correct some typos of DERIVED_DATA in readme

README.md

@@ -41,7 +41,7 @@ Recommended to run with a sample of the dataset.
 Train models through `bin/main.py`, e.g. to train the model used in the paper use
 
 ```sh
-python bin/main.py --config src/ml_downscaling_emulator/score_sde_pytorch/configs/subvpsde/ukcp_local_pr_12em_cncsnpp_continuous.py --workdir ${DERVIED_DATA}/path/to/models/paper-12em --mode train
+python bin/main.py --config src/ml_downscaling_emulator/score_sde_pytorch/configs/subvpsde/ukcp_local_pr_12em_cncsnpp_continuous.py --workdir ${DERIVED_DATA}/path/to/models/paper-12em --mode train
 ```
 
 ```sh
@@ -74,12 +74,12 @@ Functionalities can be configured through config files, or more conveniently, th
 Once have trained a model create samples from it with `bin/predict.py`, e.g.
 
 ```sh
-python bin/predict.py --checkpoint epoch_20 --dataset bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season --split test  --ensemble-member 01 --input-transform-dataset bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season --input-transform-key pixelmmsstan --num-samples 1 ${DERVIED_DATA}/path/to/models/paper-12em
+python bin/predict.py --checkpoint epoch_20 --dataset bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season --split test  --ensemble-member 01 --input-transform-dataset bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season --input-transform-key pixelmmsstan --num-samples 1 ${DERIVED_DATA}/path/to/models/paper-12em
 ```
 
 This example command will:
-* use the checkpoint of the model in `${DERVIED_DATA}/path/to/models/paper-12em/checkpoints/{checkpoint}.pth` and model config from training `${DERVIED_DATA}/path/to/models/paper-12em/config.yml`.
-* store samples generated in `${DERVIED_DATA}/path/to/models/paper-12em/samples/{dataset}/{input_transform_data}-{input_transform_key}/{split}/{ensemble_member}/`. Sample files ar named like `predictions-{uuid}.nc`.
+* use the checkpoint of the model in `${DERIVED_DATA}/path/to/models/paper-12em/checkpoints/{checkpoint}.pth` and model config from training `${DERIVED_DATA}/path/to/models/paper-12em/config.yml`.
+* store samples generated in `${DERIVED_DATA}/path/to/models/paper-12em/samples/{dataset}/{input_transform_data}-{input_transform_key}/{split}/{ensemble_member}/`. Sample files ar named like `predictions-{uuid}.nc`.
 * generate samples conditioned on examples from ensemble member `01` in the `test` subset of the `bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season` dataset.
 * transform the inputs based on the `bham_60km-4x_12em_psl-sphum4th-temp4th-vort4th_eqvt_random-season` dataset using the `pixelmmsstan` approach.
 * generate 1 set of samples.