Spend 1 minute every day to get selected cutting-edge AI information.
The content covers but is not limited to cutting-edge AI news, AI tools, AI painting, open-source projects, and learning tutorials, etc.
Follow AI Daily to stay up-to-date with AI trends. For important information, separate posts will be made for detailed introductions.
Here is the latest AI information for July 24.
Cutting-edge News
1. The world's largest open-source large language model Llama 3.1 405B is officially released!
Meta AI officially launched the Llama 3.1 series models, including three specifications: 8B, 70B, and 405B, with 405B being the largest parameter open-source large language model in the world.
All models have a context length of 128K and perform excellently in multiple benchmarks, even surpassing closed models such as GPT-4, GPT-4o, and Claude 3.5 Sonnet.
Detailed introduction: https://ai.meta.com/blog/meta-llama-3-1/
Model download: https://huggingface.co/collections/meta-llama/llama-31-669fc079a0c406a149a5738f
In summary, the Llama 3.1 405B model has become the most powerful open-source model currently available.
If you are interested, you can experience it on HuggingChat.
Address: https://huggingface.co/chat/
2. Keling is starting to charge!
The minimum is 66 RMB/month, currently with a 50% discount at 33/month, which can generate 3300 images or 66 videos.
Ordinary users are given 66 inspiration points for logging in every day. Each video consumes 10 inspiration points, and unused points are automatically cleared daily.
Official website: https://klingai.kuaishou.com/
At the same time, the international version of Keling, Kling 1.0, is officially online!
Any email address can register without binding a mobile phone number. 66 inspiration points are given for logging in daily, so hurry up if you want to get it for free!
Official website: https://klingai.com/
3. OpenAI launches GPT-4o mini fine-tuning feature, free for a limited time!
This feature is currently open to level 4 and level 5 users and is planned to be gradually opened to all users. Before September 23, 2 million training tokens can be used for free every day.
The part exceeding 2 million training tokens will be charged at $3.00 per million tokens. From September 24, fine-tuning training will be charged at $3.00 per million tokens.
Documentation: https://platform.openai.com/docs/guides/fine-tuning
It has been observed multiple times that whenever OpenAI's competitors release products, OpenAI tends to make some moves～
Open Source Projects
1. A free, open-source, simple, and easy-to-use Mac software uninstallation tool Pearcleaner.
The tool is only 4MB in size, offering a high-value, simple, and easy-to-use interface. It can select and uninstall unnecessary software in bulk, and also clean up residual files after uninstallation very thoroughly.
GitHub: https://github.com/alienator88/Pearcleaner
It can be used as an open-source alternative to AppCleaner, and everyone might as well give it a try.
2. A free, open-source drawing tool draw.io.
It can be used to create a series of diagrams, illustrations, or graphs, including flowcharts, UML class diagrams, organizational charts, swimlane diagrams, etc., suitable for various complex professional diagrams.
GitHub: https://github.com/jgraph/drawio
Compared to ProcessOn, draw.io has richer drawing elements, and files can be saved to your cloud drive or locally, making data privacy more secure.
Additionally, a foreign developer has integrated it into VSCode and packaged it as a plugin for developers to download and use, allowing you to quickly draw perfect flowcharts, mind maps, or UML class diagrams in VSCode.
GitHub: https://github.com/hediet/vscode-drawio
3. A full-stack SaaS development template based on Cloudflare cloudflare-saas-stack.
Using this template, you can quickly build and deploy full-stack applications on Cloudflare, including database, authentication, styling, storage, and other functions.
GitHub: https://github.com/Dhravya/cloudflare-saas-stack
The author built the SuperMemory application based on this tool, which was introduced in detail before. If you are interested, you can check it out.
Detailed introduction: https://t.zsxq.com/19tFQq4ye
