AI 导读
Perplexity 通过标准化 API 暴露推理能力。 Ivy(Rust HTTP)负责 CPU 工作——解析、tokenization、模板化——并转换为 gRPC。 Tulip(Rust gRPC)为 ROSE 引擎调度批次。 ROSE(Python)实现前向传播和 CUDA graph 管理。https://t.co/GmIrKfLhGL
正文
Perplexity exposes inference via standardized APIs.
Ivy (Rust HTTP) does CPU work—parsing, tokenization, templating—and translates to gRPC.
Tulip (Rust gRPC) schedules batches for the ROSE engine.
ROSE (Python) implements forward passes and CUDA graph management. https://t.co/GmIrKfLhGL
来源:@perplexity_ai · x.com