Backend REST API#

POST /auth/#

Auth User

Authenticate user and return access and refresh tokens

Args:

response (Response): Response that will be returned to the client form_data (OAuth2PasswordRequestForm, optional): Username and Password.

Defaults to Depends().

db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB Connection.: Defaults to Depends(get_db).

Raises:

HTTPException: 401 Unauthorized if credentials are invalid HTTPException: 404 Not Found if user does not exist

Returns:

Dict[str, str]: Response containing access and refresh tokens. Credentials are saved as cookies.

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "access_token": "string",
    "refresh_token": "string",
    "token_type": "bearer"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /auth/refresh#

Get Refresh Token

Use refresh token to get new access token when access token expires.

Args:

request (Request): Incoming request response (Response): Response that will be returned to the client db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB Connection.

Defaults to Depends(get_db).

csrf (CsrfProtect, optional): Perform CSRF protection. Defaults to Depends().

Raises:

HTTPException: 404 if user does not exist HTTPException: 403 if refresh token expired CREDENTIALS_EXCEPTION: 401 if credentials are invalid HTTPException: 409 if CSRF token is invalid HTTPException: 400 otherwise

Returns:

Optional[Dict[str, str]]: Access token if refresh token is valid

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "access_token": "string",
    "refresh_token": "string",
    "token_type": "bearer"
}

DELETE /auth/logout#

Logout User

Logout user by deleting cookies

Args:: response (Response): Response that will be returned to the client
Raises:: HTTPException: 500 if an error occurs

Status Codes:

204 No Content – Successful Response

GET /auth/is_admin#

Verify Admin

This endpoint mostly exists just to test that auth code is working

Example request:

GET /auth/is_admin HTTP/1.1
Host: example.com

Status Codes:

204 No Content – Successful Response

PUT /buckets/video#

Replace Video

Replaces a video in the MinIO bucket

Args:

new_video (UploadFile, optional): New video. Defaults to Form(). old_video_location (Optional[str], optional): Location of video to replace.

Defaults to Form(None).

s3_client (Minio, optional): Minio client. Defaults to Depends(minio_api_client).

Raises:

HTTPException: 500 if something went wrong

Returns:

Dict[str, str]: Location of the new video in the bucket

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "video_location": "string"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /buckets/video#

Upload Video

Uploads a video to the MinIO bucket

Args:: video (UploadFile): Video file to upload s3_client (Minio, optional): Minio client. Defaults to Depends(minio_api_client).
Raises:: HTTPException: 500 if something went wrong
Returns:: Dict[str, str]: Location of the video in the bucket

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "video_location": "string"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /models/_db/options/filters/#

Get Available Filters

Get available filters for model zoo search page

Args:

db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.: Defaults to Depends(get_db).

Returns:

Dict[str, List[str]]: All available tags, frameworks, and tasks

Example request:

GET /models/_db/options/filters/ HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "tags": [
        "string"
    ],
    "frameworks": [
        "string"
    ],
    "tasks": [
        "string"
    ]
}

GET /models/{creator_user_id}/{model_id}#

Get Model Card By Id

Get model card by composite ID: creator_user_id/model_id

Args:: model_id (str): Model ID to search for creator_user_id (str): Creator user ID to search for db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).
Raises:: HTTPException: 404 if model card not found
Returns:: Dict: Model card

Parameters:

model_id (string) –
creator_user_id (string) –

Query Parameters:

convert_s3 (boolean) –

Example request:

GET /models/{creator_user_id}/{model_id} HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

PUT /models/{creator_user_id}/{model_id}#

Update Model Card Metadata By Id

Update model card metadata by ID

Args:

model_id (str): Model Id creator_user_id (str): Creator user id card (UpdateModelCardModel): Updated model card tasks (BackgroundTasks): Background tasks to run db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).

user (TokenData, optional): User data. Defaults to Depends(get_current_user).

Raises:

HTTPException: 404 if model card does not exist HTTPException: 403 if user does not have permission to update model card

Returns:

Optional[Dict]: Updated model card metadata

Parameters:

model_id (string) –
creator_user_id (string) –

Example request:

PUT /models/{creator_user_id}/{model_id} HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "title": "string",
    "description": "string",
    "explanation": "string",
    "usage": "string",
    "limitations": "string",
    "markdown": "string",
    "performance": "string",
    "tags": [
        "string"
    ],
    "task": "string",
    "frameworks": [
        "string"
    ],
    "pointOfContact": "string",
    "owner": "string",
    "videoLocation": "string",
    "inferenceServiceName": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ],
    "experiment": {
        "connector": "",
        "experimentId": "string"
    },
    "dataset": {
        "connector": "",
        "datasetId": "string"
    }
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "title": "string",
    "markdown": "string",
    "performance": "string",
    "task": "string",
    "inferenceServiceName": "string",
    "videoLocation": "string",
    "tags": [
        "string"
    ],
    "frameworks": [
        "string"
    ],
    "description": "string",
    "explanation": "string",
    "usage": "string",
    "limitations": "string",
    "owner": "string",
    "pointOfContact": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ],
    "experiment": {
        "connector": "",
        "experimentId": "string"
    },
    "dataset": {
        "connector": "",
        "datasetId": "string"
    },
    "_id": "string",
    "creatorUserId": "string",
    "modelId": "string",
    "created": "string",
    "lastModified": "string"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

DELETE /models/{creator_user_id}/{model_id}#

Delete Model Card By Id

Delete model card by composite key of Creator User ID and Model ID

Args:

Defaults to Depends(get_db).

user (TokenData, optional): User data. Defaults to Depends(get_current_user).

Raises:

HTTPException: 500 if arbitrary error occurs

Parameters:

model_id (string) –
creator_user_id (string) –

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /models/#

Search Cards

Search model cards

Args:: db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection. Defaults to Depends(get_db). page (int, optional): Page number. Defaults to Query(default=1, alias=”p”, gt=0). rows_per_page (int, optional): Rows per page. Defaults to Query(default=10, alias=”n”, ge=0). descending (bool, optional): Order to return results in. Defaults to Query(default=False, alias=”desc”). sort_by (str, optional): Sort by field. Defaults to Query(default=”_id”, alias=”sort”). generic_search_text (Optional[str], optional): Search through any relevant text fields. Defaults to Query(default=None, alias=”genericSearchText”). tasks (Optional[List[str]], optional): Search by task. Defaults to Query(default=None, alias=”tasks[]”). tags (Optional[List[str]], optional): Search by task. Defaults to Query(default=None, alias=”tags[]”). frameworks (Optional[List[str]], optional): Search by framework. Defaults to Query( default=None, alias=”frameworks[]” ). creator_user_id (Optional[str], optional): Search by creator. Defaults to Query(default=None, alias=”creator”). return_attr (Optional[List[str]], optional): Which fields to return. Defaults to Query(default=None, alias=”return[]”). all (Optional[bool], optional): Whether to return all results. Defaults to Query(default=None).
Returns:: Dict: A dictionary containing the results and pagination information

Query Parameters:

p (integer) –
n (integer) –
desc (boolean) –
sort (string) –
genericSearchText (string) –
title (string) –
tasks[] (array) –
tags[] (array) –
frameworks[] (array) –
creator (string) –
creatorUserIdPartial (string) –
return[] (array) –
all (boolean) –

Example request:

GET /models/ HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "results": [
        {}
    ],
    "total": 1
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /models/#

Create Model Card Metadata

Create model card metadata

Args:

card (ModelCardModelIn): Model card tasks (BackgroundTasks): Background tasks to run db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).

user (TokenData, optional): User data. Defaults to Depends(get_current_user).

Raises:

HTTPException: 409 if model card already exists

Returns:

Dict: Model card metadata

Example request:

POST /models/ HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "title": "string",
    "markdown": "string",
    "performance": "string",
    "task": "string",
    "inferenceServiceName": "string",
    "videoLocation": "string",
    "tags": [
        "string"
    ],
    "frameworks": [
        "string"
    ],
    "description": "string",
    "explanation": "string",
    "usage": "string",
    "limitations": "string",
    "owner": "string",
    "pointOfContact": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ],
    "experiment": {
        "connector": "",
        "experimentId": "string"
    },
    "dataset": {
        "connector": "",
        "datasetId": "string"
    }
}

Status Codes:

201 Created –

Successful Response

Example response:

HTTP/1.1 201 Created
Content-Type: application/json

{
    "title": "string",
    "markdown": "string",
    "performance": "string",
    "task": "string",
    "inferenceServiceName": "string",
    "videoLocation": "string",
    "tags": [
        "string"
    ],
    "frameworks": [
        "string"
    ],
    "description": "string",
    "explanation": "string",
    "usage": "string",
    "limitations": "string",
    "owner": "string",
    "pointOfContact": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ],
    "experiment": {
        "connector": "",
        "experimentId": "string"
    },
    "dataset": {
        "connector": "",
        "datasetId": "string"
    },
    "_id": "string",
    "creatorUserId": "string",
    "modelId": "string",
    "created": "string",
    "lastModified": "string"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /models/{creator_user_id}#

Get Model Cards By User

Get all model cards by a user

Args:

creator_user_id (str): Creator user id db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).

return_attr (Optional[List[str]], optional): Fields to return.: Defaults to Query(None, alias=”return”).

Returns:

Dict: Results and pagination information

Parameters:

creator_user_id (string) –

Query Parameters:

return (array) –

Example request:

GET /models/{creator_user_id} HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "results": [
        {}
    ],
    "total": 1
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

DELETE /models/multi#

Delete Multiple Model Cards

Delete multiple model cards by List of composite keys of Creator User ID and Model ID

Args:

card_package (modelCardPackage): List of Dictionaries containing models with composite ID identifiers to be removed card (UpdateModelCardModel): Updated model card tasks (BackgroundTasks): Background tasks to run db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).

user (TokenData, optional): User data. Defaults to Depends(get_current_user).

Raises:

HTTPException: 500 if arbitrary error occurs

Example request:

DELETE /models/multi HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "card_package": [
        {
            "model_id": "string",
            "creator_user_id": "string"
        }
    ]
}

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /models/export#

Export Models

Example request:

POST /models/export HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "card_package": [
        {
            "model_id": "string",
            "creator_user_id": "string"
        }
    ]
}

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /experiments/{exp_id}#

Get Experiment

Get experiment by ID.

Args:: exp_id (str): Experiment ID connector (Connector): Connector to use return_plots (bool, optional): If plots should be returned. Defaults to True. return_artifacts (bool, optional): If artifacts should be returned. Defaults to True.
Raises:: HTTPException: 404 Not Found if experiment does not exist HTTPException: 500 Internal Server Error if error occurs
Returns:: Dict: Experiment details

Parameters:

exp_id (string) –

Query Parameters:

connector (string) – (Required)
return_plots (boolean) –
return_artifacts (boolean) –

Example request:

GET /experiments/{exp_id}?connector= HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "id": "string",
    "owner": "string",
    "name": "string",
    "project_name": "string",
    "tags": [
        "string"
    ],
    "frameworks": [
        "string"
    ],
    "config": {},
    "scalars": [
        {}
    ],
    "plots": [
        {}
    ],
    "artifacts": {}
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /experiments/clone#

Clone Experiment

Clone experiment. (Not implemented yet.)

Args:: item (ClonePackageModel): Item to clone connector (Connector): Experiment connector
Returns:: Dict: Details of cloned experiment

Query Parameters:

connector (string) – (Required)

Example request:

POST /experiments/clone?connector= HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "id": "string",
    "clone_name": "string"
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /datasets/search#

Search Datasets

Search endpoint for any datasets stored in the current data connector

Args:: query (FindDatasetModel): Search query connectors (Optional[List[Connector]]): Data connector.

If not specified
Returns:: List[Dict]: List of dataset metadata

Query Parameters:

connectors[] (array) –

Example request:

POST /datasets/search HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "id": {},
    "name": "string",
    "tags": [
        "string"
    ],
    "project": "string"
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

[
    {
        "id": "string",
        "name": "string",
        "created": "string",
        "tags": [
            "string"
        ],
        "project": "string",
        "files": {},
        "default_remote": "string",
        "artifacts": [
            {
                "artifactType": "string",
                "name": "string",
                "url": "string",
                "timestamp": "string",
                "framework": "string"
            }
        ]
    }
]

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /datasets/{dataset_id}#

Get Dataset By Id

Get a dataset from it’s ID

Args:: dataset_id (str): ID of dataset (e.g ClearML Dataset ID) connector (Connector): Data connector type
Raises:: HTTPException: 404 Not Found if dataset not found
Returns:: DatasetModel: Dataset with that ID

Parameters:

dataset_id (string) –

Query Parameters:

connector (string) – (Required)

Example request:

GET /datasets/{dataset_id}?connector= HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "id": "string",
    "name": "string",
    "created": "string",
    "tags": [
        "string"
    ],
    "project": "string",
    "files": {},
    "default_remote": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ]
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /datasets/#

Create Dataset

Create a new dataset, based on a dataset file uploaded to it

Args:: file (UploadFile, optional): Archived dataset (e.g zip). Defaults to File(…). dataset_name (str, optional): Name of dataset. Defaults to Form(…). project_name (str, optional): Name of project to uplaod to. Defaults to Form(…). connector(Connector): Data connector to use. output_url (Optional[str], optional): Remote URL to upload file to. Defaults to Form(default=None).
Raises:: HTTPException: 413 Request Entity Too Large if dataset size is too large HTTPException: 415 Unsupported Media Type if wrong file type HTTPException: 500 Internal Server Error if any IOErrors
Returns:: DatasetModel : Created dataset

Status Codes:

201 Created –

Successful Response

Example response:

HTTP/1.1 201 Created
Content-Type: application/json

{
    "id": "string",
    "name": "string",
    "created": "string",
    "tags": [
        "string"
    ],
    "project": "string",
    "files": {},
    "default_remote": "string",
    "artifacts": [
        {
            "artifactType": "string",
            "name": "string",
            "url": "string",
            "timestamp": "string",
            "framework": "string"
        }
    ]
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /iam/add#

Add User

Function to add a user to the MongoDB after reciveing a POST call to the endpoint

Args:: item (UserInsert): The item that contains the user information to be sent to the database for storage db (Any = Depends(get_db)): This retieves the client that is connected to the database.
Returns:: JSONResponse: A Response object with a corresponding code depending on the success of the call or a failure for whatever reason.

Example request:

POST /iam/add HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "name": "string",
    "user_id": "string",
    "password": "string",
    "password_confirm": "string",
    "admin_priv": true
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

DELETE /iam/delete#

Delete User

Delete a user from the database

Args:: userid (UserRemoval): User ID to delete db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).
Raises:: HTTPException: 404 if user not found

Example request:

DELETE /iam/delete HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "users": [
        "string"
    ]
}

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

PUT /iam/edit#

Update User

Example request:

PUT /iam/edit HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "name": "string",
    "user_id": "string",
    "password": "string",
    "password_confirm": "string",
    "admin_priv": true
}

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

PUT /iam/edit/multi#

Update Many User

Example request:

PUT /iam/edit/multi HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "users": [
        "string"
    ],
    "priv": true
}

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /iam/#

Get Users

Get a list of users from the database

Args:: pages_user (UserPage): Pagination input descending (bool, optional): Order of results. Defaults to Query(default=True, alias=”desc”). sort_by (str, optional): Sort field. Defaults to Query(default=”lastModified”, alias=”sort”). db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).
Raises:: HTTPException: 422 if unable to get users HTTPException: 404 if user not found
Returns:: Dict: _description_

Query Parameters:

desc (boolean) –
sort (string) –

Example request:

POST /iam/ HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "page_num": 1,
    "user_num": 1,
    "name": "string",
    "userId": "string",
    "admin_priv": 1,
    "last_modified_range": {},
    "date_created_range": {}
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /engines/{service_name}/logs#

Get Inference Engine Service Logs

Get logs for an inference service

Args:: service_name (str): Name of the service request (Request): FastAPI Request object k8s_client (ApiClient, optional): K8S Client. Defaults to Depends(get_k8s_client).
Raises:: HTTPException: 404 Not Found if service does not exist HTTPException: 500 Internal Server Error if there is an error getting the service logs
Returns:: EventSourceResponse: SSE response with logs

Parameters:

service_name (string) –

Example request:

GET /engines/{service_name}/logs HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

PATCH /engines/{service_name}/scale/{replicas}#

Scale Inference Engine Deployments

Scale the number of replicas of the deployment

Args:: service_name (str): Name of the service replicas (int, optional): Number of replicas. Defaults to Path(ge=0, le=3). k8s_client (ApiClient, optional): K8S Client. Defaults to Depends(get_k8s_client).
Raises:: HTTPException: 404 Not Found if service does not exist HTTPException: 500 Internal Server Error if there is an error scaling the service

Parameters:

service_name (string) –
replicas (integer) –

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /engines/{service_name}#

Get Inference Engine Service

Get Inference Engine Service

Args:: service_name (str): Name of the service db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).
Raises:: HTTPException: 404 Not Found if service does not exist
Returns:: Dict: Service details

Parameters:

service_name (string) –

Example request:

GET /engines/{service_name} HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "modelId": "string",
    "imageUri": "string",
    "containerPort": 1,
    "env": {},
    "numGpus": 1.0,
    "_id": "string",
    "inferenceUrl": "string",
    "ownerId": "string",
    "serviceName": "string",
    "created": "2024-03-05T02:54:36.571224",
    "lastModified": "2024-03-05T02:54:36.571224",
    "host": "string",
    "path": "string",
    "protocol": "string",
    "backend": "knative"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

DELETE /engines/{service_name}#

Delete Inference Engine Service

Delete a deployed inference engine from the K8S cluster.

Parameters:

service_name (string) – Name of KNative service, defaults to Path(description=”Name of KService to Delete”)
k8s_client (ApiClient, optional) – Python K8S Client, defaults to Depends(get_k8s_client)
service_name – Name of KService to Delete

Raises HTTPException:

500 Internal Server Error if deletion fails

Status Codes:

204 No Content – Successful Response

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

PATCH /engines/{service_name}#

Update Inference Engine Service

Update an existing inference engine inside the K8S cluster

Parameters:

service (InferenceEngineService) – Configuration (service name and Image URI) of updated Inference Engine
k8s_client (ApiClient, optional) – Python K8S client, defaults to Depends(get_k8s_client)
service_name (string) –

Raises HTTPException:

500 Internal Server Error if failed to update

Example request:

PATCH /engines/{service_name} HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "imageUri": "string",
    "containerPort": 1,
    "env": {},
    "numGpus": 1.0
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /engines/{service_name}/status#

Get Inference Engine Service Status

Get status of an inference service. This is typically used to give liveness/readiness probes for the service.

Args:: service_name (str): Name of the service k8s_client (ApiClient, optional): K8S Client. Defaults to Depends(get_k8s_client).
Raises:: HTTPException: 404 Not Found if service does not exist HTTPException: 500 Internal Server Error if there is an error getting the service status
Returns:: Dict: Service status

Parameters:

service_name (string) –

Example request:

GET /engines/{service_name}/status HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "serviceName": "string",
    "status": "Pending",
    "message": "string",
    "ready": true,
    "schedulable": true,
    "expectedReplicas": 1
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

GET /engines/#

Get Available Inference Engine Services

Get all available inference engine services

Args:

db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional):: MongoDB Connection. Defaults to Depends(get_db).

Raises:

HTTPException: 500 Internal Server Error if there is an error getting the services

Returns:

list : List of available services

Example request:

GET /engines/ HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

POST /engines/#

Create Inference Engine Service

Create an inference engine service

Args:

service (CreateInferenceEngineService): Service details k8s_client (ApiClient, optional): K8S client. Defaults to Depends(get_k8s_client). db (Tuple[AsyncIOMotorDatabase, AsyncIOMotorClient], optional): MongoDB connection.

Defaults to Depends(get_db).

user (TokenData, optional): User details. Defaults to Depends(get_current_user).

Raises:

HTTPException: 500 Internal Server Error if there is an error creating the service HTTPException: 500 Internal Server Error if the API has no access to the K8S cluster HTTPException: 500 Internal Server Error if there is an error creating the service in the DB HTTPException: 422 Unprocessable Entity if the service already exists HTTPException: 422 Unprocessable Entity if the namespace is not set

Returns:

Dict: Service details

Example request:

POST /engines/ HTTP/1.1
Host: example.com
Content-Type: application/json

{
    "modelId": "string",
    "imageUri": "string",
    "containerPort": 1,
    "env": {},
    "numGpus": 1.0
}

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{
    "modelId": "string",
    "imageUri": "string",
    "containerPort": 1,
    "env": {},
    "numGpus": 1.0,
    "_id": "string",
    "inferenceUrl": "string",
    "ownerId": "string",
    "serviceName": "string",
    "created": "2024-03-05T02:54:36.571224",
    "lastModified": "2024-03-05T02:54:36.571224",
    "host": "string",
    "path": "string",
    "protocol": "string",
    "backend": "knative"
}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

POST /engines/{service_name}/restore#

Restore Inference Engine Service

Restore a deleted service (i.e someone accidently removed the deployment). This will do the following: 1. Grab information about the service from the database 2. Call the delete endpoint to delete the service 3. Call the create endpoint to create the service using the information from the database

TODO: Refactor so that I don’t have to repeat code. Preferably factor out stuff into a controller submodule (split up the code for adding to mongodb and k8s) to allow router to simply compose functions together for better reusability of code

Parameters:

service_name (string) –

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}

422 Unprocessable Entity –

Validation Error

Example response:

HTTP/1.1 422 Unprocessable Entity
Content-Type: application/json

{
    "detail": [
        {
            "loc": [
                "string",
                1
            ],
            "msg": "string",
            "type": "string"
        }
    ]
}

DELETE /engines/admin/clear#

Wipe Orphaned Services

Status Codes:

204 No Content – Successful Response

GET /#

Root

Return a simple message to test if the server is running.

Returns:: JSONResponse: A simple message that says “Hello World”

Example request:

GET / HTTP/1.1
Host: example.com

Status Codes:

200 OK –

Successful Response

Example response:

HTTP/1.1 200 OK
Content-Type: application/json

{}