Title: Teuken 7B Instruct v0.6 %E2%80%94 teuken-7b-instruct-v0-6 Canonical: https://llmeu.com/et/models/teuken-7b-instruct-v0-6/ Markdown mirror: https://llmeu.com/et/models/teuken-7b-instruct-v0-6.md Machine-readable: https://llmeu.com/llms.txt (index) - https://llmeu.com/llms-full.txt (full catalogue) - https://llmeu.com/llms-catalogue.json Provider: OpenGPT-X (Teuken) Licence: cc-by-nc-4.0 Hosting: catalogue only %E2%80%94 not hosted on LLM EU yet Sources: https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6 https://arxiv.org/abs/2410.08928 # Teuken 7B Instruct v0.6 [OpenGPT-X (Teuken)](/et/providers/opengptx/) %C2%B7 Pakkujad EU Viimati %C3%BCle vaadatud: 14. okt 2026 ## Suver%C3%A4%C3%A4nsus ja majutus See mudel on kataloogis viitamiseks. Seda ei hostita veel LLM EU-s ja %C3%BCkski esimese osapoole l%C3%B5pp-punkt ei teeninda seda siin. P%C3%A4ringud sellele suunatakse otse pakkujale, vastavalt pakkuja enda tingimustele. ## Kokkuv%C3%B5te Teuken 7B Instruct v0.6 on Saksamaa avalikult rahastatud mudel, mis on treenitud k%C3%B5igis 24 ametlikus EL-i keeles ja mille on v%C3%A4lja t%C3%B6%C3%B6tanud Fraunhofer, Forschungszentrum J%C3%BClich, TU Dresden ja DFKI. See on avaldatud CC-BY-NC-4.0 litsentsi all, seega on kaalud allalaaditavad, kuid neid tohib kasutada ainult isiklikel, mitte%C3%A4rilistel, uurimis- ja hariduslikel eesm%C3%A4rkidel. Projekti enda EU21 hindamisel on selle keskmine tulemus 0,57, edestades Llama 3.1 8B Instructi 0,563 ja Mistral 7B Instruct v0.3 0,527 samal testikomplektil. ## Spetsifikatsioonid - Pakkuja OpenGPT-X (Teuken) (DE) - Konteksti aken 4.1k - Parameetrid 7.45B - Modaalsused text - Litsents cc-by-nc-4.0 %C2%B7 [licence](https://creativecommons.org/licenses/by-nc/4.0/) - Avatud kaalud yes - V%C3%A4ljalaske kuup%C3%A4ev 28. juuli 2025 - %C3%9Clesanded [Saksa keel](/et/tasks/german/), [Mitmekeelne tugi](/et/tasks/multilingual-support/), [T%C3%B5lge](/et/tasks/translation/), [Kokkuv%C3%B5tete tegemine](/et/tasks/summarization/), [Klienditugi](/et/tasks/customer-support/) ## Keeled **Tugev:** de, en, fr, es, it, nl, pl, pt, sv, cs, da, fi, el, ro **Piisav:** bg, et, ga, hr, hu, lt, lv, mt, sk, sl #### Tugev - de - en - fr - es - it - nl - pl - pt - sv - cs - da - fi - el - ro #### Piisav - bg - et - ga - hr - hu - lt - lv - mt - sk - sl Kinnitamata: pakkuja ei ole seda v%C3%A4idet dokumenteerinud. %E2%80%94 a language is listed as strong only where a source claims it; the card links the evidence where one exists. ## Millal kasutada #### Tugevused - trained on all 24 official EU languages rather than English-first data - published EU21 evaluation against Llama 3.1 8B and Mistral 7B - small enough to run on one mid-range accelerator - German public funding, with evaluation code and data documentation #### Piirangud - CC-BY-NC-4.0 forbids commercial use of the released checkpoint - 4096-token context window - a 7B model, so reasoning and coding lag larger releases ## Korduma kippuvad k%C3%BCsimused ### Who built Teuken 7B Instruct v0.6? The card credits Fraunhofer, Forschungszentrum J%C3%BClich, TU Dresden, DFKI; funded by the German Federal Ministry for Economic Affairs and Climate Action. Source: https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6. ### Can Teuken 7B Instruct v0.6 be used commercially? No. It is published under CC-BY-NC-4.0 (https://creativecommons.org/licenses/by-nc/4.0/); the weights are downloadable, but the card says the licence "forbids commercial use of the released checkpoint" and allows only private, non-commercial, research and educational use. ### What evaluation does Teuken 7B Instruct v0.6 publish? The project's own EU21 evaluation (translated ARC, HellaSwag, TruthfulQA, MMLU) records 0.57 (project-reported) average, 0.59 (project-reported) on ARC and 0.66 (project-reported) on HellaSwag. The card states the project reports these itself and that they beat Llama 3.1 8B Instruct (0.563) and Mistral 7B Instruct v0.3 (0.527) on the same test set. Source: https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6. ### How large is Teuken 7B Instruct v0.6, and how many languages does it cover? The card records 7.45B (safetensors total 7,452,725,248) (source: https://huggingface.co/api/models/openGPT-X/Teuken-7B-instruct-v0.6) over a 4096-token context window (source: https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6/blob/main/config.json), trained on all 24 EU official languages: 14 recorded as strong and 10 more as supported. ### Is Teuken 7B Instruct v0.6 served on LLM EU? It has no endpoint in the catalogue, so it is catalogue only %E2%80%94 not hosted on LLM EU yet, with no region and no retention to report. A card whose status is "listed" is a catalogue entry, not a running deployment. ### Hinnastamine See mudel on kataloogis viitamiseks. Seda ei hostita veel LLM EU-s ja %C3%BCkski esimese osapoole l%C3%B5pp-punkt ei teeninda seda siin. P%C3%A4ringud sellele suunatakse otse pakkujale, vastavalt pakkuja enda tingimustele. ### Kinnitamise olek Kui v%C3%A4ljal pole allikat, n%C3%A4itab kaart seda kinnitamata, mitte ei t%C3%A4ida seda. T%C3%BChi v%C3%A4li on informatsioon. LLM EU Verified t%C3%A4hendab dokumenteeritust, mitte juriidilist sertifitseerimist. See ei ole tehisintellekti m%C3%A4%C3%A4ruse vastavushindamine ja LLM EU ei ole ELi asutus. ### Versioonid ja variandid - catalogue %C2%B7 Ainult kataloogis %C2%B7 [weights](https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6) ### N%C3%A4idisp%C3%A4ring ``` curl https://api.llmeu.com/v1/chat/completions \ -H "Authorization: Bearer $LLMEU_API_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"teuken-7b-instruct-v0-6","messages":[{"role":"user","content":"Hello"}]}' ``` Selle mudeli jaoks pole m%C3%A4nguv%C3%A4ljakut juurutatud: %C3%BCkski l%C3%B5pp-punkt ei teeninda seda, seega p%C3%A4ringul poleks kuhugi minna. ### Benchmarkid - EU21 average (translated ARC, HellaSwag, TruthfulQA, MMLU): **0.57 (project-reported)** [source](https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6) - EU21-ARC: **0.59 (project-reported)** [source](https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6) - EU21-HellaSwag: **0.66 (project-reported)** [source](https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6) ### Allikad - [huggingface.co](https://huggingface.co/openGPT-X/Teuken-7B-instruct-v0.6) - [arxiv.org](https://arxiv.org/abs/2410.08928) ### Kaardi muudatuste logi - 28. juuli 2025 %E2%80%94 V%C3%A4ljalaske kuup%C3%A4ev - 14. okt 2026 %E2%80%94 Viimati %C3%BCle vaadatud Viimati %C3%BCle vaadatud: 14. okt 2026 ### %C3%9Clesanded - [Saksa keel](/et/tasks/german/) - [Mitmekeelne tugi](/et/tasks/multilingual-support/) - [Klienditugi](/et/tasks/customer-support/)