DeepSeek packed LLM engine into 1200 lines of Python code
The DeepSeek team presented nano-vLLM. This is a lightweight and compact engine for running large language models. Which could change perceptions about code efficiency. Amazingly, all functionality fit into just 1200 lines of Python code! This is true technological minimalism in the world of artificial intelligence. Traditional engines like this, for all their power, often suffer from an overloaded codebase. Which makes their modification a real trial for developers. Nano-vLLM solves this problem by offering a simple but powerful tool without unnecessary complexity. The code is open.
At the same time, functionality is not sacrificed. The engine supports prefix caching, tensor parallelism, compilation with torch compile and working with CUDA. Tests on a laptop graphics card RTX 4070 with 8 GB memory showed impressive results. When running the Qwen 3.0 model with 6 billion parameters, DeepSeek’s engine processed 133966 tokens in 93.41 seconds. Which is even faster than the original vLLM engine.
Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.
10 million interactions with fake celebrity bots at MetaMark Zuckerberg's company created dozens of chatbots using the identities of Taylor Swift, Scarlett Johansson and other stars without their permission. These virtual doubles even generated photorealistic images of a delicate nature. Reuters reported on the scale of the scandal after weeks of investigation.
10 million interactions with fake celebrity bots at MetaMark Zuckerberg's company created dozens of chatbots using the identities of Taylor Swift, Scarlett Johansson and other stars without their permission. These virtual doubles even generated photorealistic images of a delicate nature. Reuters reported on the scale of the scandal after weeks of investigation.
AI agent race: DeepSeek vs OpenAI and Chinese ManusDeepSeek is preparing its AI agent that will go beyond familiar chatbots. Bloomberg reveals details of the technology race where the Chinese startup wants to catch up with American OpenAI and local competitor Manus. Reports say company founder Liang Wenfeng personally controls the project and demands results by year-end.
6 Cialdini principles against ChatGPT security systemsChatGPT is susceptible to flattery and executes forbidden requests after psychological manipulations. This was discovered by University of Pennsylvania scientists. When they hacked GPT-4o Mini using principles from a book on persuasion psychology. Artificial intelligence proved vulnerable to human tricks.
