Skip to content

Usage¤

Pretraining¤

Data Preparation¤

First install snakemake following these instructions

To download and regrid a CMIP6 dataset to a common resolution (e.g., 1.406525 degree), go to the corresponding directory inside snakemake_configs and run

snakemake all --configfile config_2m_temperature.yml --cores 8
This script will download and regrid the 2m_temperature data in parallel using 8 CPU cores. Modify configfile for other variables. After downloading and regrdding, run the following script to preprocess the .nc files into .npz format for pretraining ClimaX
python src/data_preprocessing/nc2np_equally_cmip6.py \
    --dataset mpi
    --path /data/CMIP6/MPI-ESM/1.40625deg/
    --num_shards 10
    --save_dir /data/CMIP6/MPI-ESM/1.40625deg_np_10shards
in which num_shards denotes the number of chunks to break each .nc file into.

Training¤

python src/climax/pretrain/train.py --config <path/to/config>
For example, to pretrain ClimaX on MPI-ESM dataset on 8 GPUs use
python src/climax/pretrain/train.py --config configs/pretrain_climax.yaml \
    --trainer.strategy=ddp --trainer.devices=8 \
    --trainer.max_epochs=100 \
    --data.batch_size=16 \
    --model.lr=5e-4 --model.beta_1="0.9" --model.beta_2="0.95" \
    --model.weight_decay=1e-5

Tip

Make sure to update the paths of the data directories in the config files (or override them via the CLI).

Pretrained checkpoints¤

We provide two pretrained checkpoints, one was pretrained on 5.625deg data, and the other was pretrained on 1.40625deg data. Both checkpoints were pretrained using all 5 CMIP6 datasets.

Usage: We can load the checkpoint by passing the checkpoint url to the training script. See below for examples.

Global Forecasting¤

Data Preparation¤

First, download ERA5 data from WeatherBench. The data directory should look like the following

5.625deg
   |-- 10m_u_component_of_wind
   |-- 10m_v_component_of_wind
   |-- 2m_temperature
   |-- constants.nc
   |-- geopotential
   |-- relative_humidity
   |-- specific_humidity
   |-- temperature
   |-- toa_incident_solar_radiation
   |-- total_precipitation
   |-- u_component_of_wind
   |-- v_component_of_wind

Then, preprocess the netcdf data into small numpy files and compute important statistics

python src/data_preprocessing/nc2np_equally_era5.py \
    --root_dir /mnt/data/5.625deg \
    --save_dir /mnt/data/5.625deg_npz \
    --start_train_year 1979 --start_val_year 2016 \
    --start_test_year 2017 --end_year 2019 --num_shards 8

The preprocessed data directory will look like the following

5.625deg_npz
   |-- train
   |-- val
   |-- test
   |-- normalize_mean.npz
   |-- normalize_std.npz
   |-- lat.npy
   |-- lon.npy

Training¤

To finetune ClimaX for global forecasting, use

python src/climax/global_forecast/train.py --config <path/to/config>
For example, to finetune ClimaX on 8 GPUs use
python src/climax/global_forecast/train.py --config configs/global_forecast_climax.yaml \
    --trainer.strategy=ddp --trainer.devices=8 \
    --trainer.max_epochs=50 \
    --data.root_dir=/mnt/data/5.625deg_npz \
    --data.predict_range=72 --data.out_variables=['z_500','t_850','t2m'] \
    --data.batch_size=16 \
    --model.pretrained_path='https://huggingface.co/tungnd/climax/resolve/main/5.625deg.ckpt' \
    --model.lr=5e-7 --model.beta_1="0.9" --model.beta_2="0.99" \
    --model.weight_decay=1e-5
To train ClimaX from scratch, set --model.pretrained_path="".

Regional Forecasting¤

Data Preparation¤

We use the same ERA5 data as in global forecasting and extract the regional data on the fly during training. If you have already downloaded and preprocessed the data, you do not have to do it again.

Training¤

To finetune ClimaX for regional forecasting, use

python src/climax/regional_forecast/train.py --config <path/to/config>
For example, to finetune ClimaX on North America using 8 GPUs, use
python src/climax/regional_forecast/train.py --config configs/regional_forecast_climax.yaml \
    --trainer.strategy=ddp --trainer.devices=8 \
    --trainer.max_epochs=50 \
    --data.root_dir=/mnt/data/5.625deg_npz \
    --data.region="NorthAmerica"
    --data.predict_range=72 --data.out_variables=['z_500','t_850','t2m'] \
    --data.batch_size=16 \
    --model.pretrained_path='https://huggingface.co/tungnd/climax/resolve/main/1.40625deg.ckpt' \
    --model.lr=5e-7 --model.beta_1="0.9" --model.beta_2="0.99" \
    --model.weight_decay=1e-5
To train ClimaX from scratch, set --model.pretrained_path="".

Climate Projection¤

Data Preparation¤

First, download ClimateBench data. ClimaX can work with either the original ClimateBench data or the regridded version. In the experiment in the paper, we regridded to ClimateBench data to 5.625 degree. To do that, run

python src/data_preprocessing/regrid_climatebench.py /mnt/data/climatebench/train_val \
    --save_path /mnt/data/climatebench/5.625deg/train_val --ddeg_out 5.625
and
python src/data_preprocessing/regrid_climatebench.py /mnt/data/climatebench/test \
    --save_path /mnt/data/climatebench/5.625deg/test --ddeg_out 5.625

Training¤

To finetune ClimaX for climate projection, use

python src/climax/climate_projection/train.py --config <path/to/config>
For example, to finetune ClimaX on 8 GPUs use
python python src/climax/climate_projection/train.py --config configs/climate_projection.yaml \
    --trainer.strategy=ddp --trainer.devices=8 \
    --trainer.max_epochs=50 \
    --data.root_dir=/mnt/data/climatebench/5.625deg \
    --data.out_variables="tas" \
    --data.batch_size=16 \
    --model.pretrained_path='https://huggingface.co/tungnd/climax/resolve/main/5.625deg.ckpt' \
    --model.out_vars="tas" \
    --model.lr=5e-4 --model.beta_1="0.9" --model.beta_2="0.99" \
    --model.weight_decay=1e-5
To train ClimaX from scratch, set --model.pretrained_path="".

Visualization¤

Coming soon