← 研踪 PaperTrailarXiv:2605.31463 · v2
PithTrain: A Compact and Agent-Native MoE Training System
首次提交 2026-05-29 23:52
版本更新 2026-11-07 22:51
雷达观测 2026-11-10 11:14
摘要
Mixture-of-Experts (MoE) has become the dominant architecture for frontier language models. To meet this demand, production frameworks have built optimized MoE training stacks over years of engineering effort. Yet evolving these stacks for new architectures and system optimizations remains expensive. With the rise of AI coding agents, they could automate parts of training-framework development and accelerate this evolution. But applying them to these existing frameworks carries hidden costs, invisible to today's throughput-only evaluations. We name this missing dimension agent-task efficiency (ATE): the cost of using coding agents to understand, operate, and extend a framework. Grounded in four agent-native design principles, we build PithTrain, a compact, agent-native MoE training framework supporting pretraining, supervised fine-tuning, and reinforcement learning. We further introduce ATE-Bench, covering real-world training-framework tasks. Our evaluation shows PithTrain achieves throughput competitive with production frameworks and higher agent-task efficiency on ATE-Bench. On new-feature tasks, PithTrain uses up to 62% fewer Agent Turns than Megatron-LM, and up to 44% and 64% less Active GPU Time than Megatron-LM and TorchTitan, respectively.
作者备注：NeurIPS 2026
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-07 22:51 · 最新观测版本
PithTrain: A Compact and Agent-Native MoE Training System
Mixture-of-Experts (MoE) has become the dominant architecture for frontier language models. To meet this demand, production frameworks have built optimized MoE training stacks over years of engineering effort. Yet evolving these stacks for new architectures and system optimizations remains expensive. With the rise of AI coding agents, they could automate parts of training-framework development and accelerate this evolution. But applying them to these existing frameworks carries hidden costs, invisible to today's throughput-only evaluations. We name this missing dimension agent-task efficiency (ATE): the cost of using coding agents to understand, operate, and extend a framework. Grounded in four agent-native design principles, we build PithTrain, a compact, agent-native MoE training framework supporting pretraining, supervised fine-tuning, and reinforcement learning. We further introduce ATE-Bench, covering real-world training-framework tasks. Our evaluation shows PithTrain achieves throughput competitive with production frameworks and higher agent-task efficiency on ATE-Bench. On new-feature tasks, PithTrain uses up to 62% fewer Agent Turns than Megatron-LM, and up to 44% and 64% less Active GPU Time than Megatron-LM and TorchTitan, respectively.
查看此版本解读 →查看此版本原文 ↗