Result Types

Workspace

Workspace details for a specific workspace given by a provided secret key.

Click to expand

Applicable Functions

client.get_info()

Attributes

Workspace(
    id="ws_1c8aab980f174b0296c7e35e88665b13",
    name="Raighne's Workspace",
    owner="user_6323fea23e292439f31c58cd",
    tier="Developer",
    create_date=1701927649302
)

Name	Type	Description
`id`	`str`	Unique ID of the current workspace.
`name`	`str`	Name of the current workspace. You can modify this on Nexus directly by clicking on the Settings button on the top right of the Workspace Dashboard.
`owner`	`str`	Unique user ID of the workspace owner. This is tied to the email used to sign up for the Nexus account.
`tier`	`str`	Current workspace tier that determines access to advanced features and increased resource quotas. To enjoy these benefits, check out how to Upgrade Your Plan.
`create_date`	`int`	UNIX timestamp of workspace creation date.

Project

Project results for a specific project given by the project key.

Click to expand

Applicable Functions

Attributes

Project(
    id='proj_9004a21df7b040ace4674c4879603fe8',
    name='keypoints',
    workspace_id='ws_1c8aab980f174b0296c7e35e88665b13',
    type='ObjectDetection',
    create_date=1701927649302,
    localization='MULTI',
    tags=['cat faces'],
    groups=['main', 'cats'],
    statistic=Statistic(
        tags_count=[TagsCountItem(name='cat faces', count=0)],
        total_assets=28,
        annotated_assets=0,
        total_annotations=0
    )
)

Attribute	Type	Description
`id`	`str`	Project ID.
`name`	`str`	Project name.
`workspace_id`	`str`	Workspace ID of the project.
`type`	`str`	Project task type, one of [`Classification`, `ObjectDetection`, `Segmentation`]
`create_date`	`int`	UNIX timestamp of project creation date.
`localization`	`str`	Region(s) for data localization.
`tags`	`list[str]`	List of tag names.
`groups`	`list[str]`	List of assets groups.
`statistic`	`dict`	Contains project statistics for the following categories: - `tags_count`: List of tag counts - `asset_total`: Total number of assets in the project. - `asset_annotated`: Total number of annotated assets in the project. - `annotation_total`: Total number of annotations in the project.

Statistic

Project-level statistics on tags, assets, and annotations.

Click to expand

Applicable Functions

project.get_info()

Attributes

Statistic(
    tags_count=[TagsCountItem(name="cat faces", count=0)],
    total_assets=28,
    annotated_assets=0,
    total_annotations=0
)

Name	Type	Description
`tags_count`	`list[TagCountItem]`	List of `TagCountItem` objects representing tag counts for each tag in the project.
`total_assets`	`int`	Total number of assets in the project.
`annotated_assets`	`int`	Total number of annotated assets in the project.
`total_annotations`	`int`	Total number of annotations in the project.

User

User metadata.

Applicable Functions

project.list_users()

ProjectUser(
    id='user_6323fea23e292439f31c58cd',
    access_type='Owner',
    email='raighne@datature.io',
    nickname='raighne',
    picture='https://s.gravatar.com/avatar/avatars%2Fra.png'
)

Attribute	Type	Description
`id`	`str`	User ID.
`access_type`	`str`	The access type of the current project, one of [`Owner`, `Collaborator`, `Labeller`] .
`email`	`str`	User email.
`nickname`	`str`	User nickname.
`picture`	`str`	User profile picture.

AssetResults

Data results for a specific asset.

Applicable Functions

Asset(
    id='asset_8208740a-2d9c-46e8-abb9-5777371bdcd3',
    filename='boat180.png',
    project='proj_cd067221d5a6e4007ccbb4afb5966535',
    status='None',
    create_date=1701927649302,
    url='',
    metadata=AssetMetadata(
        file_size=186497,
        mime_type='image/png',
        height=243,
        width=400,
        groups=['main'],
        custom_metadata={'captureAt': '2021-03-10T09:00:00Z'}
    ),
    statistic=AssetAnnotationsStatistic(
        tags_count=[],
        total_annotations=0
    )
)

Attribute	Type	Description
`id`	`str`	Asset ID.
`filename`	`str`	File name of the asset.
`project`	`str`	Project ID in which the asset is contained.
`status`	`str`	The status of the asset, enum: [`Annotated`, `Review`, `Completed`, `Tofix`, `None`]
`create_date`	`int`	UNIX timestamp of when the asset was uploaded.
`url`	`str`	URL to the raw asset file.
`metadata`	`dict`	Asset metadata.
`statistic`	`dict`	Asset annotation statistics.

AssetMetadata

Metadata for a specific asset.

Applicable Functions

AssetMetadata(
    file_size=186497,
    mime_type='image/png',
    height=243,
    width=400,
    groups=['main'],
    custom_metadata={'captureAt': '2021-03-10T09:00:00Z'}
)

Attribute	Type	Description
`file_size`	`int`	Size of the asset in bytes.
`mime_type`	`str`	Media type and format of the asset.
`height`	`int`	Pixel height of the asset.
`width`	`int`	Pixel width of the asset.
`groups`	`List[str]`	The groups of the asset.
`custom_metadata`	`dict`	The custom metadata of the asset.

AssetStatistics

Data statistics for a specific asset.

Applicable Functions

AssetAnnotationsStatistic(
    tags_count= [
        TagsCountItem(name="tagName1", count=1)
    ],
    total_annotations= 2
)

Attribute	Type	Description
`tags_count`	`list[dict]`	List of tag counts.
`total_annotations`	`int`	Total number of annotations in the asset.

GroupStatistics

Asset group statistics.

Applicable Functions

project.assets.list_groups()

[
    AssetGroup(
        group='1', 
        statistic=AssetGroupStatistic(
            total_assets=1, 
            annotated_assets=0, 
            reviewed_assets=0, 
            to_fixed_assets=0, 
            completed_assets=0
        )
    )
]

Attribute	Type	Description
`group`	`str`	Name of the asset group.
`statistic`	`dict`	Contains asset counts of the following categories: - `total_assets`: Total number of assets in the asset group - `annotated_assets`: Total number of annotated assets in the asset group. - `reviewed_assets`: Total number of reviewed assets in the asset group. - `to_fixed_assets`: Total number of assets in which annotations need to be fixed in the asset group. - `completed_assets`: Total number of assets that have completed the annotation pipeline in the asset group.

TagCountItem

Total count of instances of a tag.

Applicable Functions

project.get_info()

Attributes

TagCountItem(
  name="tagName1",
  count=1
)

Name	Type	Description
`name`	`str`	Tag name.
`count`	`int`	Total count of instances of the tag.

AnnotationMetadata

Metadata for a specific annotation.

Applicable Functions

Annotation(
    id='annot_a9ff9b21-c0e2-49ff-8a69-773aaf00a6f8',
    project_id='proj_cd067221d5a6e4007ccbb4afb5966535',
    asset_id='asset_f4dcb429-0332-4dd6-a1b4-fee794031ba6',
    tag='boat',
    bound_type='Rectangle',
    create_date=1701927649302,
    bound=[
        [0.2772511848341232, 0.34635416666666663],
        [0.2772511848341232, 0.46875],
        [0.54739336492891, 0.46875],
        [0.54739336492891, 0.34635416666666663]
    ]
)

Attribute	Type	Description
`id`	`str`	Unique ID of the annotation.
`project_id`	`str`	ID of the project containing the annotation.
`asset_id`	`str`	ID of the asset containing the annotation.
`tag`	`str`	Tag name of the annotation.
`bound_type`	`str`	Bound type of the annotation, one of [`rectangle`, `polygon`].
`bound`	`list[list[float]]`	Bound vertices with the following format: `[[x1, y1], [x2, y2], ... , [xn, yn]]`

WorkflowMetadata

Metadata for a training workflow.

Applicable Functions

Workflow(
    id='flow_64e812a7e47592ef374cbbc2',
    project_id='proj_cd067221d5a6e4007ccbb4afb5966535',
    title='Yolov8 Workflow',
    create_date=1701927649302,
    update_date=1701927649302
)

Attribute	Type	Description
`id`	`str`	ID of the workflow.
`title`	`str`	Name of the workflow.
`project_id`	`str`	Project ID containing the workflow.
`update_date`	`int`	Last updated UNIX timestamp of the workflow.

TrainingInsight

Insight metadata for a training run.

Applicable Functions

project.list_insights()

ProjectInsight(
    flow_title='Test workflow', 
    run_id='run_4a5d406d-464d-470c-bd7d-e92456621ad3', 
    dataset=InsightDataset(
        data_type='Rectangle', 
        num_classes=1, 
        average_annotations=5.19, 
        total_assets=500, 
        settings=DatasetSettings(
            split_ratio=0.3, 
            shuffle=True, 
            seed=0, 
            using_sliding_window=False
        )
    ), 
    model=InsightModel(
        name='fasterrcnn-inceptionv2-1024x1024', 
        batch_size=2, 
        training_steps=5000, 
        max_detection_per_class=100, 
        solver='momentum', 
        learning_rate=0.04, 
        momentum=0.9
    ), 
    checkpoint=RunCheckpoint(
        strategy='STRAT_ALWAYS_SAVE_LATEST', 
        evaluation_interval=250, 
        metric=None
    ), 
    artifact=InsightArtifact(
        id='artifact_65ae274540259e2a07533532', 
        is_training=False, 
        step=5000, 
        metric=ArtifactMetric(
            total_loss=0.32356, 
            classification_loss=0.012036, 
            localization_loss=0.010706, 
            regularization_loss=0.0
        )
    ),
    create_date=1705912133684
)

Attribute	Type	Description
`flow_title`	`str`	Name of the workflow.
`run_id`	`str`	ID of the training run.
`dataset`	`dict`	Contains loss metrics for the following categories:
`step`	`int`	Total number of training steps.
`create_date`	`int`	UNIX timestamp of the training creation date.
`metric`	`dict`	Contains loss metrics for the following categories: - `total_loss` - `classification_loss` - `localization_loss` - `regularization_loss`
`statistic`	`dict`	Contains dataset statistics for the following categories: - `average_annotations`: Average number of annotations per asset.
`optimizer`	`str`	Name of the optimizer used in the training.
`learning_rate`	`float`	Value of the learning rate used in the training.
`momentum`	`float`	Value of the momentum used in the training.
`epochs`	`int`	Total number of training epochs.
`batch_size`	`int`	Value of the batch size used in the training.
`model_name`	`str`	Name of the specific model architecture used in the training.
`max_detections_per_class`	`int`	Value to cap the maximum number of detections per class for the model.
`data_type`	`str`	Annotation data type.
`num_classes`	`int`	Total number of unique classes.
`split_ratio`	`float`	Train-test split ratio.
`shuffle`	`bool`	Whether the dataset was shuffled.
`seed`	`int`	Initialization seed for the training.
`checkpoint_every_n`	`int`	Epoch interval to generate checkpoints.
`metric_target`	`str`	Metric used to determine best checkpoint saved.

TrainingMetadata

Metadata for training runs.

Bases

dict

Applicable Functions

{
    "id": "run_63eb212ff0f856bf95085095",
    "object": "run",
    "project_id": "proj_cd067221d5a6e4007ccbb4afb5966535",
    "flow_id": "flow_63bbd3bf8a78eb906f417396",
    "status": {
        "conditions": [
            {
                "condition": "TrainingStarted",
                "last_updated": 1676353954729,
                "status": "finished"
            },
            {
                "condition": "TrainingFinished",
                "last_updated": 1676356061724,
                "status": "finished"
            }
        ],
        "last_updated": 1676356061724
    },
    "execution": {
        "accelerator": {
            "name": "GPU_T4",
            "count": 2
        },
        "checkpoint": {
            "strategy": "STRAT_LOWEST_VALIDATION_LOSS",
            "evaluation_interval": 250,
            "metric": "Loss/total_loss"
        }
    },
    "features": {
        "matrix": true,
        "preview": true
    },
    "create_date": 1676353954729,
    "last_modified_date": 1676356061724,
    "logs": [
        "log_63eb212ff0f856bf95085095"
    ]
}

Attribute	Type	Description
`id`	`str`	Training run ID.
`object`	`str`	Type of object.
`project_id`	`str`	ID of the project containing the training run.
`flow_id`	`str`	ID of the workflow used for the training run.
`status`	`dict`	Status of completion of the different stages in the training run, contains the following categories: - `conditions`: List of dictionaries that describe the training conditions (`TrainingStarted`, `TrainingFinished`), last updated UNIX timestamp of the operation, and the status of completion. - `last_updated`: UNIX timestamp of when the training statuses were last updated.
`execution`	`dict`	Contains training configuration parameters for the following categories: - `accelerator`: Dictionary containing the type and number of GPUs used. - `checkpoint`: Dictionary containing the checkpoint strategy, evaluation interval, and metric used to save the best checkpoint.
`features`	`dict`	Contains the activation status of certain advanced visualization features such as Evaluation Preview and Confusion Matrix.
`create_date`	`int`	UNIX timestamp of the training creation date.
`last_modified_date`	`int`	UNIX timestamp of the last modified date of the training.
`logs`	`list[str]`	List of training log IDs that can be used to view training logs via `datature.Run.log()`

LogMetadata

Metadata for training logs.

Bases

dict

Applicable Functions

datature.Run.log()

{
    "id": "log_63eb212ff0f856bf95085095",
    "object": "log",
    "event": [
        {
            "ev": "memoryUsage",
            "pl": {},
            "t": 1675669392000
        }
    ]
}

Attribute	Type	Description
`id`	`str`	Log ID.
`object`	`str`	Type of object.
`event`	`list[dict]`	List of training event logs containing the following categories: - `ev`: Type of event tracked. - `pl`: Detailed description of the logs. - `t`: UNIX timestamp of the event.

ArtifactMetadata

Metadata for artifacts.

Bases

dict

Applicable Functions

{
    "id": "artifact_63bd140e67b42dc9f431ffe2",
    "object": "artifact",
    "is_training": false,
    "step": 3000,
    "flow_title": "Blood Cell Detector",
    "run_id": "run_63bd08d8cdf700575fa4dd01",
    "files": [
        {
            "name": "ckpt-13.data-00000-of-00001",
            "md5": "5a96886e53f98daae379787ee0f22bda"
        }
    ],
    "project_id": "proj_cd067221d5a6e4007ccbb4afb5966535",
    "artifact_name": "ckpt-13",
    "create_date": 1673335822851,
    "metric": {
        "total_loss": 0.548,
        "classification_loss": 0.511,
        "localization_loss": 0.006,
        "regularization_loss": 0.03
    },
    "is_deployed": false,
    "exports": ["onnx", "tflite"],
    "model_type": "efficientdet-d1-640x640",
    "exportable_formats": ["tensorflow", "tflite", "onnx", "pytorch"]
}

Attribute	Type	Description
`id`	`str`	Artifact ID.
`object`	`str`	Type of object.
`is_training`	`bool`	Whether the training is still running.
`step`	`int`	Total number of training steps.
`flow_title`	`str`	Title of the workflow.
`run_id`	`str`	ID of the training run of the current artifact.
`files`	`list[dict]`	List of artifact checkpoint files containing the following categories: - `name`: Name of the checkpoint file. - `md5`: MD5 hash value of the checkpoint file.
`project_id`	`str`	ID of the project containing the current artifact.
`artifact_name`	`str`	Checkpoint name of the artifact.
`create_date`	`int`	UNIX timestamp of the artifact creation date.
`metric`	`dict`	Dictionary containing the following metrics: - `total_loss` - `classification_loss` - `localization_loss` - `regularization_loss`
`is_deployed`	`bool`	Whether the current artifact has an active deployment.
`exports`	`list[str]`	List of model formats that the artifact has been exported in.
`model_type`	`str`	Model architecture name.
`exportable_formats`	`list[str]`	List of all exportable model formats for the artifact.

ExportedMetadata

Metadata of exported models.

Bases

dict

Applicable Functions

datature.Artifact.list_exported()

{
    "id": "model_d15aba68872b045e27ac3db06a401da3",
    "object": "model",
    "status": "Finished",
    "format": "tensorflow",
    "create_date": 1673336054173,
    "download": {
        "method": "GET",
        "expiry": 1673339505871,
        "url": "https://storage.googleapis.com/exports.datature.ioa2d89"
    }
}

Attribute	Type	Description
`id`	`str`	ID of the exported model.
`object`	`str`	Type of object.
`status`	`str`	Status of the model export.
`format`	`str`	Exported model format.
`create_date`	`int`	UNIX timestamp of the creation date of the exported model.
`download`	`dict`	Dictionary containing the download metadata of the exported model: - `method`: Request method - `expiry`: UNIX timestamp of the expiry of the download link. - `url`: Download link of the exported model.

DeploymentMetadata

Metadata for active deployments.

Bases

dict

Applicable Functions

{
    "id": "deploy_30922d5e-b2f6-43dc-b7b4-e29e2c30fb45",
    "object": "deploy",
    "project_id": "proj_cd067221d5a6e4007ccbb4afb5966535",
    "model_id": "model_5cbcf0fb9f692621f9cfc1ea8f1d68c7",
    "scaling": {
        "mode": "fixed",
        "num_instances": 1
    },
    "status": {
        "overview": "AVAILABLE",
        "message": "Created service successfully",
        "status_date": 1673936127872
    },
    "create_date": 1673935589700,
    "last_modified_date": 1673936127872,
    "locked": false,
    "url": "https://inference.datature.io/neural/3092245/predict"
}

Attribute	Type	Description
`id`	`str`	ID of the active deployment.
`object`	`str`	Type of object.
`project_id`	`str`	Project ID containing the active deployment.
`model_id`	`str`	ID of the model used for the deployment.
`scaling`	`dict`	Dictionary containing the following categories: - `mode`: Instance scaling mode of the deployment. - `num_instances`: Number of instances of the deployment.
`status`	`dict`	Dictionary of the deployment status containing the following categories: - `overview`: Overview status of the deployment. - `message`: Status message of the deployment. - `status_date`: UNIX timestamp of the last update of the deployment status.
`create_date`	`int`	UNIX timestamp of the creation date of the deployment.
`last_modified_date`	`int`	UNIX timestamp of the last modified date of the deployment.
`locked`	`bool`	Whether the deployment is locked.
`url`	`str`	API URL endpoint for prediction requests.

OperationMetadata

Metadata for background operations.

Bases

dict

Applicable Functions

datature.Operation.retrieve()

{
    "id": "op_508fc5d1-e908-486d-9e7b-1dca99b80024",
    "object": "operation",
    "op_link": "users/api|affaf/proje-1dca99b80024",
    "status": {
        "overview": "Queued",
        "message": "Operation queued",
        "time_updated": 1676621361765,
        "time_scheduled": 1676621361765,
        "progress": {
            "unit": "whole operation",
            "with_status": {
                "queued": 1,
                "running": 0,
                "finished": 0,
                "cancelled": 0,
                "errored": 0
            }
        }
    }
}

Attribute	Type	Description
`id`	`str`	Unique operation ID.
`object`	`str`	Type of object.
`op_link`	`str`	Operation link used to retrieve operation status.
`status`	`dict`	Operation status metadata.

OperationStatus

Metadata of operation status.

Bases

dict

{
    "overview": "Queued",
    "message": "Operation queued",
    "time_updated": 1676621361765,
    "time_scheduled": 1676621361765,
    "progress": {
        "unit": "whole operation",
        "with_status": {
            "queued": 1,
            "running": 0,
            "finished": 0,
            "cancelled": 0,
            "errored": 0
        }
    }
}

Attribute	Type	Description
`overview`	`str`	Overview status of current operation.
`message`	`str`	Status message of current operation.
`time_updated`	`int`	Last updated UNIX timestamp of current operation status.
`time_scheduled`	`int`	UNIX timestamp of when the operation was first scheduled.
`progress`	`dict`	Operation progress status indicator.