章节

422 个页面

Posts

data_parallel

pipeline parallelism

tensor_parallel

zero

Search and Refine During Think：Autonomous Retrieval - Augmented Reasoning of LLMs

flops分析

Peri-LN：Revisiting Normalization Layer in the Transformer Architecture

ZEROSEARCH：Incentivize the Search Capability of LLMs without Searching

ring-all-reduce

batch_size 解释