Skip to content
This repository has been archived by the owner on Jun 8, 2023. It is now read-only.

Latest commit

 

History

History
364 lines (287 loc) · 11.1 KB

README.md

File metadata and controls

364 lines (287 loc) · 11.1 KB
Raw

Mantium API Client Library

Build Status

Table of Contents

Quickstart:

Read the getting started guide for more information on how to use Mantium.

Authentication

  • Make an account by visiting app.mantiumai.com and select Register.
  • Enter your email address and create a password. After you've verified the email, you'll be able to sign in to the Mantium application. You'll also need your username and password to obtain a token for API use.

Installation

To install the python Library please use the following command.

pip install mantiumapi

Usage

Set authentication credentials in your environment. Setting MANTIUM_USER and MANTIUM_PASSWORD will allow the client to obtain the authentication token, which will re-authenticate when the token expires.

It's also possible to directly set the token through the MANTIUM_TOKEN environment variable. Documentation for authenticating can be found here

  • Linux/MacOS
export MANTIUM_USER='<username>'
export MANTIUM_PASSWORD='<password>'
  • Windows
set MANTIUM_USER="<username>"
set MANTIUM_PASSWORD="<password>"
  • Windows Powershell
$env:MANTIUM_USER="<username>"
$env:MANTIUM_PASSWORD="<password>"

AI Methods

Get all of the supported ai_methods for a provider

List Methods

Requires AI Provider name

>>> from mantiumapi import AiMethods
>>> ai_methods = AiMethods.get_list("openai")
>>> print([method.name for method in ai_methods])
['answers', 'classifications', 'completion', 'search']

AI Engines

Get available AI engines

Get All AI Engines

Get all of the configured and available AI engines

>>> from mantiumapi import AiEngine
>>> ai_engines = AiEngine.get_list()
>>> print([(engine.ai_provider, engine.name) for engine in ai_engines])
[('OpenAI', 'content-filter-alpha-c4'), ...]

Get Ai Engines for a single provider

>>> ai_engines = AiEngine.from_provider("Cohere")
>>> print([(engine.ai_provider, engine.name) for engine in ai_engines])
[('Cohere', 'baseline-otter'), ...]

Go to Table of Contents

Tags

List Tags

Get all of the tags for your selected organization.

Query Params Page Page number Size Page size. If not supplied, returns all the results in a single page for certain APIs.

>>> from mantiumapi import Tag
>>> tags = Tag.get_list()
>>> print([tag.name for tag in tags])
['Tag 1', 'Tag 2']

Passing URL parameters

>>> tags = Tag.get_list(params={"page":1,"size":1})
>>> print([tag.name for tag in tags])
['Tag 1']

Get Tag by ID

>>> tag = Tag.from_id('8c809efe-4738-4c7a-93e1-ea933ef84172')
>>> tag.name
['Tag 1']

Create Tag

>>> new_tag = Tag()
>>> new_tag.name = "My new tag"
>>> new_tag.description = "This is a new tag"
>>> new_tag.create()

Update Tag

>>> tag = Tag.from_id('8c809efe-4738-4c7a-93e1-ea933ef84172')
>>> tag.name = "New tag name"
>>> tag.save()

Delete Tag

>>> tag = Tag.from_id('8c809efe-4738-4c7a-93e1-ea933ef84172')
>>> tag.refres()
>>> tag.delete()

Go to Table of Contents

Prompts

List Prompts

List all of your organization's prompts.

Optional Query String Parameters

  • page - The page of records to return. Optional, defaults to page 1.
  • size - the number of records to return for each page. Optional, defaults to 20 - prompts a page.
  • schema_class - not used, exclude.
  • tags - A list of Tag IDs separated by comma used to filter the results, optional.

Document link

>>> from mantiumapi import Prompt
>>> prompts = Prompt.get_list(params={"size":2})
>>> print([prompt.name for prompt in prompts])
['My first prompt', 'Test prompt']

Get Prompt by ID

>>> prompt = Prompt.from_id('3184ba90-c2b1-4604-bbd2-6ed436ca5f52')
>>> prompt.name
'My first prompt'

Create Prompt

>>> new_prompt.name = "A new prompt"
>>> new_prompt.prompt_text = "Prompt text"
>>> new_prompt.ai_provider = "OpenAI"
>>> new_prompt.default_engine = "davinci"
>>> new_prompt.ai_method = "completion"
>>> new_prompt.ai_engine_id = "b2ffecf7-4fee-42e7-b85d-a5e28d939396"
>>> new_prompt.text = "Prompt text"
>>> new_prompt.prompt_parameters = {"basic_settings": {"temperature": 1,"max_tokens": 16,"frequency_penalty": 0,"presence_penalty": 0,"top_p": 1},"advanced_settings": {"best_of": 1,"n": 1,"echo": "false","stream": "false"}}
>>> new_prompt.create()

Update Prompt

>>> prompt = Prompt.from_id('3184ba90-c2b1-4604-bbd2-6ed436ca5f52')
>>> prompt.name = "New Prompt name"
>>> prompt.update()

Delete Prompt

>>> prompt = Prompt.from_id('3184ba90-c2b1-4604-bbd2-6ed436ca5f52')
>>> prompt.refresh()
>>> prompt.delete()

Execute Prompt

Asynchronously submit input to a Prompt for execution. Returns a PromptExecute object to manage the result.

As this is an asynchronous endpoint, the first result returned will not be a finished result. The result can be updated by calling .refresh() on the PromptExecute object.

  • Input (string)- Data to append to Prompt for execution
>>> prompt = Prompt.from_id('0aab4c37-d931-4726-8768-b7ff91776ce6')
>>> result = Prompt.execute('Data for the Prompt')
>>> result.__dict__
{'prompt_execution_id': '221202da-9d76-4711-91f0-e0d8b50a57d5', 'prompt_id': '0aab4c37-d931-4726-8768-b7ff91776ce6', 'input': 'Data for the Prompt', 'output': '', 'reason': '', 'status': 'RUNNING', 'error': '', 'warning_message': '', 'hitl_info': None}
>>> result.refresh()
>>> result.refresh()
>>> result.__dict__
{'prompt_execution_id': '221202da-9d76-4711-91f0-e0d8b50a57d5', 'prompt_id': '0aab4c37-d931-4726-8768-b7ff91776ce6', 'input': 'Data for the Prompt', 'output': 'Output from the Prompt Exeuction', 'reason': '', 'status': 'COMPLETED', 'error': '', 'warning_message': '', 'hitl_info': None}

Go to Table of Contents

Intelets

Intelets organize multiple prompts by grouping them together sequentially so that the output of one prompt feeds into the input of the next - this enables the creation of complex AI data pipelines for processing text.

List Intelets

List all of your organization's prompts.

Optional Query String Parameters

  • page - The page of records to return. Optional, defaults to page 1.
  • size - the number of records to return for each page. Optional, defaults to 20 - prompts a page.
>>> from mantiumapi import Intelet
>>> intelets = Intelet.get_list()
>>> print([i.name for i in intelets])
['My first Intelet', 'Processing pipeline', ...]

Get Intelet by ID

>>> intelet = Intelet.from_id('0aab4c37-d931-4726-8768-b7ff91776ce6')
>>> print(intelet.name)
'My first Intelet'

Create Intelet

The order of Prompts in the prompts list dictate in which order they will be executed in the pipeline.

>>> new_intelet=Intelet()
>>> new_intelet.name = "Name of new Intelet"
>>> new_intelet.description = "Description of the Intelet"
>>> new_intelet.prompts = ['3184ba90-c2b1-4604-bbd2-6ed436ca5f52']
>>> new_intelet.save()

Update Intelet

>>> intelet = Intelet.from_id('3184ba90-c2b1-4604-bbd2-6ed436ca5f52')
>>> intelet.name = "New Intelet name"
>>> intelet.update()

Delete Intelet

>>> intelet = intelet.from_id('3184ba90-c2b1-4604-bbd2-6ed436ca5f52')
>>> intelet.refresh()
>>> intelet.delete()

Execute Intelet

Asynchronously submit input to an Intelet for execution. Returns a InteletExecute object to manage the result.

As this is an asynchronous endpoint, the first result returned will not be a finished result. The result can be updated by calling .refresh() on the InteletExecute object.

  • Input (string)- Data to append to Prompt for execution
>>> intelet = Intelet.from_id('3c3a14f3-aeaa-468b-8b8f-8d8a053f1719')
>>> result = Intelet.execute('Sample input text')
>>> result.__dict__
{'intelet_execution_id': 'd81fba13-d3fc-41af-9f0b-10816b7ca5df', 'intelet_id': '3c3a14f3-aeaa-468b-8b8f-8d8a053f1719', 'input': 'Sample input text', 'output': '', 'reason': '', 'status': 'QUEUED', 'error': '', 'executed_prompts': [], 'pending_prompts': ['760311e5-9137-4ed6-dc2a-34f077b44131', '3a063314-1fe5-42f2-a276-c25404071c3d', '6004bfd0-5bec-4f41-bd89-1e4a6ca3f76b'], 'results': []}
>>> result.refresh()
>>> result.__dict__
{'intelet_execution_id': 'd81fba13-d3fc-41af-9f0b-10816b7ca5df', 'intelet_id': '3c3a14f3-aeaa-468b-8b8f-8d8a053f1719', 'input': 'Sample input text', 'output': 'Output of Intelet', 'reason': '', 'status': 'COMPLETED', 'error': '', 'executed_prompts': ['760311e5-9137-4ed6-dc2a-34f077b44131', '3a063314-1fe5-42f2-a276-c25404071c3d', '6004bfd0-5bec-4f41-bd89-1e4a6ca3f76b'], 'pending_prompts': [], 'results': []}

Go to Table of Contents

Logs

List Logs

Query Params

  • page (int) - Page number
  • size (int) - Page size. If not supplied, returns all the results in a single page for certain APIs.
  • after_date (string) - After Date
  • before_date (string) Before Date
  • log_type (string) LogType, An enumeration. [AUTH | DEFAULT | PROMPT | INTELET FILE]
  • log_level (string) Log Level
  • log_status (string) Log Status
>>> from mantiumapi import Log
>>> logs = Log.get_list(params={"size":5, "log_type":"PROMPT"})
>>> print([(l.log_type, l.log_payload) for l in logs])
[('PROMPT', {'to': 'completion', 'name': 'OpenAI Completion', 'error': '', 'input': ...]

Get Log by ID

>>> log = Log.from_id('b372d92c-028d-463b-98ba-3cec3f170af5')
>>> log.id
'b372d92c-028d-463b-98ba-3cec3f170af5'
>>> log.log_type
'PROMPT'

Go to Table of Contents