Deepseek R1 0528 Qwen3 8B GPTQ Int4 Int8Mix

🚀 DeepSeek-R1-0528-Qwen3-8B-GPTQ-Int4-Int8Mix

This is a quantized version based on the deepseek-ai/DeepSeek-R1-0528-Qwen3-8B.

📅【Model Update Date】

2025-05-29
1. First commit

📦【Dependencies】

vllm==0.9.0
transformers==4.52.3

⚠️ Important Note

💡【Notes for the New Version of VLLM】

1. It is recommended to use the V0 inference mode

Before starting vllm, set the environment variable first:

export VLLM_USE_V1=0

📋【Model List】

💾【Model Download】

from modelscope import snapshot_download
snapshot_download('tclf90/DeepSeek-R1-0528-Qwen3-8B-GPTQ-Int4-Int8Mix', cache_dir="Local path")

🌟【Introduction】

Paper Link👁️

1. Introduction

The DeepSeek R1 model has undergone a minor version upgrade, with the current version being DeepSeek-R1-0528. In the latest update, DeepSeek R1 has significantly improved its depth of reasoning and inference capabilities by leveraging increased computational resources and introducing algorithmic optimization mechanisms during post-training. The model has demonstrated outstanding performance across various benchmark evaluations, including mathematics, programming, and general logic. Its overall performance is now approaching that of leading models, such as O3 and Gemini 2.5 Pro.

Compared to the previous version, the upgraded model shows significant improvements in handling complex reasoning tasks. For instance, in the AIME 2025 test, the model’s accuracy has increased from 70% in the previous version to 87.5% in the current version. This advancement stems from enhanced thinking depth during the reasoning process: in the AIME test set, the previous model used an average of 12K tokens per question, whereas the new version averages 23K tokens per question.

Beyond its improved reasoning capabilities, this version also offers a reduced hallucination rate, enhanced support for function calling, and better experience for vibe coding.

2. Evaluation Results

DeepSeek-R1-0528

For all our models, the maximum generation length is set to 64K tokens. For benchmarks requiring sampling, we use a temperature of $0.6$, a top-p value of $0.95$, and generate 16 responses per query to estimate pass@1.

DeepSeek-R1-0528-Qwen3-8B

Meanwhile, we distilled the chain - of - thought from DeepSeek - R1 - 0528 to post - train Qwen3 8B Base, obtaining DeepSeek - R1 - 0528 - Qwen3 - 8B. This model achieves state - of - the - art (SOTA) performance among open - source models on the AIME 2024, surpassing Qwen3 8B by +10.0% and matching the performance of Qwen3 - 235B - thinking. We believe that the chain - of - thought from DeepSeek - R1 - 0528 will hold significant importance for both academic research on reasoning models and industrial development focused on small - scale models.

📄 License

This code repository is licensed under MIT License. The use of DeepSeek - R1 models is also subject to MIT License. DeepSeek - R1 series (including Base and Chat) supports commercial use and distillation.

📚 Citation

@misc{deepseekai2025deepseekr1incentivizingreasoningcapability,
      title={DeepSeek - R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning}, 
      author={DeepSeek - AI},
      year={2025},
      eprint={2501.12948},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2501.12948}, 
}

📞 Contact

If you have any questions, please raise an issue or contact us at service@deepseek.com.