Automated crystal identification system developed at EMBL Grenoble Marquez group by Aurélien Personnaz.
Create a python virtual environment and install the dependencies.
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
The models are pulled and stored in the HuggingFace repository
You will need to sign up and get a write token.
If you want to log the training in wandb you need to login first:
wandb login
The wandb entity and project name can be passed by setting the environment variables WANDB_ENTITY and WANDB_PROJECT
export WANDB_ENTITY=*your entity*
export WANDB_PROJECT=*your project*
The HuggingFace token must be passed with the environment variable HUGGINGFACE_TOKEN
export HUGGINGFACE_TOKEN=*your token*
Then run a training with the training.py script.
Example:
python AXIS_training.py --n_epoch 2 --train_dataset ~/data/CRIMS-v1 --test_dataset ~/data/CRIMS-test
The HuggingFace token must be passed with the environment variable HUGGINGFACE_TOKEN
export HUGGINGFACE_TOKEN=*your token*
You can run a simple inference by executing
python AXIS_inference.py --model_checkpoint apersonnaz/AXIS-CRIMS_v3-vis --image_path ~/data/CRIMS-test/vis/other/1000.jpg