Enterprise Plans
AI Application Development
AI Translator API
AI SEO/GEO Service
Web Scraping Service
Log in
const { OpenAI } = require('openai');
const api = new OpenAI({
baseURL: 'https://api.ai.cc/v1',
apiKey: '',
});
const main = async () => {
const result = await api.chat.completions.create({
model: 'Qwen/Qwen1.5-72B-Chat',
messages: [
{
role: 'system',
content: 'You are an AI assistant who knows everything.',
},
{
role: 'user',
content: 'Tell me, why is the sky blue?'
}
],
});
const message = result.choices[0].message.content;
console.log(`Assistant: ${message}`);
};
main();
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.ai.cc/v1",
api_key="",
)
response = client.chat.completions.create(
model="Qwen/Qwen1.5-72B-Chat",
messages=[
{
"role": "system",
"content": "You are an AI assistant who knows everything.",
},
{
"role": "user",
"content": "Tell me, why is the sky blue?"
},
],
)
message = response.choices[0].message.content
print(f"Assistant: {message}")
-
AI Playground
Test all API models in the sandbox environment before you integrate.
We provide more than 300 models to integrate into your app.
.webp)
Product Detail
Unveiling Qwen 1.5 Chat (72B): The Next-Gen Language Model
Introducing Qwen1.5-72B-Chat, the highly anticipated beta version of Qwen2. This advanced transformer-based, decoder-only language model, developed by researchers at Alibaba Cloud, represents a significant leap forward in AI capabilities. It delivers substantial improvements over its predecessor, Qwen, particularly in aligning with human preferences for chat models.
Key enhancements include robust multilingual support and stable handling of an expansive 32K context length, making it a versatile tool for a wide range of natural language processing tasks.
✨ The Sophisticated Architecture of Qwen1.5-72B-Chat
Qwen1.5-72B-Chat is a finely tuned version of the base Qwen1.5-72B model, specifically recommended for all conversational and instruction-following applications. It builds upon the proven Transformer architecture, integrating state-of-the-art features such as SwiGLU activation, attention QKV bias, and efficient group query attention.
The model also incorporates a mixture of sliding window attention and full attention for optimized performance. Crucially, it features an improved tokenizer designed to be highly adaptive to a multitude of natural languages and various coding structures, significantly boosting its global applicability.
💡 Key Use Cases for Qwen 1.5 Chat (72B)
✅ Superior Text Generation Capabilities
Leveraging its large scale and impressive language generation prowess, Qwen1.5-72B-Chat is excellently suited for developing advanced chatbots. It facilitates engaging and contextually aware conversations across diverse topics, providing a natural and human-like interaction experience.
🔍 Retrieval Augmented Generation (RAG) and Function Calling
This model demonstrates strong applicability for Retrieval-Augmented Generation (RAG) setups, effectively mitigating issues such as hallucination and limitations due to real-time data. It consistently showcases robust performance in RAG benchmarks, surpassing Llama-2-70B and Mixtral 8x7b on many metrics.
Moreover, Qwen1.5-72B-Chat excels in function calling tasks (tool-use benchmark), precisely selecting and utilizing external tools, making it an optimal choice for AI agents that require seamless interaction with external systems.
🛡️ Efficient Content Moderation
Harness Qwen1.5-72B-Chat's profound language understanding for highly effective content moderation. Its extensive analytical capacity enables accurate detection and filtering of inappropriate or harmful content across various platforms and applications.
🌐 Broad Multilingual Application Support
With remarkable multilingual capabilities, Qwen1.5-72B-Chat is perfectly suited for global tasks, catering to a diverse spectrum of language needs. Evaluated across 12 different languages, including French, Spanish, and Japanese, it consistently delivers strong performance, making it indispensable for applications requiring extensive linguistic support.
🚀 Qwen 1.5 Chat (72B): Outperforming the Competition
Qwen1.5-72B-Chat has been rigorously benchmarked against leading competitors such as Llama-2-70B and GPT-4, consistently demonstrating promising instruction-following capabilities.
The model has achieved notably strong performance in aligning with human preferences on the two common benchmarks: MT-Bench and Alpaca-Eval. Its stable support for a large 32K context length further enhances its competitive advantage.
🛠️ Getting Started & Licensing for Qwen1.5-72B-Chat
Integration & Tips for Implementation
You can seamlessly integrate Qwen1.5-72B-Chat into your applications by signing up for AI/ML API access directly on this website.
For local evaluation, it's recommended to install the latest Huggingface transformers library (version >= 4.37.0). Pay close attention to the provided hyper-parameters in generation_config.json to optimize performance and address specific issues like code switching. For more details, consult the model's repository.
Understanding the License Agreement
The Qwen1.5-72B-Chat model is governed by the Tongyi Qianwen license agreement. Commercial use is permissible without submitting a request, unless your product or service exceeds 100 million monthly active users. For the comprehensive agreement, visit the model's repository on GitHub or Huggingface.
🎯 Conclusion: A Landmark in Open-Source AI
In essence, Qwen1.5-72B-Chat marks a monumental advancement in the realm of open-source transformer-based language models. Its significantly improved performance and unparalleled versatility across diverse natural language processing tasks establish a robust foundation for future innovations, heralding even greater capabilities in AI-driven language processing.
❓ Frequently Asked Questions (FAQ)
Q1: What makes Qwen1.5-72B-Chat a significant advancement?
A1: It's the beta version of Qwen2 by Alibaba Cloud, offering enhanced human preference alignment for chat, extensive multilingual support, and a stable 32K context length, significantly improving upon its predecessor.
Q2: What are the core applications of this model?
A2: Its primary uses include building advanced chatbots, Retrieval-Augmented Generation (RAG), function calling for intelligent AI agents, robust content moderation, and various multilingual tasks.
Q3: How does Qwen1.5-72B-Chat compare to other leading language models?
A3: It exhibits strong instruction-following and human preference alignment, often outperforming models like Llama-2-70B and Mixtral 8x7b in RAG benchmarks, and compares favorably with GPT-4 in instruction capabilities.
Q4: Is it suitable for projects requiring multilingual support?
A4: Yes, absolutely. With support for 12 diverse languages, Qwen1.5-72B-Chat is exceptionally well-suited for multilingual communication, translation, and content understanding across various linguistic contexts.
Q5: What are the terms for commercial use of Qwen1.5-72B-Chat?
A5: The model is under the Tongyi Qianwen license. Commercial use is generally permitted without a specific request, unless your product or service reaches more than 100 million monthly active users, at which point a request becomes necessary.
Learn how you can transformyour company with AICC APIs
