Instructions to use elte-nlp/Racka-4B-MLX-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use elte-nlp/Racka-4B-MLX-4bit with MLX:

# Make sure mlx-lm is installed
# pip install --upgrade mlx-lm

# Generate text with mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("elte-nlp/Racka-4B-MLX-4bit")

prompt = "Write a story about Einstein"
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
    messages, add_generation_prompt=True
)

text = generate(model, tokenizer, prompt=prompt, verbose=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings
LM Studio

How to use elte-nlp/Racka-4B-MLX-4bit with Pi:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "elte-nlp/Racka-4B-MLX-4bit"

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "mlx-lm": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "elte-nlp/Racka-4B-MLX-4bit"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use elte-nlp/Racka-4B-MLX-4bit with Hermes Agent:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "elte-nlp/Racka-4B-MLX-4bit"

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default elte-nlp/Racka-4B-MLX-4bit

Run Hermes

hermes

OpenClaw new

How to use elte-nlp/Racka-4B-MLX-4bit with OpenClaw:

Start the MLX server

# Install MLX LM:
uv tool install mlx-lm
# Start a local OpenAI-compatible server:
mlx_lm.server --model "elte-nlp/Racka-4B-MLX-4bit"

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "elte-nlp/Racka-4B-MLX-4bit" \
  --custom-provider-id mlx-lm \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

MLX LM

How to use elte-nlp/Racka-4B-MLX-4bit with MLX LM:

Generate or start a chat session

# Install MLX LM
uv tool install mlx-lm
# Interactive chat REPL
mlx_lm.chat --model "elte-nlp/Racka-4B-MLX-4bit"

Run an OpenAI-compatible server

# Install MLX LM
uv tool install mlx-lm
# Start the server
mlx_lm.server --model "elte-nlp/Racka-4B-MLX-4bit"
# Calling the OpenAI-compatible server with curl
curl -X POST "http://localhost:8000/v1/chat/completions" \
   -H "Content-Type: application/json" \
   --data '{
     "model": "elte-nlp/Racka-4B-MLX-4bit",
     "messages": [
       {"role": "user", "content": "Hello"}
     ]
   }'

Access Request for Research-Only Model

Please provide your professional details and acknowledge the terms of use to request access.

By requesting access, you acknowledge that this model is provided solely for research purposes, is offered 'as-is' without any guarantees, and cannot be utilized for for-profit tasks or commercial applications.

Racka-4B-MLX-4bit Model Card

Racka

Racka (Regionális Adatokon Célzottan Kialakított Alapmodell) is a continually pretrained large language model designed to bridge the resource gap between Hungarian and high-resource languages. It employs parameter-efficient continual pretraining via Low-Rank Adaptation (LoRA) on a Qwen3-4B (reasoning/instruct) backbone.

The model was trained on a mixture of 160B tokens (44% Hungarian, 24% German, 21% English, 11% Code) on the Komondor HPC. To better match the training distribution, Racka uses an adapted tokenizer that achieves substantially improved tokenization fertility for Hungarian while maintaining competitive performance in English and German.

Note: This is an MLX-quantized model, converted for use on Apple Silicon. The full-precision model is available at elte-nlp/Racka-4B. GGUF conversions are available at elte-nlp/Racka-4B-GGUF.

We would like to thank Raphael Várszegi for contributing the MLX model conversions.

To learn more about the details of our project, please check our FAQ.

Model Details

Model: elte-nlp/Racka-4B
Quantization: MLX affine, 4-bit weight codes
Intended use case: Low-complexity edge tasks, optimized few-shot pipelines, task-specific Hungarian fine-tuning.

MLX Conversion

Generated with: MLX 0.32.0, MLX-LM 0.31.3
Method: affine weight quantization, group size 64
Approximate size: 2.1 GB
MLX-reported effective bits/weight: 4.501

The effective bits/weight value includes per-group scale and bias metadata. The weight codes themselves use the bit width shown in the repo name.

Other MLX variants

Repository	Format	Approximate size	Effective bits/weight
Racka-4B-MLX-3bit	MLX affine, 3-bit	1.7 GB	3.501
Racka-4B-MLX-4bit	MLX affine, 4-bit	2.1 GB	4.501
Racka-4B-MLX-5bit	MLX affine, 5-bit	2.6 GB	5.501
Racka-4B-MLX-6bit	MLX affine, 6-bit	3.1 GB	6.500
Racka-4B-MLX-8bit	MLX affine, 8-bit	4.0 GB	8.500
Racka-4B-MLX-16bit	MLX float16, unquantized	7.5 GB	16

Benchmarks

Limitations

The model is capable of both instruction following chat and English reasoning using the original Qwen settings, this is a preserved capability with no direct training targetting this functionality.
The model has not been aligned and is unsafe for use with end-users.
This model is only to be used for research purposes, commercial or for-profit usage is not permitted.

Team

In alphabetical order:

Zsolt Csibi (ELTE-IK, AI Dept.)
Bence Gortka (ELTE-BTK, DH-Lab)
Natabara Gyöngyössy (ELTE-IK, AI Dept.)
Kornél Nagy (ELTE-BTK, DH-Lab)
Dávid Nemeskey (ELTE-BTK, DH-Lab)
Gábor Palkó (ELTE-BTK, DH-Lab)
Martin Sallai (ELTE-BTK, DH-Lab)
András Simonyi (ELTE-IK, AI Dept.)
András Szekeres (ELTE-BTK, DH-Lab)

Acknowledgements

We acknowledge the Digital Government Development and Project Management Ltd. for awarding us access to the Komondor HPC facility based in Hungary.

This research was supported by the EKÖP-24 University Excellence Scholarship Program of the Ministry for Culture and Innovation, funded by the National Research, Development and Innovation Fund.

The authors acknowledge the support of the National Laboratory for Digital Heritage. Project no. 2022-2.1.1-NL-2022-00009 has been implemented with the support provided by the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund, financed under the 2022-2.1.1-NL funding scheme.

We would like to thank Levente Szabados for the name idea and initial informal discussions.

Citation

@article{racka2026,
  title={Racka: Efficient Hungarian LLM Adaptation on Academic Infrastructure},
  author={Csibi, Zsolt and Gortka, Bence Gy\"orgy and Nagy, Korn\'el and Nemeskey, D\'avid M\'ark and Sallai, Martin and Simonyi, Andr\'as and Szekeres, Andr\'as M\'ark and Palk\'o, G\'abor},
  journal={Proceedings of the XXII. Hungarian Computational Linguistics Conference},
  year={2026}
}

FAQ

Q: Is this model intended for use as a general-purpose chat agent?

A: No. While Racka-4B currently stands as one of the highest-scoring small/edge models on Hungarian benchmarks, it remains a compact, edge-oriented model. Conducting a full language adaptation for larger architectures would demand computing infrastructure beyond the Komondor HPC (AI training is just really expensive on the large scale).

Our team is actively working on adapting larger models; in the meantime, we recommend utilizing models of at least 27B–30B parameters for complex or resource-heavy reasoning tasks.

Q: Why is this a gated-access model with a restricted license?

A: Due to Hungarian and EU regulations, as well as the licensing terms of our source data, certain components of our training corpus cannot be used to generate derivative works for commercial or for-profit activities.

The volume of globally available, high-quality Hungarian text data remains limited. Restricting our training data exclusively to open, unrestricted corpora would not have yielded a sufficient dataset for effective language adaptation. Consequently, we require users to fill in an access agreement confirming the model will be used strictly for non-profit purposes. We believe the academic and non-profit communities will still derive value from this work.

Q: What kind of financial funding did this project receive?

A: As noted in our acknowledgements, our team consists of AI researchers and enthusiasts who volunteered their time. We did not receive dedicated financial funding or grants to build this model. Instead, the project was made possible through institutional and operational support:

Worktime Allocations & Academic Grants: Several ELTE researchers and PhD students were permitted to dedicate official university hours and existing fellowship time to this initiative.
Compute & HPC Access: Our academic project proposal was accepted by the Komondor HPC team (this is a public opportunity available to all Hungarian academics), granting us vital GPU time. The facility's engineering team also provided essential, specialized HPC support.
Voluntary Contribution: Driven by a shared passion for open-source NLP, project members frequently dedicated their personal time to bring this model to completion.

Q: How would you improve on your model?

A: The current model has undergone continual pre-training during language adaptation, but it has not been through any SFT or alignment steps, which are required for highly capable models. Our aim is to translate, create and synthetize higher complexity datasets as part of our next projects. This will allow us to train models that are more capable in everyday use cases.