Instructions to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with Transformers:

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

pipe = pipeline("text-generation", model="4eJIoBek/ruGPT3_small_nujdiki_stage1")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("4eJIoBek/ruGPT3_small_nujdiki_stage1")
model = AutoModelForCausalLM.from_pretrained("4eJIoBek/ruGPT3_small_nujdiki_stage1", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
# Run inference directly in the terminal:
llama cli -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
# Run inference directly in the terminal:
llama cli -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

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 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

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 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

Use Docker

docker model run hf.co/4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

LM Studio
Jan

vLLM

How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "4eJIoBek/ruGPT3_small_nujdiki_stage1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "4eJIoBek/ruGPT3_small_nujdiki_stage1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

SGLang

How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 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 "4eJIoBek/ruGPT3_small_nujdiki_stage1" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "4eJIoBek/ruGPT3_small_nujdiki_stage1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "4eJIoBek/ruGPT3_small_nujdiki_stage1" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "4eJIoBek/ruGPT3_small_nujdiki_stage1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Ollama
How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with Ollama:
```
ollama run hf.co/4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
```
Unsloth Desktop
Docker Model Runner
How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with Docker Model Runner:
```
docker model run hf.co/4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0
```

Lemonade

How to use 4eJIoBek/ruGPT3_small_nujdiki_stage1 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull 4eJIoBek/ruGPT3_small_nujdiki_stage1:Q8_0

Run and chat with the model

lemonade run user.ruGPT3_small_nujdiki_stage1-Q8_0

List all available models

lemonade list

Atomic Chat

это файнтюн sberai ruGPT3 small (125 млн параметров) на всех оригинальных нуждиках (2 часа транскрибированные через openai whisper medium). размер блока при файнтюне 1024, 25 эпох. все скрипты по инференсу модели тут https://github.com/ai-forever/ru-gpts, через transformers вполне себе работает на 4 гб видеопамяти, на 2 думаю тоже заработает.

-как запустить через transformers? запускаем строки ниже в jupyterе

from transformers import pipeline, set_seed

set_seed(32) generator = pipeline('text-generation', model="4eJIoBek/ruGPT3_small_nujdiki_fithah", do_sample=True, max_length=350) generator("Александр Сергеевич Пушкин известен также благодаря своим сказкам, которые включают в себя: ")

и всё работает и вообще нихуёво

Downloads last month: 223

Safetensors

Model size

0.2B params

Tensor type

F32

4eJIoBek
/

ruGPT3_small_nujdiki_stage1

Space using 4eJIoBek/ruGPT3_small_nujdiki_stage1 1