Reproduce¶
To reproduce the results of a project, we clone the code and then retrieve the data managed with DVC:
$ git clone https://github.com/veit/dvc-example.git
$ cd dvc-example
$ uv sync
$ uv run dvc pull -TR
Collecting |23.0 [00:02, 7.77entry/s]
Fetching
Building workspace index |1.00 [00:00, 119entry/s]
Comparing indexes |25.0 [00:00, 1.92kentry/s]
Applying changes |14.0 [00:00, 4.10kfile/s]
A data/features/
A data/prepared/
A eval/
A data/data.xml
A model.pkl
17 files fetched and 14 files added
$ tree data
data
├── data.xml
├── data.xml.dvc
├── features
│ ├── test.pkl
│ └── train.pkl
└── prepared
├── test.tsv
└── train.tsv
You can then easily reproduce the results with dvc repro:
$ uv run dvc repro
'data/data.xml.dvc' didn't change, skipping
Stage 'prepare' didn't change, skipping
Stage 'featurize' didn't change, skipping
Stage 'train' didn't change, skipping
Running stage 'evaluate':
> uv run python src/dvc_example/evaluate.py model.pkl data/features
Use `dvc exp run` to save experiment.