Instructions to use Data-Dream/CryChic-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Data-Dream/CryChic-v2 with Transformers:

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

pipe = pipeline("text-generation", model="Data-Dream/CryChic-v2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Data-Dream/CryChic-v2")
model = AutoModelForCausalLM.from_pretrained("Data-Dream/CryChic-v2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Data-Dream/CryChic-v2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Data-Dream/CryChic-v2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Data-Dream/CryChic-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Data-Dream/CryChic-v2

SGLang

How to use Data-Dream/CryChic-v2 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 "Data-Dream/CryChic-v2" \
    --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": "Data-Dream/CryChic-v2",
		"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 "Data-Dream/CryChic-v2" \
        --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": "Data-Dream/CryChic-v2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Data-Dream/CryChic-v2 with Docker Model Runner:
```
docker model run hf.co/Data-Dream/CryChic-v2
```

CryChic - Music Generation Model

Github | Handbook | Website | Contact | Twitter Account

Model Description

CryChic-v2 is a lightweight, transformer-based model designed for generating short, melodic musical pieces. It is optimized for performance in resource-constrained environments, such as mobile devices or embedded systems.

How to Use

You can generate music by prompt.

Limitations and Bias

CryChic-v2 operates within certain creative constraints:

Primarily generates melodies up to 30 seconds.
Better performance with genres like Classical and Jazz due to training data limitations.

Users are encouraged to be aware of these limitations when evaluating the output.

Training Procedure

The model was trained on a dataset comprising diverse musical genres but heavily features classical and jazz pieces, which might influence its generative style.

Training Data

The model was trained using a proprietary dataset of labeled melodies that include a variety of musical styles.

Ethical Considerations

While CryChic-v2 is designed for creativity, it should be used responsibly. The model is not capable of replicating specific artists' styles without explicit conditioning and should not be used to generate deceptive or misleading content.

Citing CryChic-v2

If you use this model in your research, please cite it as follows:

@misc{crychicv2,
title={CryChic-v2: A Lightweight Music Generation Model},
author={Data Dream},
year={2025},
howpublished={Hugging Face},
}

Downloads last month: 2

Safetensors

Model size

8B params

Tensor type

BF16