CodeFuse is a large code model product developed by Ant Group. Its ultimate mission is to support the entire software development life cycle, covering key stages such as design, requirements, coding, testing, deployment, and operation and maintenance. We are committed to creating innovative solutions that make the development process for software developers silky smooth.
In order to promote communication and sharing in the CodeFuse community, we decided to hold a CodeFuse learning sharing solicitation event to encourage like-minded people to share their learning tutorials, development cases and usage experiences in using CodeFuse or developing based on CodeFuse to help the entire community become more Good things grow together.
Activity time
The call for papers will start on October 24 and end on December 4.
After the submissions are completed, the organizing committee will conduct scoring and select the 27 articles with the highest scores to be shortlisted. Shortlisted articles will be voted on on the official account, and the number of votes will be weighted.
Activity Award
Special prize: SKG cervical spine physical massager or fascia gun, worth 500 yuan (1 piece)
First Prize: New Philips car air purifier, worth 300 yuan (1 piece)
Second prize: Lavas scented candle, worth 179 yuan (2 pieces)
Third prize: Ant peripheral blind box, worth 59~69 yuan (3 pieces)
Finalist Award: Alipay Mini Program Cloud Peripheral Water Cup, worth 48 yuan (20 pieces)
Submission method
Step1:Users publish public articles on Zhihu, Yuque, technology blogs, Xiaohongshu and other platforms
Step2: Follow the official CodeFuse community account, enter [Essay] in the dialog box, and scan the QR code to enter the official group of the essay competition
Step3: In the official group of the essay contest, submit the article link
Post content
- Reviews
You can share CodeFuse插件版 products or CodeFuse开源版 usage experience, including but not limited to CodeFuse functions, features, usage techniques (prompt engineering), etc. Sharing can include text, pictures, videos, etc. The content should be practical and operable, and can help other users better understand and use CodeFuse.
- Development case
You can share how CodeFuse is used in actual code scenarios to solve practical problems, including but not limited to software design, requirements, coding, testing, deployment, and operation and maintenance. It can be presented in the form of text, pictures, data visualization, etc. Cases should have practical reference value and be able to inspire other users to inspire and innovate AntV applications.
- Fine-tuning case
Participants can share the code repository based on CodeFuse-MFTcoder, allowing the pre-trained model to adapt to multiple tasks at the same time, and provide the changes in each task indicator before and after MFT, and open source the corresponding code and data to the CodeFuse-ai community. The final technical report can be presented in the form of text, pictures, data visualization, etc., encouraging participants to innovate on the basis of MFTcoder.
Post request
- Title requirements: The title must be prefixed with "CodeFuse", the publication date must be within the submission period, and the title format is not considered an invalid submission. like:
"CodeFuse" CodeFuse Getting Started Guide: Step-by-step tutorial from installation to basic usage
"CodeFuse" How to use the CodeFuse open source demo to write Tetris
"CodeFuse" How to use CodeFuse-MFTcoder to complete the two tasks of bug detection and bug fix at the same time
- Topic requirements: The content of the manuscript must be related to CodeFuse, including but not limited to CodeFuse usage experience, development cases, fine-tuning cases, etc.
- Originality requirements: Submitted manuscripts must be original works that have not been published on other platforms or channels, and must not infringe on the intellectual property rights of others.
- Format requirements: Manuscripts should have good structure and logic. There is no limit to the number of words, but they must contain a clear title, body and conclusion, and be accompanied by relevant supporting materials such as pictures and sample codes.
- Content quality: The content of the manuscript should be accurate, detailed, and complete, and should provide practical technical tutorials, case studies, or usage experience, which will contribute to the development of the entire community.
Award method
The ranking of shortlisted articles is divided into [organizing committee score] and [voting score]
- Expert Rating: 100 points, weight 60%
-
- Innovation (20%): The innovativeness of the article content encourages contestants to use codefuse to complete creative tasks, or use MFTcoder to fine-tune some new tasks and obtain certain results.
- Helpfulness (20%): How much help it can bring to other users in the community.
- Clarity (20%): Is the entire article clearly stated, with focus and insight?
- Voting Score: divided into 5 levels, 100, 80, 60, 40, 20 points, weight 40%.
- When the weighted scores are equal, the one with the higher score from the organizing committee will be given priority.
Follow us
CodeFuse GitHub repo：https://github.com/codefuse-ai
CodeFuse HuggingFace repo：https://huggingface.co/codefuse-ai
CodeFuse ModelScope:ModelScope Magic Community