Model page

Qwen/Qwen2.5-VL-32B-Instruct warn

downloads 1.5Mlikes 499license apache-2.0arch qwen2_5_vlparams 33452.7Mupdated 2025-04-14

chat template: present · view on Hugging Face ↗

Scan coverage

Ingot runs three batteries against a model. What each one checks →

BatteryLooks atStatus
Static batteryMetadata & packagingcomplete 2026-08-22
Weights batteryWeights forensics — no GPU, no downloadcomplete 2026-08-22152,064-token embedding scanned · 6557 undertrained
Behavioral batteryLive-inference differentialsnot run

Findings

Scanned 2026-08-22 · published from a community scan.

medium Undertrained (glitch) token surface in vocabulary

Embedding-norm scan flagged 6557 undertrained tokens (norm < 0.3× the vocabulary median of 1.281), including 153 plain-ASCII strings that can appear in ordinary input as identifiers — e.g. "<unk>", "ForCanBeConverted", "ForCanBeConvertedToF", "$PostalCodesNL", "PostalCodesNL", "(stypy", "TokenNameIdentifier", "thuisontvangst". In models where this class was tested behaviorally, such tokens silently rewrote user input into confident, schema-valid, wrong output. These are candidates from the weights alone; behavioral confirmation requires the behavioral battery.

How to fixruntime guardweight-level

Keep the affected token strings out of the model's input — the scan-derived runtime guard carries this model's exact blocklist.

  1. Fetch this model's guard artifact (`/api/v1/guard/<owner>/<model>`): the confirmed corrupting tokens and the low-norm candidate list, derived from the published scan.
  2. Screen inbound text with it (the `@ingotai/guard` package is a reference implementation) and route flagged records to a different model or human review — verbatim-copy tasks on flagged strings are the failure mode.
  3. The underlying cause is undertrained embeddings in the weights; a true fix is weight-level (continued pretraining on the affected tokens) — that is not a patch, it's a training job.

Remediation guidance addresses the documented findings only. It is evidence-driven repair, not a safety certification of the model.

Fingerprint

The durable profile of this model: measured weights-and-metadata facts, rebuilt on every scan and battery run. Updated 2026-08-22.

architectureqwen2_5_vl · 64 layers · 5120-dim
parameters33452.7M
vocabulary152,064 tokens
licenseapache-2.0
serializationsafetensors
chat templatepresent · sha256:a0bc6f6fc7a29a80
glitch-token surface6,557 undertrained candidates, 153 plain-ASCII
Full measured fingerprint
architecturesQwen2_5_VLForConditionalGeneration
librarytransformers
pipelineimage-text-to-text
repo files32
revision7cfb30d71a1f
HF snapshot1.4M downloads · 499 likes · updated 2025-04-14 · captured 2026-08-21
embedding tensormodel.embed_tokens.weight · BF16 · 152,064×5120
embedding normsmedian 1.2808 · mean 1.1943
lineage checkno claimed base model
glitch-token samples"<unk>", "ForCanBeConverted", "ForCanBeConvertedToF", "$PostalCodesNL", "PostalCodesNL", "(stypy", "TokenNameIdentifier", "thuisontvangst", "useRalative", "useRal", "webElementX", "Cumhurba"

Battery runs

The run trace behind the findings above: every deep-battery job for this model, with what each run measured or why it failed. Findings are only as good as the runs that produced them.

batterystatusqueueddurationattempts
weightscomplete2026-08-21 07:4270s1
weights run 2026-08-21 measurements
probes runglitch-norm-scan, lineage-norm-correlation
embedding tensormodel.embed_tokens.weight · BF16 · 152,064×5120
glitch surface6,557 undertrained, 153 plain-ASCII
lineage checknot checked (no claimed base model)

Verdict badge

Ship the verdict in your README — it always shows the latest published analysis:

Ingot verdict: warn

[![Ingot scan](https://ingot.tools/api/v1/models/Qwen/Qwen2.5-VL-32B-Instruct/badge.svg)](https://ingot.tools/models/Qwen/Qwen2.5-VL-32B-Instruct)
Gate it in CI