OmniSVG/OmniSVG1.1_8B warn
Weights only ship in a format that can run code when loaded; the chat template was dropped from its base model, which changes behavior; its tokenizer differs from its claimed base model. Plus 2 minor notes.
Could not load this model from the Hugging Face API (private, gated, or nonexistent). Findings below are from our archive.
Scan coverageStatic battery2026-08-26Weights battery2026-08-26Behavioral batterynot rundetails
| Battery | Looks at | Status |
|---|---|---|
| Static battery | Metadata & packaging | complete 2026-08-26 |
| Weights battery | Weights forensics — no GPU, no download | complete 2026-08-26197,000-token embedding scanned · 1 undertrained · lineage inconsistent · pickle audit clean |
| Behavioral battery | Live-inference differentials | not run |
Ingot runs three batteries against a model. What each one checks →
Findings
Scanned 2026-08-26 · published from a community scan.
medium Pickle-serialized weights, no safetensors
Weights ship only as pickle-based files (pytorch_model.bin). Loading pickle executes arbitrary code from the file — prefer a safetensors release or load in a sandbox.
How to fix
Convert the weights to safetensors before loading them anywhere that matters.
- Do not load the pickle files in-process — pickle deserialization executes arbitrary code from the file.
- Convert locally in a sandbox: `pip install safetensors` and use `safetensors.torch.save_file` on a state dict loaded with `torch.load(..., weights_only=True)` (refuses most code-execution payloads), or use Hugging Face's `convert.py` space/script.
- Pin the exact revision hash you converted from, and load only your converted safetensors artifact from then on.
medium Chat template dropped vs parent
Qwen/Qwen2.5-VL-7B-Instruct ships a chat template; this repo does not. Serving stacks will silently fall back to a generic template, changing behavior. (In our 296-model census, 78% of pure quantization re-releases changed or dropped the template.)
How to fixingot patch
Restore the parent's chat template in `tokenizer_config.json` — a pure metadata fix.
- Run `ingot patch <owner/model>` — the patch manifest carries the parent's template and applies it to a local copy's `tokenizer_config.json`.
- Or fix by hand: copy the `chat_template` value from the parent repo's `tokenizer_config.json` into this model's, and pin your serving stack to that file.
- If the drift was intentional (the author retrained on a new template), confirm that in the model card before "fixing" it — restoring the parent template on retrained weights changes behavior too.
medium Vocabulary size differs from claimed parent (197000 vs 152064)
A changed vocab means changed tokenization: strings will split differently than on Qwen/Qwen2.5-VL-7B-Instruct, which can shift behavior on identifiers, codes, and non-English text.
How to fixweight-level
Not patchable: the vocab size mirrors the embedding matrix in the weights. Verify the change was intentional.
- Do not edit `vocab_size` in config.json to "match the parent" — it must equal the embedding table in the shipped weights or the model won't load.
- Diff the tokenizers (`tokenizer.json` / added_tokens) against the parent to see what was added or removed, and test your own identifiers, codes, and non-English text through both.
- If the drift is unexplained by the model card, treat tokenization-sensitive behavior as unvalidated on this model.
low Undertrained tokens in vocabulary (non-ASCII tail)
Embedding-norm scan flagged 1 undertrained tokens (norm < 0.3× the vocabulary median), but 0 decode to plain-ASCII strings, so exposure in English-language pipelines is limited. 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.
- 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.
- 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.
- 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.
info Pickle static analysis clean
Opcode-level parse of pytorch_model.bin (no code executed) found only standard serialization globals (3 distinct, all torch/collections/numpy). Pickle remains an executable format — this verifies the current bytes, not future uploads; prefer a safetensors release.
low Weights diverge from claimed parent Qwen/Qwen2.5-VL-7B-Instruct
This model declares Qwen/Qwen2.5-VL-7B-Instruct as its base (relation: unspecified), but mean cosine similarity of 49 sampled token-embedding rows against that parent is only 0.001 (true finetunes, merges, and quantizations sit above 0.8; independently trained weights sit near 0). Either the lineage label is wrong, or the model was so heavily re-trained, pruned, or distilled that the parent's properties (safety posture, evaluated behavior, licensing basis) should not be assumed to carry over. Verify provenance before relying on the parent's reputation.
How to fix
Fix or verify the `base_model` declaration so lineage checks can run.
- If you own the repo: correct the `base_model` field in the model card metadata to the real, public parent.
- If you don't: identify the true parent (config architecture + weight shapes narrow it fast) and re-scan with that lineage in mind.
Fix it
Some findings are metadata-level and patchable — apply the fixes to your local copy (your weights never leave your machine):
npx @ingotai/scan patch OmniSVG/OmniSVG1.1_8B
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 2026-08-26.
| architecture | qwen2_5_vl · 28 layers · 3584-dim |
| vocabulary | 197,000 tokens |
| license | apache-2.0 |
| serialization | no safetensors pickle |
| chat template | none |
| claimed lineage | Qwen/Qwen2.5-VL-7B-Instruct |
| lineage verified | inconsistent vs Qwen/Qwen2.5-VL-7B-Instruct — embedding-row cosine 0.001 |
| glitch-token surface | 1 undertrained candidates, 0 plain-ASCII |
Full measured fingerprint
| architectures | Qwen2_5_VLForConditionalGeneration |
| pipeline | text-generation |
| repo files | 4 — pickle: pytorch_model.bin |
| revision | f45d8cb9622e |
| HF snapshot | 139 downloads · 25 likes · updated 2026-03-30 · captured 2026-08-25 |
| pickle audit | pytorch_model.bin — 3 standard global(s) |
| embedding tensor | transformer.model.embed_tokens.weight · BF16 · 197,000×3584 |
| embedding norms | median 1.0345 · mean 1.0857 |
| lineage check | inconsistent — cosine 0.0013 over 49 sampled rows vs Qwen/Qwen2.5-VL-7B-Instruct |
Battery runs (1)the run trace behind the findings — what each job measured
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| weights | complete | 2026-08-25 21:51 | 5m | 1 |
weights run 2026-08-25 — measurements
| probes run | glitch-norm-scan, zero-template-token-scan, pickle-static-analysis, lineage-norm-correlation |
| probes skipped | token-decode: no tokenizer.json |
| embedding tensor | transformer.model.embed_tokens.weight · BF16 · 197,000×3584 |
| glitch surface | 1 undertrained, 0 plain-ASCII |
| lineage check | inconsistent — cosine 0.0013 over 49 rows vs Qwen/Qwen2.5-VL-7B-Instruct |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/OmniSVG/OmniSVG1.1_8B)