PyTorch Load Failed / Unpickling Error
Checkpoint load fails when files are corrupted or format is mismatched.
Checkpoint load fails when files are corrupted or format is mismatched.
What this failure is
PyTorch Load Failed / Unpickling Error is a Data Integrity failure seen during ML training runs. Checkpoint load fails when files are corrupted or format is mismatched. Common tags: Checkpoint, Load, Torch, Safetensors.
Is this what broke your run? Paste your log.
You're reading about PyTorch Load Failed / Unpickling Error. 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)
Checkpoint file is truncated: save was interrupted mid-write (torn write). Checkpoint was saved with a different model architecture or optimizer configuration. File was corrupted during network transfer (S3, GCS, NFS). The checkpoint was saved with GPU tensors and loaded on CPU (or vice versa). 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
- torch.load() raises UnpicklingError
- Checkpoint loads but model output is garbage
- Optimizer state is missing or mismatched
Common symptoms and what they mean
| Symptom | Why it happens |
|---|---|
| UnpicklingError: unsupported pickle protocol | Checkpoint file is truncated: save was interrupted mid-write (torn write) |
| OSError: [Errno 2] No such file or directory: checkpoint.pt | Checkpoint was saved with a different model architecture or optimizer configuration |
| RuntimeError: Error(s) in loading state_dict for Model: | File was corrupted during network transfer (S3, GCS, NFS) |
| KeyError: missing keys: optimizer state doesn't match model keys | The checkpoint was saved with GPU tensors and loaded on CPU (or vice versa) |
Which systems are affected
- Training checkpoint resumption
- Distributed checkpoint loading
- Model parallelism with sharded checkpoints
- Multi-cloud checkpoint storage
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: UnpicklingError: unsupported pickle protocol
- ✓Verified signal present: OSError: [Errno 2] No such file or directory: checkpoint.pt
- ✓Verified signal present: RuntimeError: Error(s) in loading state_dict for Model:
- ✓Verified signal present: KeyError: missing keys: optimizer state doesn't match model keys
- ✓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
- Checkpoint file is truncated: save was interrupted mid-write (torn write)
- Checkpoint was saved with a different model architecture or optimizer configuration
- File was corrupted during network transfer (S3, GCS, NFS)
- The checkpoint was saved with GPU tensors and loaded on CPU (or vice versa)
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.