Is this a hardware or a software issue?

Software failures can look like either layer depending on the symptom. The discriminator is the resolution of the symptom to the recommended fix: if applying the steps under "How to fix it" eliminates the symptom, this is a software/configuration class issue regardless of the hardware it surfaced on. Repeat the fix on a fresh node to confirm.

How long does the fix typically take?

Most resolutions from this entry take 10 to 30 minutes of engineer time once the diagnosis is confirmed. Long fixes come from skipping the diagnosis step and chasing apparent symptoms (validation drift, memory growth, slow throughput) instead of the actual root cause listed above. Apply the fix in a small experiment first, then roll it forward.

Does this affect inference or only training?

Primarily this is a software failure observed during training, but the same misconfiguration can leak into downstream evaluation and serving. Treat the prevention checklist as a guardrail: any pipeline that touches a trained checkpoint should re-validate the failure mode before promoting the model.

What's the cheapest way to confirm this is the problem?

Cheapest signal: reproduce the symptom on a 1-GPU toy run with the same configuration. If the toy run exhibits the same behavior, the cause is the configuration rather than the cluster. If only the multi-GPU run is affected, the failure is a distributed interaction, so Megatron-LM is the most likely layer to inspect.

Can the Denpex platform detect this automatically?

Yes. Denpex correlates the symptoms in this entry against your training logs in real time and surfaces the failure taxonomy, root cause, and exact remediation command on the console. Three full diagnoses are included free during signup, and the encyclopedia entry you are reading is generated from the same signal library that powers the platform.

What related failures should I check next?

Closely related failures cluster around #RNG State Divergence. After applying this fix, validate against the related entries below to confirm no overlapping cause is masking as this one. Engineers routinely discover that the dominant failure mode is actually a second, sibling failure compounded on top.

Will this recur after the fix?

Applied once with the prevention checklist, this failure does not recur under normal operation. The recurring variant most engineers encounter is the re-emergence of the same symptom on a different model or stack after RNG State Divergence is upgraded. Pin compatible versions and rerun the prevention smoke test on every dependency change.

Is there a fingerprint I can monitor?

Yes. Training hangs during the backward pass when using Tensor Parallelism combined with activation checkpointing (recomputation). is the highest-signal fingerprint. Wire a metric or alert on that signal (or its absence after the fix) so you can detect the recurrence automatically. The Denpex platform encodes this exact fingerprint and surfaces it directly in the dashboard.

What is the official documentation that covers this?

Reference documentation is listed in the References section at the bottom of this entry. The most authoritative URLs are the upstream framework docs and the issue trackers cited; the Denpex entry synthesizes both into the executive, engineer, and on-call perspectives so each audience can act without reading every linked source.

Does this entry include working code?

Yes. The "Code Examples" section contains copy-paste-ready snippets that target the most common stacks (PyTorch, Hugging Face Trainer, DeepSpeed, Megatron-LM, vLLM/Transformers). Each snippet is annotated with why it works so you can adapt the pattern to your own framework without losing fidelity.

Tensor Parallel RNG State Context Divergence Deadlock

Q: What does this failure look like in production?

In production this failure typically appears during the first tens-to-hundreds of training steps: loss curves become non-monotonic, validation metrics plateau above the baseline, and downstream checkpoints either diverge or fail to improve. Engineers frequently misread it as data drift when the underlying mechanism is in tensor parallelism (tp), ranks within the same tp group must execute identical control flows for collective operations. if the `cudarngstatestracker` is misconfigured or activation offloading fails to perfectly restore the rng state, dropout masks or stochastic operations will differ across tp ranks. this causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of nccl calls and deadlocking.. Validating against the symptom list in this entry is the fastest way to confirm the diagnosis.

Q: How does this scale from 1 GPU to 1000 GPUs?

At single-GPU scale the symptom usually looks like slow convergence or a single anomalous loss value. At 64+ GPU scale, distributed interactions amplify in tensor parallelism (tp), ranks within the same tp group must execute identical control flows for collective operations. if the `cudarngstatestracker` is misconfigured or activation offloading fails to perfectly restore the rng state, dropout masks or stochastic operations will differ across tp ranks. this causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of nccl calls and deadlocking. through collectives, and the failure can manifest as a full job abort. Always reproduce the fix at the smallest cluster configuration that still exhibits the symptom before promoting to production scale.

In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of NCCL calls and deadlocking.

Quick answer

In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations.

Symptom: RuntimeError: RNG state changed within GPU RNG context
Root cause: In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.
Recommended fix: Ensure strict RNG Tracker usage tensor_parallel.get_cuda_rng_tracker().add('model-parallel-rng', seed) Megatron-LM requires stochastic operations in TP regions to use the specific RNG tracker context so that seeds are perfectly synchronized across the TP group.
How Denpex helps: Denpex investigates Tensor Parallel RNG State Context Divergence Deadlock using the evidence you provide or your connected workload collects. Earlier rank, host or application evidence is needed to distinguish an initiating failure from a downstream report.

Software#RNG State Divergence

What this failure is

Tensor Parallel RNG State Context Divergence Deadlock is a Software failure seen during ML training runs. In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of NCCL calls and deadlocking. Common tags: RNG State Divergence.

Live diagnosis, no signup

Is this what broke your run? Paste your log.

You're reading about Tensor Parallel RNG State Context Divergence Deadlock. Paste your own traceback and relevant evidence for an investigation of your workload, with a next action or a specific missing fact. A reference entry does not establish your cause. No account or card for the free diagnosis. Review data handling before submitting sensitive logs.

Before uploading, review cloud data handling and local options.

training_logs.txt

No log to hand? Try one:

Upload or drop a plain-text log

Upload your log file. Large files are processed in your browser. Only selected evidence is submitted when you run diagnosis.

3 free diagnoses/day

Why it happens (the mechanism)

The deadlock manifests during the backward pass AllReduce, but the actual corruption happened much earlier during the forward pass RNG sampling or activation recomputation.

What you'll observe

RuntimeError: RNG state changed within GPU RNG context
NCCL AllReduce timeout during backward pass
Process stuck in torch.distributed.all_reduce

Common symptoms and what they mean

Symptom	Why it happens
Training hangs during the backward pass when using Tensor Parallelism combined with activation checkpointing (recomputation).	In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of NCCL calls and deadlocking.
Sometimes accompanied by silent math divergence before the hang.	In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of NCCL calls and deadlocking.

Symptom

Why it happens

Training hangs during the backward pass when using Tensor Parallelism combined with activation checkpointing (recomputation).

Sometimes accompanied by silent math divergence before the hang.

Which systems are affected

Megatron-LM
PyTorch CUDA RNG
NCCL

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.

✓Check for 'RNG state changed' warnings in the logs.
✓Print and compare `torch.cuda.get_rng_state()` hashes across TP ranks at the start of the backward pass.
✓Disable activation checkpointing to see if the deadlock disappears.

Root cause

In Tensor Parallelism (TP), ranks within the same TP group must execute identical control flows for collective operations. If the `CudaRNGStatesTracker` is misconfigured or activation offloading fails to perfectly restore the RNG state, dropout masks or stochastic operations will differ across TP ranks. This causes divergent control flow (e.g., dropping different tokens in sequence parallel), leading to ranks submitting different sequences of NCCL calls and deadlocking.

The fix and how to prevent it

Searchable error signature

search key

RuntimeError: RNG state changed within GPU RNG context

Use this text as a lookup key in logs and upstream issue trackers. It is not presented as a captured customer log. Confirm the cause from your own preceding events, versions, configuration and the cited references.

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.

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

References

↗Megatron-LM RNG State Context Testing Issues

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.

Diagnose your logs free Compare a solved incident Browse all failures

Related Software errors

Watchdog Timeout from Inconsistent Tensor Shapes

Software · medium

Xid 13: Graphics Engine Exception due to Kernel Out-of-Bounds Access

Software · high

Adam Epsilon Underflow in FP16 Mixed Precision

Software · critical

DeepSpeed ZeRO-3 Stale Sync during Grad Clipping

Software · medium