Nowadays, so many open-source AI models, such as Gemma, Mistral 7B, Llama, etc., generally need to know some programming knowledge such as python if they want to deploy them to local PCs, and now some software and open-source tools have appeared, which can already deploy AI models to local with one click, and many of them are fool-proof operations that do not require corresponding programming knowledge. The LM Studio introduced below is a desktop software for deploying open-source AI models with one click for fools.
1. What is LM Studio?
LM Studio is a desktop application that supports Windows, Apple macOS and Linux operating systems for running local large language models on computers. It uses a local large language model through an OpenAI-like HTTP server running on localhost, which gives you more control and privacy.
Key features of LM Studio:
Search and download a variety of AI large language models;
Run the AI model on a local machine with a GPU;
Integrate with AnythingLLM to create chatbot interfaces.
Official website: lmstudio.ai
2. What are the minimum computer requirements?
• Mac computer with macOS 13.6 or later (M1/M2/M3)
• Have a Windows/Linux PC that supports AVX2 processors
• Support NVIDIA/AMD GPU graphics card, 16GB or more video memory is recommended.
• For computers, 6GB+ RAM is recommended
Generally, the newer CPUs of Intel and AMD support the AVX2 instruction set, if you are unsure, you can use the CPUZ software to check the instruction set supported by your computer's CPU.
3. Install LM Studio on Windows
Enter the official website page, click on the corresponding operating system, and download.
Open the downloaded installation file and click Run
It will be automatically installed on the C drive
3. 运行LM Studio
Open LM Studio directly, and you can download the recommended models directly from the software homepage.
You can also search for models on Hugging Face on the search page. And in the upper right corner of the page, it shows the estimated memory and video memory capacity required.
5. Chat with your AI
Go to the chat page, load your downloaded model, enter a prompt, and wait for it to answer.
You can also load multiple models at once to compare different responses from different AI models.
LM Studio allows you to easily deploy your own private AI large language model, no longer need to think about the network, cost, and interact with AI anytime, anywhere, the only imperfect is the all-English interface, no Chinese function. The official subsequent version will also add text embedding (POST /v1/embeddings) support to achieve more customization functions.