Skip to content

Code for training a fine-grained species classification model using data from GBIF

License

Notifications You must be signed in to change notification settings

RolnickLab/gbif-species-trainer

Repository files navigation

*** NOTE: This code is deprecated and being slowly merged to ami-ml repoistory. ***

Species model trainer from GBIF

This repository contains the code to download image and metadata for a list of species from Global Biodiversity Information Facility (GBIF) and train a deep learning model using the downloaded data.

Setup python environment

Install conda and prepare a python environment using the following steps:

Build a new conda environment

conda create -n gbif_species_trainer python=3.9

Activate the conda environment:

conda activate gbif_species_trainer

Install cuda toolkit and pytorch

conda install pytorch torchvision torchaudio pytorch-cuda=11.7 -c pytorch -c nvidia

Install additional libraries using pip:

python3 -m pip install -r requirements.txt

Download Data

Follow instructions in the data_download folder to download data for a list of species.

Train Model

Train a model using scripts in model_training folder.

About

Code for training a fine-grained species classification model using data from GBIF

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages