Qwen-14B and the Qwen Technical Report
Qwen-14B follows Qwen-7B; the technical report (arXiv 2309.16609, 2023-09-28) documents the base, RLHF-tuned chat and tool-use design.
Second Qwen open-weights size, shortly followed by the Qwen Technical Report with 48 authors describing Qwen base models, Qwen-Chat trained with SFT and RLHF, and code and math specialists. Tongyi Qianwen license.
- Date
- Monday, 25 September 2023
- Lab
- Alibaba (Qwen)
- Kind
- open-weights
- Access
- open weights (restricted license)
Figures
| Measure | Value | Measured by |
|---|---|---|
| Technical report | arXiv 2309.16609 Submitted 2023-09-28 | company |
Date is Hugging Face repo creation for Qwen-14B (2023-09-24 UTC). The report's tool-use claim (competitive with bigger models on code-interpreter tasks) is the authors' own.
Sources
- arxiv.org/abs/2309.16609
- huggingface.co/api/models?author=Qwen&sort=createdAt&direction=1&limit=100
- github.com/QwenLM/Qwen
This record was checked against its sources on 6 October 2026. How we check