Minjia Zhang
where else to find them
no profile links on file
what they talk about
where they keep going back
no series more than once — all 4 in the filters
on a stage with
all 25 in the filters
who talks about the same things
nobody close enough yet
| title | conference | date |
|---|---|---|
| Universal Checkpointing: A Flexible and Efficient Distributed Checkpointing... | USENIX ATC '25 | 2025-07-07 |
Universal Checkpointing decouples on-disk state from model parallel layout by converting checkpoints into an atomic checkpoint format that stores parameters… | ||
| Parcae: Proactive, Liveput-Optimized DNN Training on Preemptible Instances | 21st USENIX Symposium on Networked Systems Design and Implementation (NSDI '24) | 2024-04-16 |
Parcae is a system that proactively adapts hybrid data+pipeline parallelization to predicted spot-instance changes, using a liveput metric that balances… | ||
| Bamboo: Making Preemptible Instances Resilient for Affordable Training of Large DNNs | NSDI 2023 | 2023-04-17 |
Bamboo makes preemptible (spot) instances usable for large DNN training by building a distributed system that inserts redundant computations into idle pipeline… | ||
| ZeRO-Offload: Democratizing Billion-Scale Model Training | USENIX ATC '21 | 2021-07-14 |
ZeRO-Offload offloads fp16 gradients and fp32 optimizer states and runs parameter updates on CPU while keeping forward/backward on GPU, enabling 30 billion… | ||
| nothing matches | ||