# Parameter Efficiency

A fine-tuned version based on the Qwen2.5-3B-Instruct model on medical and health datasets, designed to generate medical-related responses and provide health guidance to users.

Compressible version of Qwen2.5-7B provided by the ACIP project, supporting dynamic compression rate adjustment while maintaining model performance

Compressible version of Llama-7B model provided by the ACIP project, supports dynamic adjustment of compression ratio

This model is an efficient fine-tuning model optimized based on the TRL (Transformer Reinforcement Learning) library, focusing on the application of reinforcement learning in Transformer models.

TextNet is a lightweight and efficient architecture specifically designed for text detection, achieving an excellent balance between detection accuracy and inference speed through three variants.

This is a dialogue summarization model based on the T5-base architecture, fine-tuned using LoRA technology, specifically designed for efficient dialogue summarization.

Chinese ELECTRA is a pre-trained model developed by the Harbin Institute of Technology-iFLYTEK Joint Lab based on Google's ELECTRA architecture, with only 1/10 the parameters of BERT but comparable performance.

Chinese ELECTRA is a pre-trained model developed by the HIT-iFLYTEK Joint Lab based on Google's ELECTRA model, featuring a small parameter size but superior performance.