Skip to content

EMA Decay Too High

EMA (Exponential Moving Average) decay values that are too high or too low cause poor model averaging.

Quick answer

EMA (Exponential Moving Average) decay values that are too high or too low cause poor model averaging.

Training Stability#ema#exponential-moving-average#model-averaging#training-stability#generalization

What this failure is

EMA Decay Too High is a Training Stability failure seen during ML training runs. EMA (Exponential Moving Average) decay values that are too high or too low cause poor model averaging. Common tags: Ema, Exponential Moving Average, Model Averaging, Training Stability.

Live diagnosis, no signup

Is this what broke your run? Paste your log.

You're reading about EMA Decay Too High. 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.

training_logs.txt
No log to hand? Try one:

3 free diagnoses/day

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.

Evaluate one incident

Why it happens (the mechanism)

EMA decay too high (e.g., 0.9999) means EMA updates too slowly. EMA decay too low (e.g., 0.9) means EMA tracks noisy training. EMA momentum not tuned for training duration. EMA not applied to all parameters. 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

  • EMA model is too slow to track training model
  • EMA model is too noisy
  • EMA model diverges from training model

Common symptoms and what they mean

SymptomWhy it happens
EMA model parameters are very different from training modelEMA decay too high (e.g., 0.9999) means EMA updates too slowly
EMA model doesn't improve generalizationEMA decay too low (e.g., 0.9) means EMA tracks noisy training
EMA model is identical to initial modelEMA momentum not tuned for training duration

Which systems are affected

  • Training with EMA for better generalization
  • Fine-tuning with EMA
  • Large model training with weight averaging

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: EMA model parameters are very different from training model
  • Verified signal present: EMA model doesn't improve generalization
  • Verified signal present: EMA model is identical to initial model
  • 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 analysis

No 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 extension

Root cause

  • EMA decay too high (e.g., 0.9999) means EMA updates too slowly
  • EMA decay too low (e.g., 0.9) means EMA tracks noisy training
  • EMA momentum not tuned for training duration
  • EMA not applied to all parameters

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.

We send a single-use code tied to that address. Static provider and TLD rules do not reject valid addresses. Account trust determines the benefit after signup.

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.