nvidia/Kimi-K2.7-Code-NVFP4 warn
claims base: moonshotai/Kimi-K2.7-Code · 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 2026-08-22 |
| Weights battery | Weights forensics — no GPU, no download | complete 2026-08-22163,840-token embedding scanned · 226 undertrained · lineage consistent |
| Behavioral battery | Live-inference differentials | not run |
Findings
Scanned 2026-08-22 · published from a community scan.
medium Repo ships executable Python (trust_remote_code)
The repository contains custom code files that run in-process when loaded with trust_remote_code=True. Pin the revision hash and review the code before loading.
How to fix
Review and pin the custom code; never float on `main` with trust_remote_code=True.
- Read every `.py` file in the repo before first load — this code runs in your process.
- Pin the revision: `from_pretrained(model_id, revision="<commit sha>", trust_remote_code=True)` so a later push can't swap the code under you.
- Prefer a version of the architecture already in `transformers` if one exists, which removes the remote-code requirement entirely.
low Undertrained tokens in vocabulary (non-ASCII tail)
Embedding-norm scan flagged 226 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 moonshotai/Kimi-K2.7-Code
Mean cosine similarity of 64 sampled token-embedding rows against moonshotai/Kimi-K2.7-Code 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 2026-08-22.
| architecture | kimi_k25 |
| vocabulary | 163,840 tokens |
| license | other |
| serialization | safetensors custom code |
| chat template | none |
| claimed lineage | moonshotai/Kimi-K2.7-Code |
| lineage verified | consistent vs moonshotai/Kimi-K2.7-Code — embedding-row cosine 1.000 |
| glitch-token surface | 226 undertrained candidates, 0 plain-ASCII |
Full measured fingerprint
| architectures | KimiK25ForConditionalGeneration |
| library | Model Optimizer |
| pipeline | text-generation |
| repo files | 87 |
| revision | 9f28d60908c9 |
| HF snapshot | 1.1M downloads · 9 likes · updated 2026-07-06 · captured 2026-08-21 |
| embedding tensor | language_model.model.embed_tokens.weight · BF16 · 163,840×7168 |
| embedding norms | median 0.7619 · mean 0.747 |
| lineage check | consistent — cosine 1 over 64 sampled rows vs moonshotai/Kimi-K2.7-Code |
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 07:43 | 3m | 1 |
weights run 2026-08-21 — measurements
| probes run | glitch-norm-scan, lineage-norm-correlation |
| probes skipped | token-decode: no tokenizer.json |
| embedding tensor | language_model.model.embed_tokens.weight · BF16 · 163,840×7168 |
| glitch surface | 226 undertrained, 0 plain-ASCII |
| lineage check | consistent — cosine 1 over 64 rows vs moonshotai/Kimi-K2.7-Code |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/nvidia/Kimi-K2.7-Code-NVFP4)