Agents read more than they write, so DeepSeek shrank what they remember. Read the technical breakdow...
DeepLearning.AI(@DeepLearningAI)136 字 (约 1 分钟)
85
DeepSeek通过减少缓存大小和提升计算效率,显著优化了模型性能。
入选理由:DeepSeek模型缓存大小降至890字节/标记,比V1版本缩小437倍
FeaturedTweet#LLMs#AIAgents#DeepSeek#模型优化英文
