Instructions to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="QuantFactory/OmniLing-V1-8b-experimental-GGUF")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("QuantFactory/OmniLing-V1-8b-experimental-GGUF", dtype="auto")

llama-cpp-python

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="QuantFactory/OmniLing-V1-8b-experimental-GGUF",
	filename="OmniLing-V1-8b-experimental.Q2_K.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 QuantFactory/OmniLing-V1-8b-experimental-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF: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 QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF: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 QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M

Use Docker

docker model run hf.co/QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "QuantFactory/OmniLing-V1-8b-experimental-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/OmniLing-V1-8b-experimental-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M

SGLang

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "QuantFactory/OmniLing-V1-8b-experimental-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/OmniLing-V1-8b-experimental-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "QuantFactory/OmniLing-V1-8b-experimental-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/OmniLing-V1-8b-experimental-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with Ollama:
```
ollama run hf.co/QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
```

Unsloth Studio new

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF 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 QuantFactory/OmniLing-V1-8b-experimental-GGUF 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 QuantFactory/OmniLing-V1-8b-experimental-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for QuantFactory/OmniLing-V1-8b-experimental-GGUF to start chatting

Docker Model Runner
How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with Docker Model Runner:
```
docker model run hf.co/QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M
```

Lemonade

How to use QuantFactory/OmniLing-V1-8b-experimental-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull QuantFactory/OmniLing-V1-8b-experimental-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.OmniLing-V1-8b-experimental-GGUF-Q4_K_M

List all available models

lemonade list

QuantFactory/OmniLing-V1-8b-experimental-GGUF

This is quantized version of WoonaAI/OmniLing-V1-8b-experimental created using llama.cpp

Original Model Card

Это экспериментальный релиз, созданный исключительно в исследовательских целях! К сожалению, стабильная работа не гарантируется

Model overview

OmniLing - модель, созданная для переводов между русским и английским языком (пары RU->EN, EN->RU). Эта модель является тонкой настройкой AnatoliiPotapov/T-lite-instruct-0.1. Она находится на одном уровне с крупнейшими проприетарными моделями и может быть использована для качественного перевода художественных текстов, датасетов, сообщений пользователей, описаний и т.п При обучении были использованы собственноручно сгенерированные данные от более сильной модели.

Prompt Template - Llama 3 Instruct

<|start_header_id|>system<|end_header_id|>
You are helpful AI assistant<|eot_id|><|start_header_id|>user<|end_header_id|>
Translate this text from English to Russian:\n\n{text}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

Recommended settings

Очень рекомендуется использовать низкую температуру T=(0.1-0.3)

Не указывайте в system prompt задачу перевести этот текст, там должно быть что-то нейтральное (e.g You are a helpful AI assistant). Указывайте сообщение в user message (см. Prompt Template), модель была обучена именно под такой формат.

Benchmark Results

Оценка проводилась в течение 10-ти итераций, где моделям предлагалось перевести небольшой литературный текст с английского на русский. В качестве модели-судьи выступала Gemini 1.5 pro experimental

AVG:

Критерий	Claude Sonnet 3.5	Gemini 1.5 pro experiment	Gemini 1.5 Flash 8b	GPT4o mini	Gemma 27b	Mistral Nemo 12B	OmniLing-V1-8b
Точность	9.1	8.6	7.4	7.7	7.6	7.0	8.0
Стилистическое соответствие	8.4	8.1	7.0	7.2	6.8	6.3	7.3
Языковая грамотность	9.3	8.9	7.9	8.2	7.6	7.0	8.3
Передача культурных особенностей	9.4	8.9	8.4	8.3	8.3	7.5	8.7
Сохранение литературных приемов	8.3	7.9	7.0	6.9	6.7	6.2	7.1
Ритм и благозвучие	8.3	7.9	7.0	7.2	7.0	6.5	7.3
Сохранение атмосферы	9.0	8.5	7.5	7.9	7.6	7.0	7.9
Адекватность перевода идиом и фразеологизмов	9.1	8.6	7.7	7.9	7.7	7.0	8.0
Сохранение характеров персонажей	9.1	8.9	8.3	8.5	8.2	7.3	8.4
Целостность и связность текста	9.3	9.0	8.2	8.5	8.1	7.4	8.4
ИТОГО	88.7	85.4	76.7	77.6	75.4	70.7	80.1

Полные итоги: https://huggingface.co/WoonaAI/OmniLing-1-8b/blob/main/ENG-RUS-Translation%20simple%20benchmark%20(full).md

Только таблицы всех итераций: https://huggingface.co/WoonaAI/OmniLing-1-8b/blob/main/ENG-RUS-Translation%20simple%20benchmark%20(only%20tables).md

Inference

Example of using model through Transformers library:

Authors

Made by PegasusYaY(AlexBefest) and secretmoon

License

This model is released under the Creative Commons Attribution-NonCommercial 4.0 International License. If you want to use this model in your commercial product, contact with us.

Downloads last month: 81

GGUF

Model size

8B params

Architecture

llama

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

Model tree for QuantFactory/OmniLing-V1-8b-experimental-GGUF

Base model

AnatoliiPotapov/T-lite-instruct-0.1

Quantized

(12)

this model