Downloads · 30 days
44
15% of all-time downloads
dasha-v6-gguf
dariumi/dasha-v6-gguf
dasha-v6-gguf is a machine learning model from dariumi. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as llama3.2.
GGUF quantizations of Dasha's own chat model — a QLoRA fine-tune of Llama-3.2-3B-Instruct that gives her her personality and conversational style. Used as the default local "brain" runtime by DariaOS, loaded through l…
All-time downloads
288
Public
Repo size
8.8 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf8.8 GB · 100%
At a glance
- License
- llama3.2
- Access
- Public
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026
- SHA
- 485c3a87
Base models
Model card
From the Hugging Face model README
Dasha v6 (dasha-llama32-3b-instruct-v6)
GGUF quantizations of Dasha's own chat model — a QLoRA fine-tune of
Llama-3.2-3B-Instruct that gives her
her personality and conversational style. Used as the default local "brain" runtime by
DariaOS, loaded through llama.cpp
(core/runtimes/llm_runtime.py).
Files
| File | Quantization | Size | Use case |
|---|---|---|---|
dasha-v6-Q5_K_M.gguf | Q5_K_M | ~2.2 GB | Default — good quality/size tradeoff, what DariaOS installs by default |
dasha-v6-f16.gguf | F16 | ~6 GB | Full precision, for machines with the VRAM/RAM to spare |
Usage
Loaded automatically by DariaOS's installer (install.py → "Поставить llama.cpp сейчас?"),
or directly with llama.cpp:
llama-server -m dasha-v6-Q5_K_M.gguf
License
Inherits Meta's Llama 3.2 Community License from the base model.
- License
- llama3.2
- Languages
- ru
- Created
- Sep 2, 2026
- Updated
- Sep 2, 2026