Haibin Lin
where else to find them
linkedin:linhaibin
what they talk about
all 5 in the filters
where they keep going back
no series more than once — all 7 in the filters
on a stage with
all 65 in the filters
who talks about the same things
nobody close enough yet
| title | conference | date |
|---|---|---|
| OpGuard: Bitwise Alignment for Precise and General Debugging of Production LLM Training | 20th USENIX Symposium on Operating Systems Design and Implementation (OSDI '26) | 2026-07-13 |
Claims a practical system, OpGuard, that uses bitwise alignment—lightweight per-op fingerprints plus a schedule-tolerant mapper that finds the longest prefix… | ||
| Optimus: Accelerating Large-Scale Multi-Modal LLM Training by Bubble Exploitation | USENIX ATC '25 | 2025-07-07 |
Reports a system called Optimus that cuts end-to-end MLLM training time by ~20.5–21.3% on 3,072 GPUs (ViT-22B encoder + GPT-175B backbone) by turning idle GPU… | ||
| Understanding Stragglers in Large Model Training Using What-if Analysis | OSDI '25 | 2025-07-07 |
Five-month ByteDance cluster traces, with jobs up to >5,000 GPUs, drive a counterfactual what-if simulation that replaces computation times with the mean and… | ||
| Minder: Faulty Machine Detection for Large-scale Distributed Model Training | NSDI '25 | 2025-04-28 |
Minder is an automated faulty-machine detector deployed for over a year on thousand‑machine distributed training jobs; it builds per‑metric parametric models… | ||
| ByteCheckpoint: A Unified Checkpointing System for Large Foundation Model Development | NSDI '25 | 2025-04-28 |
ByteCheckpoint implements a parallelism-agnostic checkpoint representation and load-time resharding so checkpoints saved on one GPU layout can be read into… | ||
| ByteDance's Platform for Reinforcement Learning from Human Feedback | Ray Summit 2024 | 2024-09-30 |
ByteDance describes a production RLHF platform that runs LLM trajectory generation and training on Ray while keeping model-parallel LLM compute inside SPMD… | ||
| MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUs | Systems @Scale 2024 | 2024-06-12 |
Production system trains LLMs across >10,000 GPUs using a distributed profiler plus device-level torch profiling, fine-grained intra-node trunking (8 trunks)… | ||
| MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUs | 21st USENIX Symposium on Networked Systems Design and Implementation (NSDI '24) | 2024-04-16 |
MegaScale is a production system that trains LLMs on more than 10,000 GPUs and achieves concrete gains: 55.2% Model FLOPs Utilization when training a 175B… | ||
| nothing matches | ||