Description of image

models

create

Create a model for a URL or Dataset ID

Examples

gradient models create \
  --name "New Model" \
  --modelType "Tensorflow" \
  --datasetRef "dsj1g16f1z67jai"
curl -X POST 'https://api.paperspace.io/mlModels/createModelV2' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '{
   "name": "New Model",
   "datasetRef": "dsj1g16f1z67jai",
   "modelType": "Tensorflow"
}'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.create(
    name='New Model',
    dataset_ref='dsj1g16f1z67jai',
    model_type='Tensorflow'
))

Options

Name Type Attributes Description
--name string required Model name
--modelType string required Model type: Tensorflow, ONNX, Custom
--datasetRef string required Dataset ref to associate model with
--projectId string optional ID of a project
--modelSummary string optional Model summary
--notes string optional Additional notes
--tag string optional One or many tags that you want to add
--tags string optional Tags separated by comma
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

Model created with ID: moseiu8dirfro8k
{
  "id": "moseiu8dirfro8k",
  "projectId": null,
  "experimentId": null,
  "updatedByUserId": "ubx9ce8k",
  "updatedByUserEmail": "[email protected]",
  "accessorHandles": null,
  "modelType": "Tensorflow",
  "name": "New Model",
  "tag": null,
  "summary": null,
  "detail": null,
  "params": null,
  "url": null,
  "notes": null,
  "isDeleted": false,
  "isPublic": false,
  "dtCreated": "2022-03-22T01:41:44.850Z",
  "dtModified": "2022-03-22T01:41:44.850Z",
  "dtUploaded": null,
  "dtDeleted": null,
  "modelPath": null,
  "deploymentState": null,
  "tags": null,
  "datasetId": null,
  "datasetVersion": null,
  "datasetRef": "dsj1g16f1z67jai"
}
moseiu8dirfro8k

details

Show model details

Examples

gradient models details \
  --id "moseiu8dirfro8k"
curl -X GET 'https://api.paperspace.io/mlModels/getModelList/' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '{
    "filter": {
        "where": {
            "and": [
                {
                "id": "moseiu8dirfro8k"
                }
            ]
        }
    }
}'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.get(
    model_id='moseiu8dirfro8k'
))

Options

Name Type Attributes Description
--id string required Model ID
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

+------------------+-----------------+
| ID               | moseiu8dirfro8k |
+------------------+-----------------+
| Name             | New Model       |
| Project ID       | None            |
| Experiment ID    | None            |
| Model Type       | Tensorflow      |
| URL              | None            |
| Deployment State | None            |
| Tags             |                 |
+------------------+-----------------+
{
  "modelList": [
    {
      "id": "moseiu8dirfro8k",
      "projectId": null,
      "experimentId": null,
      "updatedByUserId": "ubx9ce8k",
      "updatedByUserEmail": "[email protected]",
      "accessorHandles": null,
      "modelType": "Tensorflow",
      "name": "New Model",
      "tag": null,
      "summary": null,
      "detail": null,
      "params": null,
      "url": null,
      "notes": null,
      "isDeleted": false,
      "isPublic": false,
      "dtCreated": "2022-03-22T01:41:44.850Z",
      "dtModified": "2022-03-22T01:41:44.850Z",
      "dtUploaded": null,
      "dtDeleted": null,
      "modelPath": null,
      "deploymentState": null,
      "tags": null,
      "datasetVersionId": null,
      "datasetId": null,
      "datasetVersion": null,
      "datasetRef": null
    }
  ],
  "total": 5,
  "displayTotal": 1
}
Model(
  (id = "moseiu8dirfro8k"),
  (name = "New Model"),
  (project_id = None),
  (experiment_id = None),
  (cluster_id = None),
  (tags = []),
  (model_type = "Tensorflow"),
  (url = None),
  (model_path = None),
  (deployment_state = None),
  (summary = None),
  (detail = None),
  (notes = None),
  (dataset_ref = None)
);

list

List models with optional filtering

Examples

gradient models list
curl -X GET 'https://api.paperspace.io/mlModels/getModelList/' \
-H 'x-api-key: d44808a2785d6a...'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.list())

Options

Name Type Attributes Description
--experimentId string optional Filter by experiment ID
--projectId string optional Filter by project ID
--tag string optional Filter by tag
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

+-----------+-----------------+------------+------------+---------------+
| Name      | ID              | Model Type | Project ID | Experiment ID |
+-----------+-----------------+------------+------------+---------------+
| New Model | moseiu8dirfro8k | Tensorflow | None       | None          |
+-----------+-----------------+------------+------------+---------------+
{
  "modelList": [
    {
      "id": "moseiu8dirfro8k",
      "projectId": null,
      "experimentId": null,
      "updatedByUserId": "ubx9ce8k",
      "updatedByUserEmail": "[email protected]",
      "accessorHandles": null,
      "modelType": "Tensorflow",
      "name": "New Model",
      "tag": null,
      "summary": null,
      "detail": null,
      "params": null,
      "url": null,
      "notes": null,
      "isDeleted": false,
      "isPublic": false,
      "dtCreated": "2022-03-22T01:41:44.850Z",
      "dtModified": "2022-03-22T01:41:44.850Z",
      "dtUploaded": null,
      "dtDeleted": null,
      "modelPath": null,
      "deploymentState": null,
      "tags": null,
      "datasetVersionId": null,
      "datasetId": null,
      "datasetVersion": null,
      "datasetRef": null
    }
  ],
  "total": 1,
  "displayTotal": 1
}
[
  Model(
    (id = "moseiu8dirfro8k"),
    (name = "New Model"),
    (project_id = None),
    (experiment_id = None),
    (cluster_id = None),
    (tags = []),
    (model_type = "Tensorflow"),
    (url = None),
    (model_path = None),
    (deployment_state = None),
    (summary = None),
    (detail = None),
    (notes = None),
    (dataset_ref = None)
  ),
];

download

Download model

Examples

gradient models download \
  --id "moseiu8dirfro8k" \
  --destinationDir "./"
curl -X GET 'https://api.paperspace.io/mlModels/listFiles/' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '{
    "id": "moseiu8dirfro8k",
    "links": "True"
}'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.get_model_files(
    model_id='moseiu8dirfro8k',
    links=True
))

Options

Name Type Attributes Description
--id string required Model ID
--destinationDir string required Destination directory
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

Downloading: data.txt
[
  {
    file: "data.txt",
    url: "https://osg-ny2.paperspace.io/tewr3st2z/tewr3st2z/datasets/dsj1g16f1z67jai/versions/xbreynt/data/data.txt?AWSAccessKeyId=6P0NXXIKSB49ZW7ZXESL&Expires=1647944273&Signature=h6MrW%2B2WJuA%2FqE3MyAg9Qxpq4HQ%3D&response-content-disposition=attachment%3Bfilename%3D%22data.txt%22",
  },
];
[
  ModelFile(
    (file = "data.txt"),
    (url =
      "https://osg-ny2.paperspace.io/tewr3st2z/tewr3st2z/datasets/dsj1g16f1z67jai/versions/xbreynt/data/data.txt?AWSAccessKeyId=6P0NXXIKSB49ZW7ZXESL&Expires=1647944439&Signature=Ji%2Fsc0tJYuGGMcgxvQFtRnS4Wio%3D&response-content-disposition=attachment%3Bfilename%3D%22data.txt%22"),
    (size = None)
  ),
];

upload

Upload a model file or directory

Examples

gradient models upload fashion-mnist.onnx \
  --name "New Model" \
  --modelType "ONNX"
This command is not available via CURL.
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.upload(
    path='./fashion-mnist.onnx',
    name='New Model',
    model_type='ONNX'
))

Options

Name Type Attributes Description
--name string required Model name
--modelType string required Model type: Tensorflow, ONNX, Custom
--projectId string optional ID of a project
--clusterID string optional ID of a cluster
--modelSummary string optional Model summary
--notes string optional Additional notes
--tag string optional One or many tags that you want to add
--tags string optional Tags separated by comma
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

100% (408542 of 408542) |#####################################################################################################################| Elapsed Time: 0:00:00 Time:  0:00:00
Model uploaded with ID: moc0ackf1ivmmy
This command is not available via CURL.
moc0ackf1ivmmy

delete

Delete a model

Examples

gradient models delete \
  --id "moseiu8dirfro8k"
curl -X POST 'https://api.paperspace.io/mlModels/deleteModel/' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '	{
        "id": "moseiu8dirfro8k"
    }'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.delete(
    model_id='moseiu8dirfro8k'
))

Options

Name Type Attributes Description
--id string required Model ID
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

Model deleted
None

tags add

Add tags to ml model

Examples

gradient models tags add \
  --id "moseiu8dirfro8k" \
  --tag "import202203"
curl -X PUT 'https://api.paperspace.io/entityTags/updateTags' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '{
    "entity": "mlModel",
    "entityId": "moseiu8dirfro8k",
    "tags": [
        "import202203"
    ]
}'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.add_tags(
    entity_id='moseiu8dirfro8k',
    tags=['import202203']
))

Options

Name Type Attributes Description
--id string required Model ID
--tag string optional One or more tags that you want to add to ml model
--tags string optional Separated by comma tags taht you want to add to ml model
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

Tags added to ml model
[
  {
    "tagId": 680,
    "entity": "mlModel",
    "entityId": 32023,
    "dtCreated": "2022-03-22T02:37:43.788Z",
    "dtDeleted": null,
    "userId": 750386,
    "id": 7550,
    "entity_id": 32023,
    "tag": {
      "name": "import202203",
      "dtCreated": "2022-03-15T01:29:39.877Z",
      "id": 680
    }
  }
]
None

tags remove

Remove tags from ml model

Examples

gradient models tags remove \
  --id "moseiu8dirfro8k" \
  --tag "import202203"
curl -X PUT 'https://api.paperspace.io/entityTags/updateTags' \
-H 'x-api-key: d44808a2785d6a...' \
-H 'Content-Type: application/json' \
--data-raw '{
    "entity": "mlModel",
    "entityId": "moseiu8dirfro8k",
    "tags": []
}'
from gradient import ModelsClient

api_key='d44808a2785d6a...'

models_client = ModelsClient(api_key)

print(models_client.remove_tags(
    entity_id='moseiu8dirfro8k',
    tags=['import202203']
))

Options

Name Type Attributes Description
--id string required Model ID
--tag string optional One or more tags that you want to add to ml model
--tags string optional Separated by comma tags taht you want to add to ml model
--apiKey string optional API key to use this time only
--optionsFile string optional Path to YAML with predefined options
--createOptionsFile string optional Generate template options file

Response

Tags removed from ml model
[]
None