Deepseek R1 0528 FP4

🚀 NVIDIA DeepSeek-R1-0528-FP4 Model

The NVIDIA DeepSeek-R1-0528-FP4 is a quantized version of the DeepSeek R1 0528 model, offering efficient text generation capabilities.

🚀 Quick Start

The NVIDIA DeepSeek-R1-0528-FP4 model is the quantized version of the DeepSeek AI's DeepSeek R1 0528 model, which is an auto - regressive language model that uses an optimized transformer architecture. For more information, please check here. The NVIDIA DeepSeek R1 FP4 model is quantized with TensorRT Model Optimizer. This model is ready for commercial/non - commercial use.

✨ Features

• Quantized Model: The model is quantized to FP4 data type, reducing disk size and GPU memory requirements.
• Transformer Architecture: Utilizes an optimized transformer architecture for text generation.
• Ready for Use: Suitable for both commercial and non - commercial applications.

📦 Installation

Not provided in the original document, so this section is skipped.

💻 Usage Examples

Basic Usage

To deploy the quantized FP4 checkpoint with TensorRT-LLM LLM API, follow the sample codes below (you need 8xB200 GPU and TensorRT LLM built from source with the latest main branch):

from tensorrt_llm import SamplingParams
from tensorrt_llm._torch import LLM

def main():

    prompts = [
        "Hello, my name is",
        "The president of the United States is",
        "The capital of France is",
        "The future of AI is",
    ]
    sampling_params = SamplingParams(max_tokens=32)

    llm = LLM(model="nvidia/DeepSeek-R1-0528-FP4", tensor_parallel_size=8, enable_attention_dp=True)

    outputs = llm.generate(prompts, sampling_params)

    # Print the outputs.
    for output in outputs:
        prompt = output.prompt
        generated_text = output.outputs[0].text
        print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")


# The entry point of the program need to be protected for spawning processes.
if __name__ == '__main__':
    main()

Evaluation

The accuracy benchmark results are presented in the table below:

📚 Documentation

Model Architecture

Input

Output

Software Integration

Model Version

The model is quantized with nvidia - modelopt v0.31.0

Training, Testing, and Evaluation Datasets

For training, testing, and evaluation datasets, the data collection method by dataset is Hybrid: Human, Automated, and the labeling method by dataset is also Hybrid: Human, Automated.

Calibration Datasets

• Calibration Dataset: cnn_dailymail
• Data collection method: Automated.
• Labeling method: Undisclosed.

Inference

Post Training Quantization

This model was obtained by quantizing the weights and activations of DeepSeek R1 to FP4 data type, ready for inference with TensorRT - LLM. Only the weights and activations of the linear operators within transformer blocks are quantized. This optimization reduces the number of bits per parameter from 8 to 4, reducing the disk size and GPU memory requirements by approximately 1.6x.

Model Limitations

The base model was trained on data that contains toxic language and societal biases originally crawled from the internet. Therefore, the model may amplify those biases and return toxic responses especially when prompted with toxic prompts. The model may generate answers that may be inaccurate, omit key information, or include irrelevant or redundant text producing socially unacceptable or undesirable text, even if the prompt itself does not include anything explicitly offensive.

Ethical Considerations

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. When downloaded or used in accordance with our terms of service, developers should work with their internal model team to ensure this model meets requirements for the relevant industry and use case and addresses unforeseen product misuse.

Please report security vulnerabilities or NVIDIA AI Concerns [here](https://www.nvidia.com/en - us/support/submit - security - vulnerability/).

License/Terms of Use

MIT