gpt-oss-20b-MXFP4
This seed file names one model file by its SHA-256. The player downloads it, checks every byte and runs it on your CPU, or refuses.
- Refused at start-up
Seed file
Refused at start-up, after the 11.28 GiB model file is downloaded · tested , not run again with 1.3.1 (same engine build)
Seed file · 477 bytes · already in the player's archive,catalogue folderRefused at start-up
The engine loaded a chat template that is not the one the model file declares. Another model's template would change this model's answers, so the player stops.
Tested : the refusal came at start-up, after the model file (11.28 GiB) was downloaded. Not run again with player 1.3.1, which has the same engine build.
No run command for this seed. The seed file stays published. No published answer for this seed.
Cause as recorded, in French
l'allumage est refusé : GABARIT ÉTRANGER — le moteur a chargé un gabarit (b2215de6da8ba369…, 16616 o) qui n'est pas celui que la matière déclare (a4c9919cbbd4acdd…, 16738 o) — appliquer le sien reviendrait à interroger ce modèle avec les mots d'un autreSeed gpt-oss-20b · 477 bytes
Model by openai · file converted by ggml-org (ggml-org/gpt-oss-20b-GGUF)
Apache-2.0 · No extra terms in the licence; a usage policy is also published (our reading below) · label and model card read; licence file not opened · as declared, read · not legal advice
What this seed names
- Model file
gpt-oss-20b-MXFP4.gguf- Size
- 12,109,566,624 bytes (11.28 GiB)
- SHA-256
27cd6c432c7672cb812a92f611cf3ba7bbc35928262bb1e1253ff4ee6ae35901Click once to select the whole value.- Hosted at
- Hugging Face gave the same SHA-256 for this file (checked ) · If that file changes, the player refuses it.
- Licence
- Apache-2.0 · Separate usage policy
- Seed file
gpt-oss-20b.graine.json· 477 bytes ·182c15a64fbfd777a4e1aafd5bb6cb114b85411f6bf4a3c4bce49fe9ef443fdf
What we measured
- Start-up: refused on when the seed was made (Linux, glibc 2.35).
Not measured for this seed
- AMD vs Intel
- Linux vs Windows
- Speed
- Peak memory
Not measured for any seed: see Measurements.
The seed file (JSON, 477 bytes)
{"matiere":{"nom":"gpt-oss-20b-MXFP4.gguf","octets":12109566624,"sha256":"27cd6c432c7672cb812a92f611cf3ba7bbc35928262bb1e1253ff4ee6ae35901","substrat":["hf:ggml-org/gpt-oss-20b-GGUF/gpt-oss-20b-MXFP4.gguf"]},"nom":"gpt-oss-20b","notes":"licence declaree par l hebergeur : apache-2.0. les poids ne sont pas ici : ils sont NOMMES.","recette":{"classe":"x86-64-fils8","empreinte":"869bef8788166262e4a72fe04f0809756f561a2bae03a52a95e3fb7f294b8226"},"scellee_le":"2026-08-17","v":1}Licence as we read it, in French
Aucun seuil, aucune obligation de nommage.
Ce qui suit la redistribution est ce qu'Apache 2.0 exige : joindre une copie de la licence, conserver les mentions de droit d'auteur, de brevet et d'attribution, et signaler les fichiers modifiés.
Le dépôt amont porte EN PLUS un fichier USAGE_POLICY distinct, lu en entier — « We aim for our tools to be used safely, responsibly, and democratically… By using OpenAI gpt-oss-20b, you agree to comply with all applicable law. » : conformité légale, sans seuil.
NON VÉRIFIÉ à ce jour : si ce fichier est répliqué dans le dépôt GGUF redistribué.
étiquette et carte du dépôt amont lues (étiquette croisée README brut + API) ; le fichier de licence lui-même n'a pas été ouvert