LM Studio: Run Local AI Faster Than the Cloud!
- protowoerk
- Jan 29
- 2 min read
Greetings, fellow tech enthusiast, Ready to UNLEASH THE POWER of local AI? Forget the cloud—YOU are the server now! LM Studio turns your PC into an LLM-SLINGING BEAST! Faster, better, OFFLINE! Let’s get installing before AI MegaCORP finds out!
The system loadout: Intel i9-10900X (10-core, 3.7GHz), 128GB DDR5 RAM, GTX 1080 Ti (11GB VRAM), 1TB NVME—not quite a MONSTER... Test prompt? 10 SECONDS. FASTER THAN CHATGPT! Get ready to BLOW MINDS! LET’S DO THIS! BZZT!
Step 1: Downloading LM Studio
Firstly, navigate to the official site of LM Studio (https://lmstudio.ai) and choose the Windows installer, as your operating system, then click Download to start downloading the installer file (Download LM Studio for Windows 0.3.8). Once it's downloaded, run this installer on your machine to begin installing LM Studio.
Step 2: Running Deepseek Models Locally with LMStudio
Now that we've installed LM Studio, let's run a deepseek model locally. Here are the steps you need to follow:
Open LM Studio and click on "Discover" in the left-hand panel.
In the search box, type in 'deepseek', select a model. E.g.: unsloth/DeepSeek-R1-Distill-Llama-8B-GGUF/DeepSeek-R1-Distill-Llama-8B-Q8_0.gguf or an older model deepseek-coder-6.7B-instruct-GGUF with version Q8_0
Click on "Download" button in the bottom. This will start downloading the model file (~4-8GB). Should be saved to NVME drive.
Once downloaded, click on the "Use in New Chat" button in LM Studio to load this model into your workspace.
Now that we have our model loaded, we can configure it to use a GPU if available and set the number of cores that will be used by setting up resources in LMStudio. To do so:
Click on Configuration button at the top (gear icon, left side from the purple model block).
you'll see an option for CPU Thread Pool Size and GPU Offload (if your machine has a GPU). If you have one, ensure that it is maximized.
Now, you should be ready to run your model using LMStudio. Click on "Run" in the left-hand panel and start exploring with your deepseek model.
And thus it begins our exploration - a voyage through the labyrinthine world of deepseek models and LM Studio. We hope that by following these steps, you'll be able to unlock new horizons in your language processing journey.
Comments