Instructions to use Kingrane/Bobik-2B-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Kingrane/Bobik-2B-Base with Transformers:

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

pipe = pipeline("text-generation", model="Kingrane/Bobik-2B-Base")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Kingrane/Bobik-2B-Base", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Kingrane/Bobik-2B-Base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Kingrane/Bobik-2B-Base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Kingrane/Bobik-2B-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Kingrane/Bobik-2B-Base

SGLang

How to use Kingrane/Bobik-2B-Base 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 "Kingrane/Bobik-2B-Base" \
    --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": "Kingrane/Bobik-2B-Base",
		"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 "Kingrane/Bobik-2B-Base" \
        --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": "Kingrane/Bobik-2B-Base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Desktop
Docker Model Runner
How to use Kingrane/Bobik-2B-Base with Docker Model Runner:
```
docker model run hf.co/Kingrane/Bobik-2B-Base
```

Bobik-2B-Base

Bobik-2B-Base - это русская дообученная языковая модель на базе Qwen3.5-2B. Создана разработчиком Kingrane как продолжение претрейна на русских данных.

Это BASE модель, не чат-бот. Она умеет дописывать текст, но не умеет отвечать на инструкции. Для чата используй Kingrane/Bobik-2B.

Обучение

База: unsloth/Qwen3.5-2B
Метод: Continued Pretraining (CPT) + LoRA r=32
Датасет: 50k статей русской Википедии wikimedia/wikipedia 20231101.ru
Шаги: 1500
Железо: Google Colab T4

Как использовать

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("Kingrane/Bobik-2B-Base")
tokenizer = AutoTokenizer.from_pretrained("Kingrane/Bobik-2B-Base")

text = "Квантовая физика это"
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.decode(outputs[0]))