Datasets and Models for Discovery of Inorganic Nonlinear Optical Crystals

Materials Project data extracts for training generative models and classifiers focused on nonlinear optical materials. The 952.3 MB collection includes trained models, generated virtual structures, and first-principles calculation results, created by Anchor Yu and last updated in May 2026.

Use Cases

• Training generative models for crystal structure discovery based on data from the Materials Project.
• Developing machine learning classifiers for nonlinear optical materials based on labeled datasets.
• Performing first-principles calculations on candidate compounds generated by the trained models.
• Benchmarking generative model performance for inorganic materials design.

Strengths

• Dataset size is 952.3 MB, indicating a substantial collection of models and data.
• Includes multiple file formats (CKPT, YAML, CSV, PT, ZIP) supporting different stages of the ML workflow.
• Data is sourced from the authoritative Materials Project database.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Materials Project (MP)

Collection Method

Extracted for training machine learning models.

Freshness

Last updated 2026-05-15 08:08:27