Hands-on walkthroughs of PRISM on a scientific application, grouped by application area. Each one walks a single configuration from environment and data through training.
Tutorial pages live here, in docs/tutorials/<area>/. Notebooks live in
applications/<area>/, next to the scripts they call.
For reference material on the same applications, such as pipeline designs and
evaluation invocations, see docs/applications/.
| Area | Tutorial | Modalities | Platform |
|---|---|---|---|
| Biology | BioReason: train a DNA + language model (notebook): projector alignment, SFT, GRPO, and KEGG evaluation, submitted as chained PBS jobs | DNA + text | Aurora |
| Materials science | Jointly train on material descriptions and crystal structures: predict a material property from a CIF crystal graph and a text description | Crystal graph + text | CUDA, XPU, or CPU |
- Put Markdown pages in
docs/tutorials/<area>/, where<area>is the application's directory underapplications/, or the name that directory would take if the application has no scripts there yet. - Put a notebook in
applications/<area>/, next to the scripts it runs, and link to it from this page. - Add a row to the table above and to the Tutorials section of
docs/index.md. - Run
make docsbefore pushing. It checks relative links in Markdown files the same way CI does. Links inside notebooks are not checked, so test those by hand after moving a notebook.