Instructions to use JetBrains/Mellum-4b-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use JetBrains/Mellum-4b-base with Transformers:

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

pipe = pipeline("text-generation", model="JetBrains/Mellum-4b-base")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("JetBrains/Mellum-4b-base")
model = AutoModelForCausalLM.from_pretrained("JetBrains/Mellum-4b-base")

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use JetBrains/Mellum-4b-base with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/JetBrains/Mellum-4b-base

SGLang

How to use JetBrains/Mellum-4b-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 "JetBrains/Mellum-4b-base" \
    --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": "JetBrains/Mellum-4b-base",
		"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 "JetBrains/Mellum-4b-base" \
        --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": "JetBrains/Mellum-4b-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use JetBrains/Mellum-4b-base with Docker Model Runner:
```
docker model run hf.co/JetBrains/Mellum-4b-base
```

This this what gets downloaded in IntelliJ? Why is it so much slower?

by deusaquilus - opened Aug 8, 2025

Discussion

deusaquilus

Aug 8, 2025

Is this the same thing as the model that get's downloaded in IntelliJ when I go to Tools -> AI Assistant and download the local model?
Why is so much slower when I run it in Ollama?

zerogerc

Aug 25, 2025

Hi @deusaquilus , not sure I understood your question correctly, but if you’re referring to inline completion local models (full line), then no, they’re not the same. The local models we currently use are much smaller than Mellum - around 100-400M parameters.

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment