DeepSeek

November 21st, DeepSeek announced that the preview version of its newly developed inference model DeepSeek-R1-Lite is officially online. Officially, the DeepSeek R1 series of models are trained using reinforcement learning, and the reasoning process includes a great deal of reflection and verification, and the chain of thought can be tens of thousands of words long. The series of models have achieved reasoning results comparable to OpenAI o1-preview in math, code, and a variety of complex logical reasoning tasks, and have shown users the complete thinking process that o1 has not disclosed. DeepSe...

Recently, the Chatbot Arena organized by LMSYS released the latest list update. The LMSYS Chatbot Arena ranked 11th in the overall ranking, surpassing all open source models, including Llama3-70B, Qwen2-72B, Nemotron-4-340B, Gemma2-27B, etc., and won the honor of being the first in the global open source model list. Compared with the 0507 open source Chat version, DeepSeek-V2-0628 has improved in code mathematical reasoning, command following, role playing, etc.