Model Overview

Qwen2.5-Sex is a 1.5 billion parameter language model, derived from the Qwen2.5-1.5B-Instruct architecture. Its primary distinction lies in its fine-tuning process, which involved extensive training on a large collection of Chinese erotic literature and other sensitive datasets. This specialized training makes the model particularly adept at generating text within these specific domains, with a strong emphasis on Chinese language processing due to the nature of its training data.

Key Capabilities

• Specialized Content Generation: Excels at producing text related to erotic literature and sensitive topics, primarily in Chinese.
• Chinese Language Proficiency: Optimized for handling Chinese text, reflecting its dataset composition.
• Extended Context Window: Features a 32768-token context length, allowing for more coherent and extended generations within its specialized domain.

Intended Use and Limitations

This model is explicitly provided for research and testing purposes only. Users are cautioned to adhere to local laws and regulations and are solely responsible for their actions when using the model. The developers disclaim responsibility for any misuse. Further details and updates can be found on the GitHub repository.

Training Data

The fine-tuning leveraged several sensitive datasets, including:

• Bad Data
• Toxic-All
• Erotic Literature Collection