Memory Leak in DataLoader
Memory leaks in DataLoader workers cause steadily growing memory usage over training epochs.
Memory leaks in DataLoader workers cause steadily growing memory usage over training epochs.
What this failure is
Memory Leak in DataLoader is a Memory failure seen during ML training runs. Memory leaks in DataLoader workers cause steadily growing memory usage over training epochs. Common tags: Dataloader, Memory Leak, Worker, Num Workers.
Is this what broke your run? Paste your log.
You're reading about Memory Leak in DataLoader. Paste your own crash log or traceback below and get the real root cause for YOUR run, not this generic entry. No account, no card. Logs are masked at ingress and never saved to account history.
Want 14 days on the Scale plan?
Request an evaluation code. A verified workplace organization activates up to 50 diagnoses a day, alerts, history, and follow-up questions. No credit card or automatic subscription.
Why it happens (the mechanism)
Custom Dataset holds references to data. Worker process memory not released between epochs. Global state in worker functions. Cuda memory not freed in workers. Taken together, these mechanisms explain why the failure is reproducible, why it tends to surface on specific workloads or scales, and why generic mitigation attempts often fall short without addressing the underlying cause.
What you'll observe
- Memory usage grows with each epoch
- Worker processes use more memory over time
- OOM after several epochs of training
Common symptoms and what they mean
| Symptom | Why it happens |
|---|---|
| htop shows worker processes growing | Custom Dataset holds references to data |
| Memory grows linearly with iterations | Worker process memory not released between epochs |
| OOM after long training run | Global state in worker functions |
Which systems are affected
- PyTorch DataLoader with num_workers > 0
- Training with custom Dataset classes
- Multi-worker data preprocessing
How to confirm this is the problem
Use this checklist to test the hypothesis against a small reproduction. No single line proves the root cause, so preserve the preceding events and compare one variable at a time.
- ✓Reproduce the failure from a clean checkpoint/seed: the symptom must appear without warm-up state from a previous run.
- ✓Verified signal present: htop shows worker processes growing
- ✓Verified signal present: Memory grows linearly with iterations
- ✓Verified signal present: OOM after long training run
- ✓A targeted fix from the "How to fix it" section eliminates or substantially reduces the symptom within one validation pass.
The fix and the prevention pattern
The root cause is on this page and stays free. A free account adds the exact remediation steps, saved history, and the fix on every entry in the encyclopedia.
Sign up free. Unlock the full analysisNo credit card. Daily allowance follows verified trust tier. Instant access.
Diagnose this failure in VS Code
Select the traceback or open the failed terminal, then run Denpex locally to see the initiating rank, collateral failures, exact fix, and verification command without uploading the log.
Install the free VS Code extensionRelated failures to investigate next
Root cause
- Custom Dataset holds references to data
- Worker process memory not released between epochs
- Global state in worker functions
- Cuda memory not freed in workers
The fix and how to prevent it
Evaluate Denpex on your own logs
Request a Scale evaluation code. A verified workplace organization activates 14 days with up to 50 diagnoses a day. Every account keeps its current diagnosis allowance and gets a verification path. No card or automatic subscription.
Don't just read the fix, diagnose your run
The encyclopedia tells you what went wrong. Denpex tells you what went wrong in YOUR training run. With your logs, your config, and your stack.