FM4M-demo1

FM4M-demo1

A Gradio demo from IBM Research that names three IBM materials models, smi-ted, selfies-ted and mhg-ged. For researchers and developers who want to look at how the demo is built.

Made by ibm-research on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This is a Gradio demo from IBM Research that names three IBM materials models: smi-ted, selfies-ted and mhg-ged. The README is empty, so the page gives no description. Folder names such as models, representation and plot_emb suggest it builds numeric representations of molecules and plots them. It runs on the basic CPU tier at Hugging Face.

Who it suits: Researchers in materials and chemistry machine learning who want to look at IBM's models in a demo, and developers who want to read how it is organized.

Industry | Who uses it | How they use it | Example
Research | Materials ML researchers | Try the three named models in one demo | Open the app and look at what it offers
Chemistry | Cheminformatics developers | Read the representation code | Compare how the models are loaded
Software | Developers | Study a larger Gradio app with models and data folders | Read app.py and the repository layout

Skip it if you want a documented tool, since the page says little about what the demo does.

Where to use it:

• A web browser, on the Space's page at Hugging Face
• Windows, macOS or Linux with Python, as a local copy
• Docker, since the repository holds Dockerfile files, though the page gives no Docker commands

Use case | Who | What they get back
Explore the models | Researchers | A demo page built around the three named models
Study the code | Developers | A repository with model, data and plotting folders

How to set it up:

• Nothing at all if you only use the online Space
• Python and pip, for a local copy
• The Python packages listed in requirements.txt

1. Open the Space's page at Hugging Face to use it online.
2. For a local copy, install the Hugging Face hub package and download the Space.
3. Move into the new folder and install Gradio at the pinned version.
4. Install the packages from requirements.txt.
5. Start the app with Python.

pip install -U huggingface_hub
hf download ibm-research/FM4M-demo1 --repo-type space --local-dir FM4M-demo1
cd FM4M-demo1
pip install gradio==5.4.0
pip install -r requirements.txt
python app.py

The repository is large, at 382 files, and has two Dockerfiles. The page does not say what a local run needs beyond the packages.

How to use it: Online, open the Space and use the form on its page. On your own machine, run the app file and open the page Gradio serves in your browser. The page lists no other commands.

Tool | What it does
python app.py | Starts the Gradio app on your computer

Example: You open the Space and look over the options it shows for the three models. Because the page has no description, you try one option at a time and read the output, then check the model pages for what it means.

For newcomers and experts: If you are new, open the Space online and click through it slowly. The page has no guide, so expect to learn by trying.

If you know the field, the three model names point to molecular representations, with names that suggest SMILES, SELFIES and graph inputs. The page gives no inputs, metrics or tasks, so read the model pages first.

Open the Space in a browser
Choose one of the named models
The demo builds a representation
The result is shown or plotted

Good to know:

• License: apache-2
• Cost to run: none named, and the Space sits on the basic CPU tier at Hugging Face
• Data it reaches: the data folder in the repository, and the model files from Hugging Face
• Not for: anything that needs documented behavior, since the page has no description

On Hugging Face, 13 November 2026

• Likes: 9
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/ibm-research/FM4M-demo1
• Files: https://huggingface.co/spaces/ibm-research/FM4M-demo1/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.