Mixed Precision Loss Scale
Mixed precision loss scale issues occur when GradScaler doesn't update properly, causing underflow or overflow in FP16 training.
Mixed precision loss scale issues occur when GradScaler doesn't update properly, causing underflow or overflow in FP16 training.
What this failure is
Mixed Precision Loss Scale is a Training Stability failure seen during ML training runs. Mixed precision loss scale issues occur when GradScaler doesn't update properly, causing underflow or overflow in FP16 training. Common tags: Mixed Precision, Fp16, Gradscaler, Loss Scale.
Is this what broke your run? Paste your log.
You're reading about Mixed Precision Loss Scale. 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)
GradScaler scale too small (underflow) or too large (overflow). Inf/nan check fails silently. Loss scale not updated for new loss distribution. Static loss scale not appropriate for loss profile. 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
- FP16 training has NaN loss
- Loss is zero for many iterations
- GradScaler is not scaling loss correctly
Common symptoms and what they mean
| Symptom | Why it happens |
|---|---|
| FP16 underflow causes loss of gradient | GradScaler scale too small (underflow) or too large (overflow) |
| GradScaler scale is too small or too large | Inf/nan check fails silently |
| Loss scale never updates | Loss scale not updated for new loss distribution |
Which systems are affected
- Training with torch.cuda.amp
- Mixed precision with FP16
- Volta/Turing GPU training
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: FP16 underflow causes loss of gradient
- ✓Verified signal present: GradScaler scale is too small or too large
- ✓Verified signal present: Loss scale never updates
- ✓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
- GradScaler scale too small (underflow) or too large (overflow)
- Inf/nan check fails silently
- Loss scale not updated for new loss distribution
- Static loss scale not appropriate for loss profile
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.