guifav/caramelo warn
claims base: google/gemma-3-4b-it · chat template: present · view on Hugging Face ↗
Scan coverage
Ingot runs three batteries against a model. What each one checks →
| Battery | Looks at | Status |
|---|---|---|
| Static battery | Metadata & packaging | complete 2026-08-21 |
| Weights battery | Weights forensics — no GPU, no download | complete 2026-08-21262,208-token embedding scanned · 63 undertrained · lineage consistent |
| Behavioral battery | Live-inference differentials | not run |
Findings
Scanned 2026-08-21 · published from a community scan.
medium Chat template dropped vs parent
google/gemma-3-4b-it 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.
low Undertrained tokens in vocabulary (non-ASCII tail)
Embedding-norm scan flagged 63 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 Weights consistent with claimed parent google/gemma-3-4b-it
Mean cosine similarity of 64 sampled token-embedding rows against google/gemma-3-4b-it is 1.000 — the weights plausibly descend from the declared base (relation: unspecified).
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.
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 2026-08-21.
| architecture | gemma3 |
| parameters | 4300.1M |
| vocabulary | 262,208 tokens |
| license | gemma |
| serialization | safetensors |
| chat template | none |
| claimed lineage | google/gemma-3-4b-it |
| lineage verified | consistent vs google/gemma-3-4b-it — embedding-row cosine 1.000 |
| glitch-token surface | 63 undertrained candidates, 0 plain-ASCII |
Full measured fingerprint
| architectures | Gemma3ForConditionalGeneration |
| library | transformers |
| pipeline | text-generation |
| repo files | 11 |
| revision | 76e94628bd77 |
| HF snapshot | 214 downloads · 1 likes · updated 2026-07-03 · captured 2026-08-21 |
| embedding tensor | language_model.model.embed_tokens.weight · BF16 · 262,208×2560 |
| embedding norms | median 0.9986 · mean 0.9982 |
| lineage check | consistent — cosine 1 over 64 sampled rows vs google/gemma-3-4b-it |
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.
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| weights | complete | 2026-08-21 05:17 | 81s | 1 |
weights run 2026-08-21 — measurements
| probes run | glitch-norm-scan, lineage-norm-correlation |
| embedding tensor | language_model.model.embed_tokens.weight · BF16 · 262,208×2560 |
| glitch surface | 63 undertrained, 0 plain-ASCII |
| lineage check | consistent — cosine 1 over 64 rows vs google/gemma-3-4b-it |
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 guifav/caramelo
Remediation guidance addresses the documented findings only. It is evidence-driven repair, not a safety certification of the model.
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/guifav/caramelo)