Today we continue to share three latest AI open source projects!
01 MLC LLM
The MLC LLM open source project claims: “You can compile and run the AI language model on any device.”
Now, AI large models can run on MacOS, mobile phones, and even browsers!
Programmer friends who do machine learning and computer vision must have heard of MXNet and XGBoost, and must have heard of Tianqi Chen’s name.
Yes, MLC LLM is also the latest research result of Tianqi Chen and other researchers.
With the MLC LLM solution, we can deploy and run large language models on mobile devices, such as running chatbots on iPhone:
The large model runs directly on the mobile phone without requesting a server backend.
Amazing ~!
Even these AI large models can run directly in web browsers without the need for backend services!
This is the effect of running LLM:
This is the effect of running Stable Diffusion:
Open source address:
https://github.com/mlc-ai/mlc-llm
The project also provides various online demos:
You can directly experience the effects of LLM and Stabe Diffusion in the browser.
02 AudioGPT
AudioGPT can understand and generate dialogues, music, sounds, and voice-over videos.
This project integrates multiple open source projects:
Using OpenAI’s API, we can automatically call these capabilities based on input to produce content we need.
Nowadays, various digital live-streaming hosts are gradually emerging. Generating digital human content may also be a popular direction.
Project address:
https://github.com/AIGC-Audio/AudioGPT
03 Track-Anything
As the name suggests: track everything.
This is an open source project for object tracking. Clicking on the selected target will continue to be tracked and segmented, as shown below:
The project is developed based on Segment Anything, which was just open sourced. Interested friends should not miss it.
Project address: