google_gemma-3-4b-it-Q4_K_M
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.
- Passed start-up
- Same on Linux and Windows
- Same on AMD and Intel, same OS
Use this seed
Download the playerHave the player? Run itNeeds
- RAM: the player refuses below 2.32 GiB
- Peak memory: not measured for this seed. The player compares your RAM with the file size, not with the memory it will use; on the 2 seeds we measured, the largest process peaked 49% and 77% above the file size (Linux, )
- Free disk ≥ 2.43 GiB
- x86-64 processor with AVX2, FMA and F16C
- Linux (glibc 2.35 or newer) or Windows 11 (builds 26100 and 26200 measured)
The player checks RAM, free disk and the processor before downloading, and stops if any of them falls short; a value it cannot read does not stop it.
Downloads 2.32 GiB from Hugging Face.
Seed file · 499 bytes · already in the player's archive,catalogue folderSeed gemma3-4b · 499 bytes
Model by google · file converted by bartowski (bartowski/google_gemma-3-4b-it-GGUF)
Gemma Terms of Use · Extra terms · licence text read at the source · as declared, read · not legal advice
Linux vs Windows comparison: same tokens on 3 test questions, 2 machines, .
Run it
From the unpacked player folder. --maison is the folder for this model and your conversations.
# germer: download the model file and check every byte./memown germer catalogue/gemma3-4b.graine.json --maison ~/ia/gemma3-4b Expected: a line starting with VERDICT : VERT (VERT: all four checks passed, including the byte-for-byte check)
# germer: download the model file and check every byte.\memown.exe germer catalogue\gemma3-4b.graine.json --maison "$env:USERPROFILE\ia\gemma3-4b" Expected: a line starting with VERDICT : VERT (VERT: all four checks passed, including the byte-for-byte check)
Compare your answer with ours
Ask it first, right after germer, in a new --maison folder. Type the question exactly, in French: the published fingerprint is for these exact bytes.
Expected: 11 tokens, seen on 5 machines (Linux and Windows 11), to , 0 differences · fingerprint 4f31bb7389089d47b2dafec16547c10624b6812b5c229f913e6fa879d0762e82 Click once to select the whole value.
# parler: ask a question./memown parler --maison ~/ia/gemma3-4b --dire "Quelle est la capitale de la France ?"grep -c '"ids_sha256":"4f31bb7389089d47b2dafec16547c10624b6812b5c229f913e6fa879d0762e82"' ~/ia/gemma3-4b/vies/vie/journal.json Expected: 1
# parler: ask a question.\memown.exe parler --maison "$env:USERPROFILE\ia\gemma3-4b" --dire "Quelle est la capitale de la France ?"Select-String -Quiet -SimpleMatch -Path "$env:USERPROFILE\ia\gemma3-4b\vies\vie\journal.json" -Pattern '"ids_sha256":"4f31bb7389089d47b2dafec16547c10624b6812b5c229f913e6fa879d0762e82"'Expected: True
1 or True: your tokens are the same as ours. 0 or False: they differ. Tell us either way.
To ask your own question, change the text after --dire. Keep the test question to compare with ours.
What this seed names
- Model file
google_gemma-3-4b-it-Q4_K_M.gguf- Size
- 2,489,758,112 bytes (2.32 GiB)
- SHA-256
4996030242583a40aa151ff93f49ed787ac8c25e4120c3ae4588b2e2a7d1ae94- Hosted at
- Hugging Face gave the same SHA-256 for this file (checked ) · If that file changes, the player refuses it.
- Licence
- Gemma Terms of Use · Extra terms
- Seed file
gemma3-4b.graine.json· 499 bytes ·5b3c02c8e2194e041664f2ceefe046ecdc703c7290c9af857af9403759ab8b8c
What we measured
- Start-up: passed on when the seed was made (Linux, glibc 2.35).
- Linux vs Windows: same tokens · 3 test questions · 2 machines · · data (JSON)
- AMD vs Intel, same OS: same tokens · 3 test questions · 2 machines ·
- 12 expected answers · up to 5 machines · 0 differences
- 16.8–25.5 tokens/s · 3 rented AMD EPYC servers shared with other work · 8 threads · short answers in a new conversation · player 1.2, same engines as 1.3.1 · . Your speed: not measured.
Not measured for this seed
- Peak memory
Not measured for any seed: see Measurements.
12 expected answers
| Question | Expected | Measured |
|---|---|---|
| Quelle est la capitale de la France ? (“What is the capital of France?”) | 11 tokens · SHA-256 of the token IDs 4f31bb73…62e82 | 5 machines |
| Continue la suite : 2, 4, 8, 16, (“Continue the sequence: 2, 4, 8, 16,”) | 4 tokens · SHA-256 of the token IDs bf41a994…cdd4b | 3 machines |
| What is the capital of Italy? | 22 tokens · SHA-256 of the token IDs b4e1c3ff…e64eb | 4 machines |
| Continue la suite : 2, 4, 8, 16, (“Continue the sequence: 2, 4, 8, 16,”) | 12 tokens · SHA-256 of the token IDs 866f56e4…774a8 | 4 machines |
| Explique en deux phrases pourquoi le ciel est bleu. (“Explain in two sentences why the sky is blue.”) | 70 tokens · SHA-256 of the token IDs a9509003…3ac28 | 3 machines |
| Recopie exactement : le chat dort sur le tapis rouge et la pluie tombe doucement. (“Copy exactly: the cat sleeps on the red rug and the rain falls gently.”) | 15 tokens · SHA-256 of the token IDs bbc04042…e742c | 3 machines |
| Donne trois nombres premiers entre 10 et 30. (“Give three prime numbers between 10 and 30.”) | 29 tokens · SHA-256 of the token IDs 80cbde6e…3dd6b | 3 machines |
| Traduis en anglais : la mer est calme ce soir. (“Translate into English: the sea is calm tonight.”) | 7 tokens · SHA-256 of the token IDs 1f7b8769…377a3 | 3 machines |
| Quel est le plus grand océan du monde ? (“What is the largest ocean in the world?”) | 18 tokens · SHA-256 of the token IDs be05089d…7e84f | 3 machines |
| Écris une phrase de politesse pour finir un e-mail. (“Write a polite sentence to end an email.”) | 96 tokens · SHA-256 of the token IDs 148a6163…ed327 | 3 machines |
| Résume en une phrase ce que fait un four à pain. (“Sum up in one sentence what a bread oven does.”) | 45 tokens · SHA-256 of the token IDs 3f4b5419…e7360 | 3 machines |
| Continue la suite : 3, 6, 12, 24, (“Continue the sequence: 3, 6, 12, 24,”) | 3 tokens · SHA-256 of the token IDs 510ed716…3694a | 3 machines |
The seed file (JSON, 499 bytes)
{"matiere":{"nom":"google_gemma-3-4b-it-Q4_K_M.gguf","octets":2489758112,"sha256":"4996030242583a40aa151ff93f49ed787ac8c25e4120c3ae4588b2e2a7d1ae94","substrat":["hf:bartowski/google_gemma-3-4b-it-GGUF/google_gemma-3-4b-it-Q4_K_M.gguf"]},"nom":"gemma3-4b","notes":"licence declaree par l hebergeur : gemma. 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
Toute redistribution de Gemma OU d'une dérivée doit fournir au destinataire une copie de l'accord, et être accompagnée d'un fichier « Notice » portant exactement : « Gemma is provided under and subject to the Gemma Terms of Use found at ai.google.dev/gemma/terms ».
La politique d'usage prohibé est incorporée par référence et vise explicitement les « Model Derivatives ».
Aucun seuil chiffré.
Le dépôt GGUF redistributeur (bartowski/google_gemma-3-4b-it-GGUF) ne déclare AUCUNE étiquette de licence — l'absence d'étiquette en aval ne libère rien, ce sont les termes de Google qui suivent la matière.
texte de la licence lu à la source