Spaces:
Runtime error
Apply for a GPU community grant: Personal project
I’m developing an open plant identification model that takes multiple photos, location, and time-of-year data to achieve significantly better accuracy than standard single image classifiers (like those currently on the App Store). The goal is to create a reliable tool that performs well across diverse environments and use cases. I’m also building LoRA adapters for specialised domains such as aquatic plants, weed detection, and region-specific flora, and releasing them all openly on Hugging Face.
I’ve built a prototype model trained on 2 million iNaturalist images across 14,000 species classes (example shown). It has already been downloaded over 400 times since its release last week. I’ve also published two open datasets (one iNaturalist focused and one broader GBIF based) which together have received nearly 3,000 downloads in the same period.
You can find all models and datasets on my Hugging Face profile.
A GPU grant would allow me to train the next-generation model at the required scale, evaluate multi-modal variants, and publish improved checkpoints, datasets, and demos for the community.