Create, inspect, and sync local datasets.
Local datasets are directories with a dataset.yaml under configured roots
(~/.biolm/datasets, ./.biolm/datasets). Use push / pull with
--backend mlflow to sync via the optional MLflow plugin.
Usage
biolm dataset [OPTIONS] COMMAND [ARGS]...
biolm dataset add
Copy files into a local dataset.
Examples:
Usage
biolm dataset add [OPTIONS] ID_OR_PATH FILES...
Options
- -r, --recursive
Copy directories recursively
Arguments
- ID_OR_PATH
Required argument
- FILES
Required argument(s)
biolm dataset create
Create a new local dataset under the primary root.
Examples:
Usage
biolm dataset create [OPTIONS] DATASET_ID
Options
- --type <dtype>
Dataset type label (default: files)
- --tag <tag>
Tag (repeatable)
- --description <description>
Human-readable description
- --root <root>
Root directory for the new dataset
- --force
Overwrite existing dataset.yaml
Arguments
- DATASET_ID
Required argument
biolm dataset init
Adopt an existing directory as a dataset (writes dataset.yaml).
Examples:
Usage
biolm dataset init [OPTIONS] PATH
Options
- --id <dataset_id>
Required Stable dataset id
- --type <dtype>
Dataset type label (default: files)
- --tag <tag>
Tag (repeatable)
- --description <description>
Human-readable description
- --force
Overwrite existing dataset.yaml
Arguments
- PATH
Required argument
biolm dataset list
List local datasets under configured discovery roots.
Examples:
Usage
biolm dataset list [OPTIONS]
Options
- --type <dtype>
Filter by dataset type
- --tag <tag>
Filter by tag
- --format <fmt>
Output format
- Options:
table | json
- -o, --output <output>
Save output to file
biolm dataset pull
Pull a remote dataset into a local dataset directory.
Defaults to ~/.biolm/datasets/.
Usage
biolm dataset pull [OPTIONS] DATASET_ID
Options
- --backend <backend>
Required Remote backend (e.g. mlflow)
- --path <dest_path>
Local destination directory
- --force
Overwrite conflicting local dataset.yaml
- --mlflow-uri <mlflow_uri>
MLflow tracking URI
- --experiment <experiment>
MLflow experiment name
Arguments
- DATASET_ID
Required argument
biolm dataset push
Push a local dataset to a remote backend.
Examples:
Usage
biolm dataset push [OPTIONS] ID_OR_PATH
Options
- --backend <backend>
Required Remote backend (e.g. mlflow)
- --mlflow-uri <mlflow_uri>
MLflow tracking URI
- --experiment <experiment>
MLflow experiment name
Arguments
- ID_OR_PATH
Required argument
biolm dataset show
Show metadata and files for a local dataset by id or path.
Examples:
Usage
biolm dataset show [OPTIONS] ID_OR_PATH
Options
- --format <fmt>
Output format
- Options:
table | json
- -o, --output <output>
Save output to file
Arguments
- ID_OR_PATH
Required argument