ChemVLM SFT Datasets: Multimodal Chemistry Training Data

SFT Datasets of ChemVLM, part of the MMChem Series. The dataset includes both image bytes and conversations for training multimodal large language models in chemistry. It was last updated on October 29, 2025, by the author di-zhang-fdu.

Use Cases

• Instruction-tuning multimodal LLMs for chemistry based on the provided image-text conversations.
• Benchmarking model performance on chemistry tasks using the structured training subsets.
• Developing visual question-answering systems for chemical structures and diagrams based on the multimodal content.

Strengths

• Data is structured into subsets for focused training or evaluation.
• Dataset format is specified as Apache Arrow, facilitating efficient loading with the `datasets` library.

Limitations

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

Provenance

Source

di-zhang-fdu on Hugging Face.

Collection Method

Likely compiled for training the ChemVLM model.

Freshness

Last updated 2025-10-29 03:59:47; freshness should be verified.