The Flask Application that uses the Mukalma models to run it on a Flask Web Server. This allows the model to be used for Testing and our Front-End Application. This document details the requests and the response strucuture of the API, using HTTP GET and POST Requests.
Follow the Installtion instructions on the main README to learn how to install the API
Use this request to get a reply to a message from the model.
| Path | <API_URL>/reply |
| Type | HTTP POST |
| Request Body | JSON |
Request Parameters
| Parameter | Type | Description |
|---|---|---|
message |
String |
Your text message for which we need the reply |
async |
Boolean |
Toggle API request method. Please use False for this when connecting from anything other than Front-End |
m_id |
Integer |
Message ID. Can be sent as 0. Only used by Front-End |
Response Structure
| Key | Value | Description |
|---|---|---|
response |
String |
Response to the message |
candidates |
Array |
Candidate responses from the model |
knowledge_sent |
String |
Knowledge span used for the response |
knowledge_source |
String |
Knowledge source used for the response |
topic |
Object |
Dictionary containing information about the topic used by the model |
Use this request to test your connection to the API.
| Path | <API_URL>/connect |
| Type | HTTP GET |
Use this request to clear the current context of the model.
| Path | <API_URL>/connect |
| Type | HTTP GET |
GNU General Public License v3.0