nlp

Enhance Bilingual Conversations with ChatGLM3 Actions

27 Apr 2025
Enhance Bilingual Conversations with ChatGLM3 Actions

In today's globalized world, effective communication across languages is essential. The ChatGLM3 6B 32K brings a powerful solution for developers looking to integrate bilingual dialogue capabilities into their applications. This open-source model, featuring 6 billion parameters, is designed for long context dialogues, making it perfect for engaging conversations in both English and Chinese. By utilizing the ChatGLM3 actions, developers can simplify the process of creating smooth, accurate, and contextually rich dialogues, enhancing user experiences in multilingual environments.

Common use cases for the ChatGLM3 actions include customer support chatbots that can seamlessly switch between languages, educational tools that facilitate language learning, and interactive applications that require natural conversation flows. This flexibility allows businesses to cater to diverse user bases and improve engagement through personalized interactions.

Prerequisites

To get started with the ChatGLM3 actions, you'll need an API key for the Cognitive Actions service and a basic understanding of making API calls.

Engage ChatGLM3 for Bilingual Dialogue

The "Engage ChatGLM3 for Bilingual Dialogue" action allows you to leverage the ChatGLM3-6B model to engage in bilingual conversations, providing a robust tool for developers focusing on natural language processing (NLP).

Purpose

This action is designed to facilitate smooth and accurate conversations in both English and Chinese, addressing the need for effective communication in multilingual settings. It supports multi-turn dialogues, tool invocation, and complex agent tasks, enabling developers to create dynamic and interactive applications.

Input Requirements

The action requires a structured input format defined by the following properties:

  • topP: A number (default 0.8) that determines the nucleus sampling parameter, influencing the variety of generated responses.
  • prompt: A string that serves as the initial input for the model, which must follow a specific format.
  • maxTokens: An integer (default 2048) indicating the maximum number of tokens to generate in the response.
  • temperature: A number (default 0.75) that controls the randomness of the output, affecting the diversity of the generated text.

Example Input:

{
  "topP": 0.8,
  "prompt": "<|system|>\nYou are a helpful assistant\n<|user|>\n请使用英文重复这段话：\"为了使模型生成最优输出，当使用 ChatGLM3-6B 时需要使用特定的输入格式，请按照示例格式组织输入。\"\n<|assistant|>\n",
  "maxTokens": 2048,
  "temperature": 0.75
}

Expected Output

The output will be a sequence of tokens that form a coherent response based on the input prompt. For example, the model might return:

[
  "\"",
  "In",
  " order",
  " to",
  " obtain",
  " the",
  " optimal",
  " output",
  " from",
  " the",
  " model",
  ",",
  " when",
  " using",
  " Chat",
  "GL",
  "M",
  3,
  "-",
  6,
  "B",
  ",",
  " you",
  " need",
  " to",
  " use",
  " specific",
  " input",
  " formats",
  ".",
  " Please",
  " organize",
  " the",
  " input",
  " according",
  " to",
  " the",
  " following",
  " example",
  " format",
  ".\""
]

Use Cases for this Specific Action

This action is particularly useful in scenarios where:

  • Customer Support: Develop chatbots that can handle inquiries in both English and Chinese, providing a seamless experience for users.
  • Language Learning Apps: Create interactive tools that help users practice bilingual conversations, making learning engaging and effective.
  • Multilingual Virtual Assistants: Build assistants that can switch languages based on user preference, enhancing accessibility for diverse audiences.

```python
import requests
import json

# Replace with your actual Cognitive Actions API key and endpoint
# Ensure your environment securely handles the API key
COGNITIVE_ACTIONS_API_KEY = "YOUR_COGNITIVE_ACTIONS_API_KEY"
# This endpoint URL is hypothetical and should be documented for users
COGNITIVE_ACTIONS_EXECUTE_URL = "https://api.cognitiveactions.com/actions/execute"

action_id = "d6fdbf4a-aaff-4d5b-b34a-e7c319083334" # Action ID for: Engage ChatGLM3 for Bilingual Dialogue

# Construct the exact input payload based on the action's requirements
# This example uses the predefined example_input for this action:
payload = {
  "topP": 0.8,
  "prompt": "<|system|>\nYou are a helpful assistant\n<|user|>\n请使用英文重复这段话：\"为了使模型生成最优输出，当使用 ChatGLM3-6B 时需要使用特定的输入格式，请按照示例格式组织输入。\"\n<|assistant|>\n",
  "maxTokens": 2048,
  "temperature": 0.75
}

headers = {
    "Authorization": f"Bearer {COGNITIVE_ACTIONS_API_KEY}",
    "Content-Type": "application/json",
    # Add any other required headers for the Cognitive Actions API
}

# Prepare the request body for the hypothetical execution endpoint
request_body = {
    "action_id": action_id,
    "inputs": payload
}

print(f"--- Calling Cognitive Action: {action.name or action_id} ---")
print(f"Endpoint: {COGNITIVE_ACTIONS_EXECUTE_URL}")
print(f"Action ID: {action_id}")
print("Payload being sent:")
print(json.dumps(request_body, indent=2))
print("------------------------------------------------")

try:
    response = requests.post(
        COGNITIVE_ACTIONS_EXECUTE_URL,
        headers=headers,
        json=request_body
    )
    response.raise_for_status() # Raise an exception for bad status codes (4xx or 5xx)

    result = response.json()
    print("Action executed successfully. Result:")
    print(json.dumps(result, indent=2))

except requests.exceptions.RequestException as e:
    print(f"Error executing action {action_id}: {e}")
    if e.response is not None:
        print(f"Response status: {e.response.status_code}")
        try:
            print(f"Response body: {e.response.json()}")
        except json.JSONDecodeError:
            print(f"Response body (non-JSON): {e.response.text}")
    print("------------------------------------------------")



## Conclusion
The ChatGLM3 actions provide developers with a powerful tool to create bilingual conversational applications, enhancing user engagement and accessibility. With its support for multi-turn dialogues and complex tasks, you can build applications that cater to diverse linguistic needs. As you explore the capabilities of the ChatGLM3 actions, consider the various use cases that can benefit from this technology, and start integrating these features to elevate your applications.
← Back to All Articles