最新文章

VoxCPM-1.5: A 500M-param open-source TTS model from China, offering 44.1kHz audio & 3s voice cloning on consumer GPUs. (VoxCPM-1.5：一款5亿参数开源TTS模型，支持44.1kHz音频和3秒声音克隆，可在消费级GPU运行。)

LEANN: Embedded vector DB framework for local RAG. On-demand computation & graph-based storage cut 97% space vs. traditional DBs. (LEANN：用于本地RAG的嵌入式向量数据库框架。按需计算和基于图的存储相比传统数据库节省97%空间。)

VoxCPM: 0.5B 端到端 TTS 模型，支持高效本地部署与高质量中文语音合成/克隆。 (VoxCPM: 0.5B端到端TTS模型，支持高效本地部署与高质量中文语音合成/克隆。)

LEANN AI Framework drastically cuts vector index storage to <5% of original size, enabling 50x compression for efficient similarity search. LEANN AI 框架将向量索引存储降至原体积<5%，实现50倍压缩，提升相似性搜索效率。

VoxCPM: Open-source TTS with tokenizer-free design, 0.5B params, real-time synthesis, and few-shot voice cloning. (VoxCPM：开源TTS，无分词器设计，5亿参数，实时合成，支持少样本语音克隆。)

LEANN: A lightweight, embedded vector DB for RAG, enabling local deployment with minimal resources. (LEANN：轻量级嵌入式向量数据库，为RAG应用实现低资源本地部署。)

VoxCPM: Tokenizer-free TTS for context-aware speech & zero-shot voice cloning via continuous diffusion autoregressive modeling. (VoxCPM：通过连续扩散自回归建模实现上下文感知语音和零样本语音克隆的无标记器TTS。)

Large AI models: Transformer-based, billions of parameters, self-supervised learning, fine-tunable for tasks. 大型AI模型：基于Transformer，数十亿参数，自监督学习，可针对任务微调。

Gemini's career evolution: from pro player (StarCraft) to championship coach (KPL) and top streamer. A case study in esports career longevity. (Gemini职业生涯演变：从职业选手到冠军教练及顶流主播，是电竞职业长寿的典型案例。)