SMolInstruct: Large-Scale Instruction Tuning Dataset for Chemistry Tasks

SMolInstruct is a large-scale instruction tuning dataset for chemistry tasks centered around small molecules. It contains over 3 million samples across 14 distinct chemistry tasks. The dataset was created by osunlp and last updated on Hugging Face in September 2024.

Use Cases

• Fine-tuning language models for chemistry question answering based on the described instruction tasks.
• Training models to generate or reason about small molecule properties and structures.
• Benchmarking model performance on a variety of specialized chemistry tasks.
• Developing AI assistants for chemistry education or research based on the instruction-response pairs.

Strengths

• Contains over 3 million samples, indicating a large scale.
• Covers 14 distinct chemistry tasks, suggesting breadth.
• Designed with a focus on small molecules, providing a specific domain.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect source or task selection bias inherent to its construction.

Provenance

Source

osunlp

Collection Method

Likely compiled for instruction tuning of language models.

Time Range

null

Freshness

Last updated 2024-09-18 15:43:27.

Geography

null