Novaciano/SEX_ROLEPLAY-3.2-1B-ao-int4wo-gs128 warn
Weights only ship in a format that can run code when loaded; the weight files reference unusual code — review before loading. Plus 1 minor note.
claims base: Novaciano/SEX_ROLEPLAY-3.2-1B · chat template: present · view on Hugging Face ↗
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 2026-08-26128,260-token embedding scanned · 0 undertrained · lineage consistent · pickle audit: 4 non-standard global(s) |
| 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.
low Padding token is the EOS token
The pad token and the (only) EOS token are the same. Fine-tuning frameworks mask pad positions out of the loss, so training on this checkpoint teaches the model to never emit EOS — the Phi-4 / Qwen 2.5 / DeepSeek R1 infinite-generation bug. Safe to serve, hazardous to fine-tune; repoint pad_token at a dedicated token first.
How to fixingot patch
Align the stop-token declarations — a pure metadata fix to `generation_config.json` (and `config.json`).
- Identify the token the chat template actually ends assistant turns with (e.g. `<|eot_id|>`, `<end_of_turn>`, `<|im_end|>`) and make sure its id is in `generation_config.json`'s `eos_token_id` list.
- Keep `config.json`'s `eos_token_id` consistent with (or a subset of) `generation_config.json`'s — runtimes differ in which file they read.
- For the pad-equals-EOS hazard: repoint `pad_token` at a dedicated padding token before fine-tuning; serving is unaffected.
- Until the repo is fixed, pass explicit stop tokens to your serving stack (e.g. vLLM `stop_token_ids`, llama.cpp `--override-kv tokenizer.ggml.eos_token_id`).
info Embedding-norm glitch scan clean
No undertrained tokens found: every non-special token's embedding norm is above 0.3× the vocabulary median (0.949). The glitch-token data-corruption class has no candidate surface in this model.
medium Pickle references non-standard globals
The pickle imports globals outside the standard torch/numpy/collections set: torchao.dtypes.affine_quantized_tensor.AffineQuantizedTensor, torchao.dtypes.uintx.int4_cpu_layout.Int4CPUAQTTensorImpl, torchao.dtypes.uintx.int4_cpu_layout.Int4CPULayout, torchao.quantization.quant_primitives.ZeroPointDomain. Common in full-model (non-state-dict) saves — each is code that runs at load time. Review before loading, or demand a safetensors release.
info Weights consistent with claimed parent Novaciano/SEX_ROLEPLAY-3.2-1B
Mean cosine similarity of 64 sampled token-embedding rows against Novaciano/SEX_ROLEPLAY-3.2-1B is 1.000 — the weights plausibly descend from the declared base (relation: unspecified).
Check every checkpoint before it ships
Use the web app, API, CLI, or CI gate to scan candidate checkpoints and catch model drift before deployment. Public-model scans publish to the open database; paid plans add the volume needed for continuous checks.
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 Novaciano/SEX_ROLEPLAY-3.2-1B-ao-int4wo-gs128
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 | llama · 16 layers · 2048-dim |
| vocabulary | 128,260 tokens |
| license | apache-2.0 |
| serialization | no safetensors pickle |
| chat template | present · sha256:d82792f95932f1c9 |
| claimed lineage | Novaciano/SEX_ROLEPLAY-3.2-1B |
| lineage verified | consistent vs Novaciano/SEX_ROLEPLAY-3.2-1B — embedding-row cosine 1.000 |
| glitch-token surface | clean no undertrained tokens |
Full measured fingerprint
| architectures | LlamaModel |
| library | transformers |
| pipeline | text-generation |
| repo files | 7 — pickle: pytorch_model.bin |
| revision | 98e5a2ac52de |
| HF snapshot | 36 downloads · 0 likes · updated 2025-04-05 · captured 2026-08-25 |
| pickle audit | pytorch_model.bin — 10 standard global(s), suspicious torchao.dtypes.affine_quantized_tensor.AffineQuantizedTensor, torchao.dtypes.uintx.int4_cpu_layout.Int4CPUAQTTensorImpl, torchao.dtypes.uintx.int4_cpu_layout.Int4CPULayout, torchao.quantization.quant_primitives.ZeroPointDomain |
| embedding tensor | embed_tokens.weight · BF16 · 128,260×2048 |
| embedding norms | median 0.9492 · mean 0.9454 |
| lineage check | consistent — cosine 1 over 64 sampled rows vs Novaciano/SEX_ROLEPLAY-3.2-1B |
Battery runs (1)the run trace behind the findings — what each job measured
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| weights | complete | 2026-08-25 22:05 | 54s | 1 |
weights run 2026-08-25 — measurements
| probes run | glitch-norm-scan, zero-template-token-scan, pickle-static-analysis, lineage-norm-correlation |
| embedding tensor | embed_tokens.weight · BF16 · 128,260×2048 |
| glitch surface | 0 undertrained, 0 plain-ASCII |
| lineage check | consistent — cosine 1 over 64 rows vs Novaciano/SEX_ROLEPLAY-3.2-1B |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/Novaciano/SEX_ROLEPLAY-3.2-1B-ao-int4wo-gs128)