lamindb.core.Settings#
- class lamindb.core.Settings(git_repo)#
Bases:
object
Settings.
Use
lamindb.settings
instead of instantiating this class yourself.Attributes
- artifact_use_virtual_keys bool#
Treat
key
parameter inArtifact
as virtual.If
True
, thekey
is not used to construct file paths, but file paths are based on theuid
of artifact.
- silence_file_run_transform_warning bool#
Silence warning about missing run & transform during file creation.
- storage Path | UPath#
Default storage location (a path to its root).
Examples:
You can set the root via:
>>> ln.settings.storage = "s3://some-bucket"
You can also pass additional fsspec kwargs via:
>>> kwargs = dict( >>> profile="some_profile", # fsspec arg >>> cache_regions=True # fsspec arg for s3 >>> ) >>> ln.settings.storage = "s3://some-bucket", kwargs
- sync_git_repo str | None#
Sync transforms with scripts in git repository.
Provide the full git repo URL.
- track_run_inputs bool#
Track files as input upon
.load()
,.cache()
and.backed()
.Requires a global run context with
track()
was created!
- transform TransformSettings#
Transform settings.
- upon_artifact_create_if_hash_exists str#
Behavior if file hash exists (default
"warn_return_existing"
).One of
["warn_return_existing", "error", "warn_create_new"]
.FAQ: What happens if I save the same artifacts & records twice?
- upon_create_search_names bool#
To speed up creating Registry objects (default
True
).If
True
, search for alternative names.FAQ: What happens if I save the same artifacts & records twice?
- upon_file_create_skip_size_hash bool#
To speed up registering high numbers of files (default
False
).This bypasses queries for size and hash to AWS & GCP.
It speeds up file creation by about a factor 100.
- verbosity str#
Logger verbosity (default ‘warning’).
‘error’: ❌ only show error messages
‘warning’: ❗ also show warning messages
‘success’: ✅ also show success and save messages
‘info’: 💡 also show info messages
‘hint’: 💡 also show hint messages
‘debug’: 🐛 also show detailed debug messages
This is based on Scanpy’s and Django’s verbosity setting.