luml.api.resources.collections

CollectionResource Objects

class CollectionResource(CollectionResourceBase, ListedResource)

get

def get(collection_value: str | None = None) -> CollectionDetails | None

Get collection by id or name.

Retrieves collection details by its id or name. Collection is related to default orbit. Search by name is case-sensitive and matches exact collection name.

Arguments:

• collection_value - The exact id or name of the collection to retrieve.

Returns:

Collection object.

Returns None if collection with the specified name or id is not found.

Raises:

• MultipleResourcesFoundError - If there are several collections with that name / id.

Example:

luml = LumlClient(api_key="luml_your_key")
collection_by_name = luml.collections.get("My Collection")
collection_by_id = luml.collections.get(
    "0199c455-21ee-74c6-b747-19a82f1a1e75"
)

Example response:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    name="My Collection",
    description="Dataset for ML models",
    collection_type='model',
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    tags=["ml", "training"],
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at=None
)

list_all

def list_all(*,
             limit: int | None = 100,
             sort_by: CollectionSortBy | None = None,
             order: SortOrder | None = SortOrder.DESC,
             search: str | None = None) -> Iterator[Collection]

List all orbit collections with auto-paging.

Arguments:

• limit - Page size (default: 100).
• sort_by - Field to sort by. Options: name, description, created_at.
• order - Sort order - "asc" or "desc" (default: "desc").
• search - Search string to filter collections by name or tags.

Returns:

Collection objects from all pages.

Example:

luml = LumlClient(
    api_key="luml_your_key",
    organization="0199c455-21ec-7c74-8efe-41470e29bae5",
    orbit="0199c455-21ed-7aba-9fe5-5231611220de"
)

for collection in luml.collections.list_all(limit=50):
    print(collection.id)

# Search by name or tags
for collection in luml.collections.list_all(search="model"):
    print(collection.name)

list

def list(*,
         start_after: str | None = None,
         limit: int | None = 100,
         sort_by: CollectionSortBy | None = None,
         order: SortOrder | None = SortOrder.DESC,
         search: str | None = None) -> CollectionsList

List all collections in the default orbit.

Arguments:

• start_after - Cursor to start listing from.
• limit - Maximum number of collections per page (default: 100).
• sort_by - Field to sort by. Options: name, description, created_at. If not provided, sorts by creation time.
• order - Sort order - "asc" or "desc" (default: "desc").
• search - Search string to filter collections by name or tags.

Returns:

CollectionsList object with items and cursor.

Example:

luml = LumlClient(api_key="luml_your_key")
result = luml.collections.list()
for collection in result.items:
    print(collection.name)

# Sort by name
result = luml.collections.list(
    sort_by=CollectionSortBy.NAME,
    order=SortOrder.ASC
)

# Search by name or tags
result = luml.collections.list(search="model")

Example response:

CollectionsList(
    items=[
        Collection(
            id="0199c455-21ee-74c6-b747-19a82f1a1e75",
            name="My Collection",
            description="Dataset for ML models",
            collection_type='model',
            orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
            tags=["ml", "training"],
            created_at='2025-01-15T10:30:00.123456Z',
            updated_at=None
        )
    ],
    cursor="WyIwMTliNDYxZmNmZDk3NTNhYjMwODJlMDUxZDkzZjVkZiIsICIyMDI1LTEyLTIyVDEyOjU0OjA4LjYwMTI5OCswMDowMCIsICJjcmVhdGVkX2F0Il0="
)

create

def create(description: str,
           name: str,
           collection_type: CollectionType,
           tags: builtins.list[str] | None = None) -> Collection

Create new collection in the default orbit.

Arguments:

• description - Description of the collection.
• name - Name of the collection.
• collection_type - Type of collection: "model", "dataset".
• tags - Optional list of tags for organizing collections.

Returns:

• Collection - Created collection object.

Example:

luml = LumlClient(api_key="luml_your_key")
collection = luml.collections.create(
    name="Training Dataset",
    description="Dataset for model training",
    collection_type=CollectionType.DATASET,
    tags=["ml", "training"]
)

Response object:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    name="Training Dataset",
    description="Dataset for model training",
    collection_type='model',
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    tags=["ml", "training"],
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at=None
)

update

@validate_collection
def update(name: str | None = None,
           description: str | None = None,
           tags: builtins.list[str] | None = None,
           *,
           collection_id: str | None = None) -> Collection

Update collection by ID or use default collection if collection_id not provided.

Updates the collection's data. Only provided parameters will be updated, others remain unchanged. If collection_id is None, the default collection from client will be used.

Arguments:

• name - New name for the collection.
• description - New description for the collection.
• tags - New list of tags.
• collection_id - ID of the collection to update. If not provided, uses the default collection set in the client.

Returns:

• Collection - Updated collection object.

Example:

luml = LumlClient(
    api_key="luml_your_key",
    organization="0199c455-21ec-7c74-8efe-41470e29bae5",
    orbit="0199c455-21ed-7aba-9fe5-5231611220de"
)
collection = luml.collections.update(
    collection_id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    name="Updated Dataset",
    tags=["ml", "updated"]
)

luml.collection = "0199c455-21ee-74c6-b747-19a82f1a1e75"
collection = luml.collections.update(
    name="Updated Dataset",
    description="Updated description"
)

Response object:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    description="Updated description",
    name="Updated Dataset",
    collection_type='model',
    tags=["ml", "updated"],
    total_models=43,
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at='2025-01-15T14:22:30.987654Z'
)

delete

@validate_collection
def delete(collection_id: str | None = None) -> None

Delete collection by ID or use default collection if collection_id not provided.

Permanently removes the collection and all its models. This action cannot be undone. If collection_id is None, the default collection from client will be used.

Arguments:

• collection_id - ID of the collection to delete. If not provided, uses the default collection set in the client.

Returns:

• None - No return value on successful deletion.

Example:

luml = LumlClient(
    api_key="luml_your_key",
    organization="0199c455-21ec-7c74-8efe-41470e29bae5",
    orbit="0199c455-21ed-7aba-9fe5-5231611220de"
)
# Delete specific collection by ID
luml.collections.delete("0199c455-21ee-74c6-b747-19a82f1a1e75")

# Set default collection
luml.collection = "0199c455-21ee-74c6-b747-19a82f1a1e75"
# Delete default collection (collection_id will be autofilled)
luml.collections.delete()

Warnings:

This operation is irreversible. All models, datasets, and data within the collection will be permanently lost. Consider backing up important data before deletion.

AsyncCollectionResource Objects

class AsyncCollectionResource(CollectionResourceBase, ListedResource)

get

async def get(collection_value: str | None = None) -> CollectionDetails | None

Get collection by id or name.

Retrieves collection details by its id or name. Collection is related to default orbit. Search by name is case-sensitive and matches exact collection name.

Arguments:

• collection_value - The exact id or name of the collection to retrieve.

Returns:

Collection object.

Returns None if collection with the specified name or id is not found.

Raises:

• MultipleResourcesFoundError - If there are several collections with that name / id.

Example:

luml = AsyncLumlClient(api_key="luml_your_key")
async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )
    collection_by_name = await luml.collections.get(
        "My Collection"
    )
    collection_by_id = await luml.collections.get(
        "0199c455-21ee-74c6-b747-19a82f1a1e75"
    )

Example response:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    name="My Collection",
    description="Dataset for ML models",
    collection_type='model',
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    tags=["ml", "training"],
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at=None
)

list_all

def list_all(*,
             limit: int | None = 100,
             sort_by: CollectionSortBy | None = None,
             order: SortOrder | None = SortOrder.DESC,
             search: str | None = None) -> AsyncIterator[Collection]

List all orbit collections with auto-paging.

Arguments:

• limit - Page size (default: 100).
• sort_by - Field to sort by. Options: name, description, created_at.
• order - Sort order - "asc" or "desc" (default: "desc").
• search - Search string to filter collections by name or tags.

Returns:

Collection objects from all pages.

Example:

luml = AsyncLumlClient(
    api_key="luml_your_key",
)

async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )

    async for collection in luml.collections.list_all(limit=50):
        print(collection.id)

    # Search by name or tags
    async for collection in luml.collections.list_all(search="model"):
        print(collection.name)

list

async def list(*,
               start_after: str | None = None,
               limit: int | None = 100,
               sort_by: CollectionSortBy | None = None,
               order: SortOrder | None = SortOrder.DESC,
               search: str | None = None) -> CollectionsList

List all collections in the default orbit.

Arguments:

• start_after - Cursor to start listing from.
• limit - Maximum number of collections per page (default: 100).
• sort_by - Field to sort by. Options: name, description, created_at. If not provided, sorts by creation time.
• order - Sort order - "asc" or "desc" (default: "desc").
• search - Search string to filter collections by name or tags.

Returns:

CollectionsList object with items and cursor.

Example:

luml = AsyncLumlClient(api_key="luml_your_key")
async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )
    result = await luml.collections.list()
    for collection in result.items:
        print(collection.name)

    # Sort by name
    result = await luml.collections.list(
        sort_by=CollectionSortBy.NAME,
        order=SortOrder.ASC
    )

    # Search by name or tags
    result = await luml.collections.list(search="model")

Example response:

CollectionsList(
    items=[
        Collection(
            id="0199c455-21ee-74c6-b747-19a82f1a1e75",
            name="My Collection",
            description="Dataset for ML models",
            collection_type='model',
            orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
            tags=["ml", "training"],
            created_at='2025-01-15T10:30:00.123456Z',
            updated_at=None
        )
    ],
    cursor="WyIwMTliNDYxZmNmZDk3NTNhYjMwODJlMDUxZDkzZjVkZiIsICIyMDI1LTEyLTIyVDEyOjU0OjA4LjYwMTI5OCswMDowMCIsICJjcmVhdGVkX2F0Il0="
)

create

async def create(description: str,
                 name: str,
                 collection_type: CollectionType,
                 tags: builtins.list[str] | None = None) -> Collection

Create new collection in the default orbit.

Arguments:

• description - Description of the collection.
• name - Name of the collection.
• collection_type - Type of collection: "model", "dataset".
• tags - Optional list of tags for organizing collections.

Returns:

• Collection - Created collection object.

Example:

luml = AsyncLumlClient(api_key="luml_your_key")

async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )
    collection = await luml.collections.create(
        name="Training Dataset",
        description="Dataset for model training",
        collection_type=CollectionType.DATASET,
        tags=["ml", "training"]
    )

Response object:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    name="Training Dataset",
    description="Dataset for model training",
    collection_type='model',
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    tags=["ml", "training"],
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at=None
)

update

@validate_collection
async def update(name: str | None = None,
                 description: str | None = None,
                 tags: builtins.list[str] | None = None,
                 *,
                 collection_id: str | None = None) -> Collection

Update collection by ID or use default collection if collection_id not provided.

Updates the collection's data. Only provided parameters will be updated, others remain unchanged. If collection_id is None, the default collection from client will be used.

Arguments:

• name - New name for the collection.
• description - New description for the collection.
• tags - New list of tags.
• collection_id - ID of the collection to update. If not provided, uses the default collection set in the client.

Returns:

• Collection - Updated collection object.

Example:

luml = AsyncLumlClient(
    api_key="luml_your_key",
)

async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )
    collection = await luml.collections.update(
        collection_id="0199c455-21ee-74c6-b747-19a82f1a1e75",
        name="Updated Dataset",
        tags=["ml", "updated"]
    )

    luml.collection = "0199c455-21ee-74c6-b747-19a82f1a1e75"
    collection = await luml.collections.update(
        name="Updated Dataset",
        description="Updated description"
    )

Response object:

Collection(
    id="0199c455-21ee-74c6-b747-19a82f1a1e75",
    orbit_id="0199c455-21ed-7aba-9fe5-5231611220de",
    description="Updated description",
    name="Updated Dataset",
    collection_type='model',
    tags=["ml", "updated"],
    total_models=43,
    created_at='2025-01-15T10:30:00.123456Z',
    updated_at='2025-01-15T14:22:30.987654Z'
)

delete

@validate_collection
async def delete(collection_id: str | None = None) -> None

Delete collection by ID or use default collection if collection_id not provided.

Permanently removes the collection and all its models. This action cannot be undone. If collection_id is None, the default collection from client will be used.

Arguments:

• collection_id - ID of the collection to delete. If not provided, uses the default collection set in the client.

Returns:

• None - No return value on successful deletion.

Example:

luml = AsyncLumlClient(
    api_key="luml_your_key",
)

async def main():
    await luml.setup_config(
        organization="0199c455-21ec-7c74-8efe-41470e29bae5",
        orbit="0199c455-21ed-7aba-9fe5-5231611220de",
    )

    # Delete specific collection by ID
    await luml.collections.delete(
        "0199c455-21ee-74c6-b747-19a82f1a1e75"
    )

    # Set default collection
    luml.collection = "0199c455-21ee-74c6-b747-19a82f1a1e56"
    # Delete default collection
    await luml.collections.delete()

Warnings:

This operation is irreversible. All models, datasets, and data within the collection will be permanently lost. Consider backing up important data before deletion.