Alibaba Cloud announced that it will open source Tongyi Qianwen's 14 billion parameter model Qwen-14B and its dialogue model, which will be free for commercial use.

Alibaba Cloud today announced an open source project called Qwen-14B, which includes a parametric model and a conversation model. This open source project allows free commercial use

This site declares: Alibaba Cloud has previously open sourced a parameter model Qwen-7B worth 7 billion. The monthly download volume has exceeded 1 million times

Based on the data provided by Alibaba Cloud, Qwen-14B has been ranked among the best in multiple authoritative reviews. Surpassing models of the same scale, some indicators are even close to Llama2-70B.

According to reports, Qwen-14B is a high-performance open source model that supports multiple languages. Its overall training data exceeds 3 trillion Tokens, has stronger reasoning, cognition, planning and memory capabilities, and supports a maximum context window length of 8k

In contrast, Qwen-14B-Chat is The dialogue model obtained through refined SFT on the base model can bring more accurate content and is more in line with human preferences, and also expands the imagination and richness of content creation.

Starting today, users can download models directly from the Alibaba Moda community, or access and call Qwen-14B and Qwen-14B-Chat through the Alibaba Cloud Lingji platform.

Alibaba Cloud CTO Zhou Jingren said that Alibaba Cloud will continue to embrace open source and openness and promote the construction of China's large model ecosystem. Alibaba Cloud firmly believes in the power of open source and takes the lead in open source self-developed large models, hoping to bring large model technology to small and medium-sized enterprises and individual developers faster.

Model Zero Code Creation Space Experience Address:

https://modelscope.cn/stud/qwen/Qwen-14B-Chat-Demo

Environment Configuration and Installation

• python 3.8 and above

• pytorch 1.12 and above, recommended 2.0 and above

• It is recommended to use CUDA 11.4 and above (GPU users need to consider this option)

Usage steps

Note: Run under the environment configuration of PAI-DSW (you can use a single card operation, the minimum requirement of video memory is 11G)

• Qwen-14B-Chat model link:

• https://modelscope.cn/models /qwen/Qwen-14B-Chat

• Qwen-14B model link:

• ##https://modelscope. cn/models/qwen/Qwen-14B

• Qwen-14B-Chat-Int4 Model link:

• https ://www.modelscope.cn/models/qwen/Qwen-14B-Chat-Int4

The community supports direct downloading of the model’s repo:

from modelscope.hub.snapshot_download import snapshot_download<br>model_dir = snapshot_download('qwen/Qwen-14B-Chat', 'v1.0.0')Advertising Statement: The external jump links (including but not limited to hyperlinks, QR codes, passwords, etc.) contained in the article are used to convey more information and save selection time. The results are for reference only. All articles on this site include this statement.

The above is the detailed content of Alibaba Cloud announced that it will open source Tongyi Qianwen's 14 billion parameter model Qwen-14B and its dialogue model, which will be free for commercial use.. For more information, please follow other related articles on the PHP Chinese website!