line 2, in

    File "<pip-setuptools-caller>", line 34, in <module>
    File "/private/var/folders/6n/9xz_rjz946b80_8s2gw2jwkw0000gn/T/pip-install-k2hyvmzz/xformers_67ce5be12fa24c4786a46aadfad7bd4d/setup.py", line 23, in <module>
    import torch
    ModuleNotFoundError: No module named 'torch'
    [end of output]
    note: This error originates from a subprocess, and is likely not a problem with pip.
    error: metadata-generation-failed

    File "<pip-setuptools-caller>", line 34, in <module>
    File "/private/var/folders/6n/9xz_rjz946b80_8s2gw2jwkw0000gn/T/pip-install-k2hyvmzz/xformers_67ce5be12fa24c4786a46aadfad7bd4d/setup.py", line 23, in <module>
    import torch
    ModuleNotFoundError: No module named 'torch'
    [end of output]
    note: This error originates from a subprocess, and is likely not a problem with pip.
    error: metadata-generation-failed

  • Slavomir96 2 months ago  | 

    Guys I have a problem with sd_solver. When I put my prompt setup into sd_solver between camera to points, appear this problem

    Invalid source /obj/geo1/sd_solver1/solver
    Error: Python error: Traceback (most recent call last):
    File "<stdin>", line 52, in <module>
    File "C:\MLOPS_2.0_H19.5/MLOPs-2.0/scripts/python\sdpipeline\image_solve.py", line 121, in run
    controlnetmodel = point.stringAttribValue("model")
    File "C:\PROGRA~1/SIDEEF~1/HOUDIN~1.752/houdini/python3.9libs\hou.py", line 70677, in stringAttribValue
    return _hou.Point_stringAttribValue(self, *args)
    hou.OperationFailed: The attempted operation failed.
    No attribute with this name exists

Please log in to leave a comment.