Goekdeniz-Guelmez/Josiefied-Qwen2.5-0.5B-Instruct-abliterated-v1-gguf 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. 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 2026-08-26pickle 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 (josiefied-qwen2.5-0.5b-instruct-abliterated-v1.fp16.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
Goekdeniz-Guelmez/Josiefied-Qwen2-0.5B-Instruct-abliterated 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.
info Pickle static analysis clean
Opcode-level parse of josiefied-qwen2.5-0.5b-instruct-abliterated-v1.fp16.bin (no code executed) found only standard serialization globals (0 distinct, all torch/collections/numpy). Pickle remains an executable format — this verifies the current bytes, not future uploads; prefer a safetensors release.
low Pickle checkpoint only partially analyzable
Static analysis could not fully parse: josiefied-qwen2.5-0.5b-instruct-abliterated-v1.fp16.bin: legacy parse stopped after 0 pickle(s): unsupported pickle opcode 0x1 at 9. Unparsed content is unverified.
low GGUF chat template differs from the source repo's
The template embedded at conversion time (4433 chars) no longer matches the source repo's current template (477 chars). Source-repo template fixes never propagate into converted GGUFs — 43% of popular GGUF repos drift this way, including quants that resurrect already-fixed launch bugs. Diff the two before deploying; re-embed with gguf-py if the source's fix matters.
How to fixingot patch
Fix the GGUF's embedded metadata in place with gguf-py — template, EOS id, and pre-tokenizer are all metadata-editable; no requant needed.
- Template or EOS drift: copy the current values from the source repo and write them into the GGUF (`gguf_set_metadata.py` / gguf-py) — the tensor data is untouched.
- Missing pre-tokenizer type: reconvert with a current `convert_hf_to_gguf.py`, or set the correct `tokenizer.ggml.pre` for the architecture.
- Until the file is fixed, override at load time: llama.cpp `--override-kv tokenizer.ggml.eos_token_id=int:<id>` and `--chat-template-file <fixed.jinja>`.
- Prefer a re-upload from the quantizer once the source repo's fix lands — already-downloaded GGUFs never pick up upstream fixes on their own.
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 Goekdeniz-Guelmez/Josiefied-Qwen2.5-0.5B-Instruct-abliterated-v1-gguf
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.
| license | apache-2.0 |
| serialization | gguf pickle |
| chat template | none |
| claimed lineage | Goekdeniz-Guelmez/Josiefied-Qwen2-0.5B-Instruct-abliterated |
| lineage verified | unverified — weights battery pending |
Full measured fingerprint
| pipeline | text-generation |
| repo files | 17 — pickle: josiefied-qwen2.5-0.5b-instruct-abliterated-v1.fp16.bin |
| revision | 269f2b80c531 |
| HF snapshot | 248 downloads · 1 likes · updated 2025-03-07 · captured 2026-08-25 |
| pickle audit | josiefied-qwen2.5-0.5b-instruct-abliterated-v1.fp16.bin — 0 standard global(s) · legacy (pre-1.6) format, head-scan only |
Battery runs (1)the run trace behind the findings — what each job measured
| battery | status | queued | duration | attempts |
|---|---|---|---|---|
| weights | complete | 2026-08-25 21:49 | 7s | 1 |
weights run 2026-08-25 — measurements
| probes run | gguf-metadata |
Verdict badge
Ship the verdict in your README — it always shows the latest published analysis:
[](https://ingot.tools/models/Goekdeniz-Guelmez/Josiefied-Qwen2.5-0.5B-Instruct-abliterated-v1-gguf)