> For the complete documentation index, see [llms.txt](https://openledger.gitbook.io/openledger/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://openledger.gitbook.io/openledger/model-factory/benchmarks.md).

# Benchmarks

<figure><img src="https://lh7-rt.googleusercontent.com/docsz/AD_4nXdZNjfukZUR19QgeKkVvpypz4k1ShSlNCAJQgs7TS4nHB6jbWWIOoqIjY0TW3842zc0A4c1q3V7aH9iShkXTPmU1uMxc_nuKlRQ1jdoHC9-aBeNu6_kWthIZ8wmcWPshhbEof1h?key=Cxo1mIkiUAwJLq-RqsOyxDS8" alt=""><figcaption><p>Performance Benchmarks</p></figcaption></figure>

Compared to traditional P-Tuning approaches, ModelFactory's LoRA tuning achieves up to 3.7 times faster training speeds while delivering superior performance, as evidenced by improved Rouge scores in advertising text generation tasks. Additionally, by utilizing advanced 4-bit quantization techniques, ModelFactory's QLoRA significantly enhances GPU memory efficiency, ensuring optimized resource utilization.

### Conclusion

ModelFactory bridges the gap between data security and model fine-tuning, making it an indispensable tool for enterprises and researchers. With its modular architecture, robust GUI, integrated chat interface, and RAG attribution, ModelFactory ensures transparency, usability, and scalability within the OpenLedger ecosystem.

<br>