Instructions to use xyzmean/llama-8B-ru with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use xyzmean/llama-8B-ru with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="xyzmean/llama-8B-ru",
	filename="unsloth.Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use xyzmean/llama-8B-ru with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf xyzmean/llama-8B-ru:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf xyzmean/llama-8B-ru:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf xyzmean/llama-8B-ru:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf xyzmean/llama-8B-ru:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf xyzmean/llama-8B-ru:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf xyzmean/llama-8B-ru:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf xyzmean/llama-8B-ru:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf xyzmean/llama-8B-ru:Q4_K_M

Use Docker

docker model run hf.co/xyzmean/llama-8B-ru:Q4_K_M

LM Studio
Jan

vLLM

How to use xyzmean/llama-8B-ru with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "xyzmean/llama-8B-ru"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "xyzmean/llama-8B-ru",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/xyzmean/llama-8B-ru:Q4_K_M

Ollama
How to use xyzmean/llama-8B-ru with Ollama:
```
ollama run hf.co/xyzmean/llama-8B-ru:Q4_K_M
```

Unsloth Studio new

How to use xyzmean/llama-8B-ru with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for xyzmean/llama-8B-ru to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for xyzmean/llama-8B-ru to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for xyzmean/llama-8B-ru to start chatting

Docker Model Runner
How to use xyzmean/llama-8B-ru with Docker Model Runner:
```
docker model run hf.co/xyzmean/llama-8B-ru:Q4_K_M
```

Lemonade

How to use xyzmean/llama-8B-ru with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull xyzmean/llama-8B-ru:Q4_K_M

Run and chat with the model

lemonade run user.llama-8B-ru-Q4_K_M

List all available models

lemonade list

DeepSeek-R1-Distill-Llama-8B-Abliterated-RU

Модель представляет собой дообученную версию DeepSeek-R1-Distill-Llama-8B-Abliterated:cite[10] на русскоязычном синтетическом датасете инструкций.

Особенности

Базовая архитектура: Основана на Llama-8B с дистилляцией знаний от DeepSeek-R1:cite[3]:cite[6]
Дообучение: Адаптация под русский язык через датасет ru-instruct (800k+ примеров)
Оптимизация:
- Удаление артефактов перевода через модель Den4ikAI/nonsense_gibberish_detector:cite[8]
- Дедупликация методом SimHash
Мультидоменность: Поддержка математических расчётов, программирования и QA:cite[8]

Датасет обучения

Комбинация 7 переведённых датасетов:

Источник	Описание
OpenOrca-ru	1.2M диалогов с детализированными ответами
OpenHermes-2.5-ru	Инструкции для сложных задач
Dolphin-ru	Мультизадачные инструкции
GSM8k-ru	Математические задачи
Boolq-ru	Вопросы с ответами Да/Нет
Conala-mined-ru	Python-сниппеты
Alpaca-cleaned-ru	Общие инструкции

Формат данных:

{
    "conversations": [
        {"role": "system", "content": "..."},
        {"role": "user", "content": "..."},
        {"role": "assistant", "content": "..."}
    ],
    "source": "название_датасета"
}

Downloads last month: 17

GGUF

Model size

8B params

Architecture

llama

Hardware compatibility

4-bit

Model tree for xyzmean/llama-8B-ru

Base model

deepseek-ai/DeepSeek-R1-Distill-Llama-8B

Finetuned

stepenZEN/DeepSeek-R1-Distill-Llama-8B-Abliterated

Quantized

(9)

this model

xyzmean
/

llama-8B-ru

DeepSeek-R1-Distill-Llama-8B-Abliterated-RU

Особенности

Датасет обучения

Model tree for xyzmean/llama-8B-ru

Dataset used to train xyzmean/llama-8B-ru