vLLM

Tin tức, mô hình và bản phát hành AI mới nhất từ vLLM.

Mô hình 🇺🇸

Mistral AI Introduces Mistral 3 Model Family, Including Mistral Large 3

Mistral AI has announced Mistral 3, a new generation of models including dense models Ministral 3 (3B, 8B, 14B) and a sparse mixture-of-experts model Mistral Large 3 (41B active, 675B total parameters). All models are released under the Apache 2.0 license and are available on multiple platforms including Hugging Face, Amazon Bedrock, and Mistral AI Studio.

Mistral

Mistral AI

NVIDIA

Red Hat

vLLM

Mistral AI27.07 · 17:04

Phần cứng & Suy luận 🇷🇺

LLM Inference: From KV-Cache to Production Deployment

An experienced MLOps engineer from hh.ru explains how to efficiently run large language models on your own hardware in 2026. The article covers key technical aspects: KV-cache, inference engines vLLM and SGLang, the evolution of attention mechanisms, and practical recommendations for model selection, parallelism, and hardware for production.

vLLM

SGLang