Working with the model manager API

Overview

Check the list of available endpoints and methods in the model manager API:

Endpoints and methods	Description
GET `/models/{name}`^{see below}	Get detailed information about a model.
PUT `/models/{name}`^{see below}	Replace the data of an existing model.
POST `/models/{name}`^{see below}	Create a new model.
DELETE `/models/{name}` ^{see below}	Delete a model.
PATCH `/models/{name}` ^{see below}	Update an existing model.
GET `/models/{name}` ^{see below}	List the available models in the current domain.
GET `/images/{name}` ^{see below}	Get the binary image of a model.

Endpoints and methods

GET `/models/{name}`

Get detailed information about a model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to get information about.

Query string parameters

Query string parameters are optionally added after the path parameters, preceded by a question mark (?) and separated by an ampersand (&)

Parameter

Type

Description

fast

boolean

If it is true, it does not show the model image in details.

Default value is -.

Example

Find below a request example in cURL language:

curl -X GET "https://api.stage.devo.com/mlmodelmanager/v3/models/MY FIRST MODEL?fast=true" -H "accept: application/json" -H "standAloneToken: cc81f6f5c73634002183d80b1fb736ec"

Response

Code

Description

200

Returns the retrieved model.

{
  "id": 22,
  "name": "MY FIRST MODEL",
  "engine": "H2O",
  "location": "domains/self/632bfd1c-f734-44fd-af66-732410b953f1",
  "description": "A BRIEF DESCRIPTION",
  "updateDate": 1683714063000,
  "creationDate": 1683714063000,
  "domainName": "self",
  "outputType": "float8",
  "image": null,
  "category": "Regression",
  "fields": [
    {
      "id": 0,
      "name": "length",
      "description": "",
      "type": "float8"
    },
    {
      "id": 1,
      "name": "entropy",
      "description": "",
      "type": "float8"
    },
    {
      "id": 2,
      "name": "p_vowels",
      "description": "",
      "type": "float8"
    }
  ],
  "clusters": [],
  "parentId": null,
  "hidden": false,
  "runtimeSize": 78608,
  "fileSize": 77728
}

404

Returns a message when the model name is not found in the server.

500

Returned when an error occurred internally in the server.

PUT `/models/{name}`

Replace the data of an existing model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and active the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to replace the data of.

Request body

The request JSON body must include an object with the following key-value pairs:

Parameter	Type	Description
`engine`	`string`	The engine used to train and execute the model.
`fileName`	`string`
`description`	`string`	The description of the model.

Example

Find below a request example in cURL language:

curl -H "standAloneToken:YOUR_TOKEN" -X GET "https://api-us.devo.com/aggregations/v2/aggrtasks/myDomain/cc_myDomain:abcDE123?skipCache=true"

Response

Code

Description

200

Returns a replaced model.

{
  "name": "string",
  "engine": "string",
  "location": "string",
  "description": "string",
  "updateDate": 0,
  "creationDate": 0,
  "domainName": "string",
  "outputType": "string",
  "image": {
    "image": "string"
  },
  "category": "string",
  "fields": [
    {
      "id": 0,
      "name": "string",
      "description": "string",
      "type": "string"
    }
  ],
  "clusters": [
    {
      "id": 0,
      "name": "string",
      "description": "string",
      "centroid": [
        0
      ]
    }
  ],
  "runtimeSize": 0,
  "fileSize": 0
}

500

Returned when an error occurred internally in the server.

POST `/models/{name}`

Create a model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to create.

Request body

The request JSON body must include an object with the following key-value pairs:

Parameter	Type	Description
`engine`	`string`	The engine used to train and execute the model.
`fileName`	`string`
`description`	`string`	The description of the model.

Example

Find below a request example in cURL language:

Response

Code

Description

200

Model has been successfully saved to the server.

{
  "name": "string",
  "engine": "string",
  "location": "string",
  "description": "string",
  "updateDate": 0,
  "creationDate": 0,
  "domainName": "string",
  "outputType": "string",
  "image": {
    "image": "string"
  },
  "category": "string",
  "fields": [
    {
      "id": 0,
      "name": "string",
      "description": "string",
      "type": "string"
    }
  ],
  "clusters": [
    {
      "id": 0,
      "name": "string",
      "description": "string",
      "centroid": [
        0
      ]
    }
  ],
  "runtimeSize": 0,
  "fileSize": 0
}

409

The model already exists.

500

Returned when an error occurred internally in the server.

DELETE `/models/{name}`

Delete a model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to delete.

Example

Find below a request example in cURL language.

curl -X DELETE "https://api.stage.devo.com/mlmodelmanager/v3/models/MY FIRST MODEL" -H "accept: */*" -H "standAloneToken: cc81f6f5c73634002183d80b1fb736ec"

Response

Code	Description
200	Model successfully deleted.
404	Model not found.
500	Returned when an error occurred internally in the server.

PATCH `/models/{name}`

Update an existing model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to update.

Request body

The request JSON body must include an object with the following key-value pairs:

Parameter	Type	Description
`name`	`string`
`engine`	`string`
`location`	`string`
`description`	`string`
`updateDate`	`integer`
`creationDate`	`integer`
`domainName`	`string`
`outputType`	`string`
`image`
`category`	`string`
`runtimeSize`	`integer`
`fileSize`	`integer`

Example

Find below a request example in cURL language.

Response

Code	Description
200	Returns the patched model.
400	Returns a message when the model name is not found in the server or the received patch JSON is incorrect.
500	Returned when an error occurred internally in the server.

GET `/models/{name}`

List the available models in the current domain.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

There are no path parameters for this request.

Example

Find below a request example in cURL language.

curl -X GET "https://api.stage.devo.com/mlmodelmanager/v3/models" -H "accept: application/json" -H "standAloneToken: cc81f6f5c73634002183d80b1fb736ec"

Response

Code

Description

200

Successful response.

[
  {
    "id": 23,
    "name": "a new model",
    "engine": "H2O",
    "location": "domains/self/e27dd5ef-aedb-42ba-b0b9-21bca5338714",
    "description": "new",
    "updateDate": 1683718486000,
    "creationDate": 1683718486000,
    "domainName": "self",
    "outputType": "float8",
    "image": null,
    "category": "Regression",
    "fields": [
      {
        "id": 0,
        "name": "length",
        "description": "",
        "type": "float8"
      },
      {
        "id": 1,
        "name": "entropy",
        "description": "",
        "type": "float8"
      },
      {
        "id": 2,
        "name": "p_vowels",
        "description": "",
        "type": "float8"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 78608,
    "fileSize": 77728
  },
  {
    "id": 21,
    "name": "dga_classifier_onnx_demo",
    "engine": "ONNX",
    "location": "domains/self/823b3522-895c-411e-a4f3-2762db27a6fc",
    "description": null,
    "updateDate": 1681900499000,
    "creationDate": 1681900499000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 880,
    "fileSize": 3881
  },
  {
    "id": 20,
    "name": "graeme_test2",
    "engine": "ONNX",
    "location": "domains/self/b96fe0d9-74f2-4f51-a301-ff7573c0f6bd",
    "description": null,
    "updateDate": 1681804104000,
    "creationDate": 1681804104000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 876,
    "fileSize": 229468
  },
  {
    "id": 19,
    "name": "graeme_test",
    "engine": "ONNX",
    "location": "domains/self/82178f93-af1b-4d1b-9438-67b842a6bee8",
    "description": null,
    "updateDate": 1681738747000,
    "creationDate": 1681738724000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 876,
    "fileSize": 229468
  },
  {
    "id": 17,
    "name": "dga_scoring",
    "engine": "ONNX",
    "location": "domains/self/85132b1f-59fa-4859-9198-e6ce49258916",
    "description": "DGA domain label scoring",
    "updateDate": 1678361980000,
    "creationDate": 1678361980000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 892,
    "fileSize": null
  },
  {
    "id": 16,
    "name": "pokemon_onnx",
    "engine": "ONNX",
    "location": "domains/self/dd374598-af8e-4542-8663-e7e8e0cdbf38",
    "description": "this is the description",
    "updateDate": 1678201475000,
    "creationDate": 1678201449000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 876,
    "fileSize": 229468
  },
  {
    "id": 4,
    "name": "gptest",
    "engine": "ONNX",
    "location": "domains/self/126e92c2-d724-40cf-a6cf-3dc9f63e9219",
    "description": "",
    "updateDate": 1676390228000,
    "creationDate": 1676390228000,
    "domainName": "self",
    "outputType": "array(float4)",
    "image": null,
    "category": "ONNX",
    "fields": [
      {
        "id": 0,
        "name": "field_0",
        "description": "",
        "type": "array(float4)"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 876,
    "fileSize": 229468
  },
  {
    "id": 2,
    "name": "RRCF",
    "engine": "IDA",
    "location": "domains/self/f868583b-011d-48ea-970c-eb059f06b951",
    "description": "Testing RRCF in Flow",
    "updateDate": 1676291578000,
    "creationDate": 1676291578000,
    "domainName": "self",
    "outputType": "float8",
    "image": null,
    "category": "Rrcf",
    "fields": [
      {
        "id": 0,
        "name": "dimension_0",
        "description": "",
        "type": "float8"
      }
    ],
    "clusters": [],
    "parentId": null,
    "hidden": false,
    "runtimeSize": 23764,
    "fileSize": 227
  }
]

500

Returned when an error occurred internally in the server.

GET `/images/{name}`

Get the binary image of a model.

In order to work with Model Management you need to activate Machine Learning permissions in your role. Go to Administration → Roles → Permissions → Machine Learning and activate the view and manage permissions in Models.

Learn more about roles and permissions in Role permissions.

Request

Path parameters

Add the following path parameters as part of the endpoint:

Parameter	Type	Description
`name` ^required	`string`	Enter the name of the model you want to get the binary image of.

Example

Find below a request example in cURL language.

curl -X GET "https://api.stage.devo.com/mlmodelmanager/v3/images/RRCF" -H "accept: */*" -H "standAloneToken: cc81f6f5c73634002183d80b1fb736ec"

Response

Code	Description
200	Returns the model file as an octet stream. { "metadata": { "id": 123456, "name": "human_usable_saved_name", "algorithm": "rrcf", "version": "1.0" }, "data": { "dimensions": 1, "shingle": 10, "treeSize": 256, "trees": 20 } }
404	Model not found.
500	Returned when an error occurred internally in the server

Working with the model manager API

Overview

Endpoints and methods

GET /models/{name}

Query string parameters

PUT /models/{name}

Request body

POST /models/{name}

Request body

DELETE /models/{name}

Path parameters

PATCH /models/{name}

GET /models/{name}

GET /images/{name}

GET `/models/{name}`

PUT `/models/{name}`

POST `/models/{name}`

DELETE `/models/{name}`

PATCH `/models/{name}`

GET `/models/{name}`

GET `/images/{name}`