Downloads · 30 days
22
11% of all-time downloads
pulsar-v0.8-flash-gguf
Sonorix/pulsar-v0.8-flash-gguf
pulsar-v0.8-flash-gguf is a text generation model from Sonorix. Use it when you need the model to write or continue text. The card lists the license as apache-2.0.
Pulsar AI from Exo — компактный русскоязычный ассистент (0.5B). База — Qwen/Qwen2.5-0.5B-Instruct, LoRA SFT (r32, 2 эпохи) на собственном датасете. Системный промт зашит в chattemplate. Q80 — ~506 МБ, Q4KM — ~379 МБ.
All-time downloads
207
Public
Repo size
929 MB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.gguf929 MB · 100%
At a glance
- Task
- Text Generation
- License
- apache-2.0
- Access
- Public
- Created
- Sep 16, 2026
- Updated
- Sep 16, 2026
- SHA
- a21b470f
Base models
Model card
From the Hugging Face model README
Pulsar v0.8 Flash (GGUF Q8_0 + Q4_K_M)
Pulsar AI from Exo — компактный русскоязычный ассистент (0.5B).
База — Qwen/Qwen2.5-0.5B-Instruct, LoRA SFT (r32, 2 эпохи) на собственном датасете.
Системный промт зашит в chat_template. Q8_0 — ~506 МБ, Q4_K_M — ~379 МБ.
Использование
Ollama
ollama run hf.co/Sonorix/pulsar-v0.8-flash-gguf:Q8_0
ollama run hf.co/Sonorix/pulsar-v0.8-flash-gguf:Q4_K_M
LM Studio
Скачай .gguf → импортируй → пресет Qwen.
llama.cpp
llama-cli -m pulsar-v0.8-flash-Q4_K_M.gguf -p "кто ты?"
Параметры генерации
temperature: 0.3-0.6,top_p: 0.9,repetition_penalty: 1.1-1.2,max_tokens: 150-250
- Task
- Text Generation
- License
- apache-2.0
- Languages
- ru
- Created
- Sep 16, 2026
- Updated
- Sep 16, 2026