Skip to main content
版本：Next

ChatGPT

ChatGPT Source Connector​

Configuration​

Before using ChatGPT source connector, you need to configure the server.

Startup​

  1. start EventMesh Runtime
  2. start eventmesh-connector-chatgpt

When finished, the ChatGPT source connector will act as an HTTP server.

Sending messages​

You can send messages to the source connector via HTTP.

connectorConfig:
  connectorName: chatgptSource
  path: /chatgpt
  port: 3756
  idleTimeout: 0
  proxyEnable: false
  parsePromptFileName: prompt
openaiConfig:
  token:
  model: gpt-3.5-turbo
  timeout: 0
  temperature: 1
  maxTokens:
  frequencyPenalty: 0
  presencePenalty: 0
  user: eventMesh
  stop: []
  logitBias: {}

openaiProxyConfig:
  host: 127.0.0.1
  port: 7890

The above example configures a URL http://localhost:3756/chatgpt in source-config.yml.

You can send messages in CHAT or PARSE request type, default CHAT,

CHAT​

The request type of "CHAT" is a conversation with ChatGPT

  • Construct a request body
    • requestType, default is CHAT.
    • type, default is cloudevents.
    • source, default is /.
    • subject, default is chatGPT.
    • datacontenttype, default is text/plain.
    • (required) text, you want to talk to ChatGPT about.

For example:

curl --location --request POST 'http://localhost:3756/chatgpt' \
--data-raw '{
    "requestType": "CHAR",
    "type": "com.example.someevent",
    "source": "/mycontext",
    "subject":"test_topic",
    "datacontenttype":"text/plain",
    "text": "can you tell me a story."
}'

PARSE​

The request type of "PARSE" is a output parse and parse result that connector will get from ChatGPT.

  • Construct a request body
    • (required)requestType, the value must be PARSE.
    • type, default is cloudevents.
    • source, default is /.
    • subject, default is chatGPT.
    • datacontenttype, default is application/json.
    • (required) text, unstructured data. e.g.,an article or a description.
    • (required) fields, field info, chatGPT according to field info Extract information from text.

For example:

curl --location --request POST 'http://localhost:3756/chatgpt' \
--data-raw '{
    "requestType": "PARSE",
    "type": "com.example.someevent",
    "source": "/mycontext",
    "subject":"test_topic",
    "datacontenttype":"application/json",
    "text": "This leaf blower is pretty amazing.  It has four settings: candle blower, gentle breeze, windy city, and tornado. It arrived in two days, just in time for my wife's anniversary present. I think my wife liked it so much she was speechless. So far I've been the only one using it, and I've been using it every other morning to clear the leaves on our lawn. It's slightly more expensive than the other leaf blowers out there, but I think it's worth it for the extra features.",
    "fields": "gift:Was the item purchased as a gift for someone else? Answer True if yes, False if not or unknown;delivery_days:How many days did it take for the product to arrive? If this information is not found, output -1;price_value:Extract any sentences about the value or price, and output them as a comma separated Python list"
}'

If datacontenttype is application/json, ChatGPT result:

{
  "gift": false,
  "delivery_days": 2,
  "price_value": ["It's slightly more expensive than the other leaf blowers out there, but I think it's worth it for the extra features."]
}

If datacontenttype is application/xml, The root node of xml is fixed as <root></root>, ChatGPT result:

<data>
  <gift>False</gift>
  <delivery_days>2</delivery_days>
  <price_value>["It's slightly more expensive than the other leaf blowers out there"]</price_value>
</data>