ScienceOne-AI/DeepSeek-671B-SFT-Guide
An open-source solution for full parameter fine-tuning of DeepSeek-V3/R1 671B, including complete code and scripts from training to inference, as well as some practical experiences and conclusions. (DeepSeek-V3/R1 满血版 671B 全参数微调的开源解决方案，包含从训练到推理的完整代码和脚本，以及实践中积累一些经验和结论。)
This project offers a comprehensive guide and tools for developers and machine learning engineers to fine-tune the DeepSeek-V3/R1 671B language model. It provides all necessary code and scripts, from setting up the training environment to performing the actual training and inference. The project takes raw text data, formatted with specific roles and optional reasoning content, and produces a specialized, fine-tuned DeepSeek model ready for deployment.
796 stars. No commits in the last 6 months.
Use this if you are an AI/ML engineer or researcher working with large language models and need to perform full parameter fine-tuning on the DeepSeek-V3/R1 671B model, especially when dealing with complex reasoning data.
Not ideal if you are an end-user looking for a pre-trained model to use directly, or if you lack the extensive computational resources and expertise in distributed machine learning required for such a large-scale fine-tuning task.
Stars
796
Forks
96
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 13, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/llm-tools/ScienceOne-AI/DeepSeek-671B-SFT-Guide"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
LLM-Red-Team/metaso-free-api
🚀 秘塔AI搜索逆向API【特长：超强检索超长输出】，支持高速流式输出、超强联网搜索（全网or学术以及简洁、深入、研究三种模式），零配置部署，多路token支持，仅供测试，如需商用请前往官方开放平台。
MaxiDonkey/DelphiDeepseek
The Deepseek API wrapper for Delphi leverages Deepseek’s advanced models to deliver powerful...
LLM-Red-Team/deepseek-free-api
🚀 DeepSeek-V3 &...
deepseek-php/deepseek-laravel
Laravel wrapper for Deepseek PHP client, to seamless deepseek API integration with laravel applications.
FareedKhan-dev/train-deepseek-r1
Building DeepSeek R1 from Scratch