Instructions to use speakleash/Bielik-11B-v3.0-DFlash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use speakleash/Bielik-11B-v3.0-DFlash with Transformers:

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

pipe = pipeline("text-generation", model="speakleash/Bielik-11B-v3.0-DFlash")

# Load model directly
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained("speakleash/Bielik-11B-v3.0-DFlash")
model = AutoModel.from_pretrained("speakleash/Bielik-11B-v3.0-DFlash")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use speakleash/Bielik-11B-v3.0-DFlash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "speakleash/Bielik-11B-v3.0-DFlash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "speakleash/Bielik-11B-v3.0-DFlash",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/speakleash/Bielik-11B-v3.0-DFlash

SGLang

How to use speakleash/Bielik-11B-v3.0-DFlash 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 "speakleash/Bielik-11B-v3.0-DFlash" \
    --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": "speakleash/Bielik-11B-v3.0-DFlash",
		"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 "speakleash/Bielik-11B-v3.0-DFlash" \
        --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": "speakleash/Bielik-11B-v3.0-DFlash",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use speakleash/Bielik-11B-v3.0-DFlash with Docker Model Runner:
```
docker model run hf.co/speakleash/Bielik-11B-v3.0-DFlash
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Bielik-11B-v3.0-DFlash

Bielik-11B-v3.0-DFlash is a DFlash draft model designed for use with Bielik-11B-v3.0-Instruct. Its development and training were supported by computational grant PLG/2026/019559 and carried out on the Helios supercomputer, providing access to the advanced infrastructure and computational resources required for large-scale machine learning workloads.

This is not a standalone language model. It is designed to be paired with the target model in a speculative decoding setup.

Model

The model was trained on the same training dataset as the target model, Bielik-11B-v3.0-Instruct. Training was performed using a custom implementation built on top of the ALLaMo framework, developed by Krzysztof Ociepa.

Developed by: SpeakLeash & ACK Cyfronet AGH
Language: Multilingual (32 European languages, optimized for Polish)
Model type: Draft model
Target model: Bielik-11B-v3.0-Instruct
License: Apache License 2.0

Quick Start

This model should be used with an inference server that supports DFlash speculative decoding.

SGLang (version > 0.5.6):

export SGLANG_ALLOW_OVERWRITE_LONGER_CONTEXT_LEN=1

python -m sglang.launch_server \
    --model-path speakleash/Bielik-11B-v3.0-Instruct \
    --speculative-algorithm DFLASH \
    --speculative-draft-model-path speakleash/Bielik-11B-v3.0-DFlash \
    --speculative-num-draft-tokens 16

vLLM (version > 0.20.0):

vllm serve speakleash/Bielik-11B-v3.0-Instruct \
  --speculative-config '{"method": "dflash", "model": "speakleash/Bielik-11B-v3.0-DFlash", "num_speculative_tokens": 15}' \
  --attention-backend flash_attn \
  --max-num-batched-tokens 32768

Credits

Krzysztof Ociepa^SpeakLeash — project lead, concept design, calibration data preparation, training pipeline development, and draft model delivery.
Bielik-11B-v3.0-Instruct development team:
- Krzysztof Ociepa^SpeakLeash - team leadership, conceptualizing, data preparation, process optimization and oversight of training
- Łukasz Flis^{Cyfronet AGH} - coordinating and supervising the training
- Remigiusz Kinas^SpeakLeash - conceptualizing, coordinating RL trainings, data preparation, benchmarking and quantizations
- Adrian Gwoździej^SpeakLeash - data preparation and ensuring data quality
- Krzysztof Wróbel^SpeakLeash - benchmarks

The model could not have been created without the commitment and work of the entire SpeakLeash team, whose contribution is invaluable. Thanks to the hard work of many individuals, it was possible to gather a large amount of content in Polish and establish collaboration between the open-science SpeakLeash project and the HPC center: ACK Cyfronet AGH. Individuals who contributed to the creation of the model: Sebastian Kondracki, Marek Magryś, Sonia Staniek, Waldemar Boszko, Igor Ciuciura.

We gratefully acknowledge Polish high-performance computing infrastructure PLGrid (HPC Center: ACK Cyfronet AGH) for providing computer facilities and support within computational grant no. PLG/2026/019559.

Citation

Please cite this model using the following format:

@misc{ociepa2025bielik11bv3multilingual,
      title={Bielik 11B v3: Multilingual Large Language Model for European Languages}, 
      author={Krzysztof Ociepa and Łukasz Flis and Remigiusz Kinas and Krzysztof Wróbel and Adrian Gwoździej},
      year={2025},
      eprint={2601.11579},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2601.11579}, 
}
@misc{Bielik11Bv3dflash,
    title     = {Bielik-11B-v3.0-DFlash model card},
    author    = {Ociepa, Krzysztof and Flis, Łukasz and Kinas, Remigiusz and Gwoździej, Adrian and Wróbel, Krzysztof and {SpeakLeash Team} and {Cyfronet Team}},
    year      = {2026},
    url       = {https://huggingface.co/speakleash/Bielik-11B-v3.0-DFlash},
    note      = {Accessed: 2026-06-24}, % change this date
    urldate   = {2026-06-24} % change this date
}

Contact Us

If you have any questions, suggestions, or feedback, please use the Discussions tab on Hugging Face. For direct communication, join the SpeakLeash Discord server or connect with the Bielik team on LinkedIn.