Downloads · 30 days
206
4% of all-time downloads
texdata/Qwen3.6-35B-A3B-Tool-Slo-GGUF
Qwen3.6-35B-A3B-Tool-Slo-GGUF is a text generation model from texdata. Use it when you need the model to write or continue text. The card lists the license as other.
Trained by MediaAtlas — LLM fine-tuning on your own data, trained in the EU, weights delivered. Pricing · All our models
Downloads · 30 days
206
4% of all-time downloads
All-time downloads
5K
Public
Repo size
60.4 GB
Likes
0
Public
Click a slice to open those files.
.gguf60.4 GB · 100%
From the Hugging Face model README
Trained by MediaAtlas — LLM fine-tuning on your own data, trained in the EU, weights delivered. Pricing · All our models
Na kratko: Slovenski model (35B MoE, 3B aktivnih) za zanesljive klice orodij skozi cel večkoračni pogovor, v Qwen XML formatu. Teče lokalno (llama.cpp, LM Studio); GGUF z vizualnim projektorjem. Sprejme tudi slike.
In short (EN): Slovenian 35B MoE model (3B active) fine-tuned for reliable native tool / function calling across multi-turn conversations (Qwen3.6 XML format). GGUF with vision projector for llama.cpp / LM Studio. Slovenian continued pretraining + instruction SFT + agentic tool-calling SFT (APIGen-MT, ToolACE, Hermes).
Vision / slike: takes images too. Load the mmproj-*.gguf next to the model (llama.cpp --mmproj; LM Studio pairs it automatically). Our training was text-only: image understanding is the base model's and was not evaluated in Slovenian.
Slovenski pogovorni / glasovni model z zanesljivim, izboljšanim klicanjem orodij (function / tool calling), grajen na Qwen3.6-35B-A3B.
Dodatno SFT fine-tunan za multi-turn agentske tool-calling delotoke, na:
Cilj: zanesljivo native function-calling skozi cel pogovor (ne le v prvi izmenjavi).
| kvant | velikost | opomba |
|---|---|---|
qwen3.6-35b-a3b-tool-slo-Q4_K_M.gguf | ~21 GB | hitrost / manj VRAM |
qwen3.6-35b-a3b-tool-slo-Q8_0.gguf | ~36 GB | kvaliteta |
mmproj-qwen3.6-35b-a3b-tool-slo-F16.gguf | ~0.9 GB | vision projektor |
<tool_call><function=...><parameter=...></parameter></function></tool_call>)llama-server -m qwen3.6-35b-a3b-tool-slo-Q4_K_M.gguf \
--mmproj mmproj-qwen3.6-35b-a3b-tool-slo-F16.gguf \
--jinja -ngl 99 -c 40960
--jinja vklopi tool-calling prek vgrajenega chat templata.
Base Qwen/Qwen3.6-35B-A3B is Apache-2.0. Training data carries its own terms: Slovenian Wikipedia (CC BY-SA 4.0), FineWeb2-sl (ODC-BY 1.0), APIGen-MT, ToolACE and Hermes function-calling datasets. Marked license: other; resolve data terms before commercial use.
Cite this work (Tadej Fius, MediaAtlas Ltd):
@misc{fius2026qwen35btoolgguf,
title = {Qwen3.6-35B-A3B Tool-Slo (GGUF)},
author = {Fius, Tadej},
year = {2026},
publisher = {MediaAtlas Ltd},
howpublished = {Hugging Face},
url = {https://huggingface.co/texdata/Qwen3.6-35B-A3B-Tool-Slo-GGUF}
}
Upstream / source citations:
@misc{qwen3, title={Qwen3 Technical Report}, author={{Qwen Team}}, year={2025}, url={https://huggingface.co/Qwen}}
@misc{wikipedia_sl, title={Slovenian Wikipedia}, author={{Wikimedia Foundation}}, note={CC BY-SA 4.0}, url={https://sl.wikipedia.org}}
@misc{penedo2024fineweb2, title={FineWeb2}, author={Penedo, G. and others}, year={2024}, url={https://huggingface.co/datasets/HuggingFaceFW/fineweb-2}}
@article{liu2024apigen, title={APIGen: Automated Pipeline for Generating Verifiable and Diverse Function-Calling Datasets}, author={Liu, Zuxin and others}, journal={NeurIPS}, year={2024}}
@misc{hermes_fc, title={Hermes Function-Calling Datasets}, author={{NousResearch}}, url={https://huggingface.co/NousResearch}}
CPT: Slovenian Wikipedia (CC BY-SA), FineWeb2-sl (ODC-BY). Tools: Hermes + APIGen/xLAM.