Agri Cm3 Vision Unsloth

Agri-CM3-Vision-Unsloth is an English vision-only dataset prepared for fine-tuning Vision Language Models using Unsloth. It is a reformatted subset of the original large-scale Chinese agricultural pest and disease benchmark HIT-Kwoo/Agri-CM3, created by farukalamai and last updated on 2026-06-30. The dataset extracts only the English vision splits, keeping all image-based tasks and formatting them in the ShareGPT conversation format compatible with Unsloth.

Use Cases

• Fine-tuning Vision Language Models for agricultural pest identification based on image-based tasks.
• Training models to generate descriptive text about plant diseases based on visual inputs.
• Benchmarking model performance on English-language agricultural vision tasks extracted from a larger Chinese benchmark.

Strengths

• Formatted specifically for compatibility with the Unsloth fine-tuning framework.
• Derived from a described large-scale benchmark for agricultural pest and disease.

Limitations

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

Provenance

Source

Subset of the HIT-Kwoo/Agri-CM3 benchmark, reformatted by farukalamai.

Collection Method

Extracted English vision splits from the original Chinese benchmark and reformatted.

Freshness

Last updated 2026-06-30 17:38:57; freshness should be verified.