Skip to content

Partial Checkpoint Save

Partial checkpoint saves write incomplete data when training is interrupted, leaving a file that appears valid but is truncated.

Quick answer

Partial checkpoint saves write incomplete data when training is interrupted, leaving a file that appears valid but is truncated.

Data Integrity#checkpoint#partial#save#atomic#corruption#storage

What this failure is

Partial Checkpoint Save is a Data Integrity failure seen during ML training runs. Partial checkpoint saves write incomplete data when training is interrupted, leaving a file that appears valid but is truncated. Common tags: Checkpoint, Partial, Save, Atomic.

Live diagnosis, no signup

Is this what broke your run? Paste your log.

You're reading about Partial Checkpoint Save. 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)

Job was preempted or killed during checkpoint write. Disk full during checkpoint write. NFS timeout during checkpoint write. Multiple processes writing to the same checkpoint file. 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

  • Checkpoint file exists but is smaller than expected
  • Resume from checkpoint crashes with pickle error
  • File size differs between saves of the same model

Common symptoms and what they mean

SymptomWhy it happens
Checkpoint file is smaller than expected (e.g., 500MB instead of 2GB)Job was preempted or killed during checkpoint write
UnpicklingError when loading checkpointDisk full during checkpoint write
torch.load fails halfway throughNFS timeout during checkpoint write

Which systems are affected

  • Training with preemptible instances
  • Long-running training with periodic checkpoints
  • Network filesystem (NFS) checkpoint targets

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: Checkpoint file is smaller than expected (e.g., 500MB instead of 2GB)
  • Verified signal present: UnpicklingError when loading checkpoint
  • Verified signal present: torch.load fails halfway through
  • 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

  • Job was preempted or killed during checkpoint write
  • Disk full during checkpoint write
  • NFS timeout during checkpoint write
  • Multiple processes writing to the same checkpoint file

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.