# Distilled Model

DeepSeek-R1-Distill-Qwen-7B is a distilled large language model released by deepseek-ai, based on the Qwen-7B architecture, suitable for text generation tasks.

A distilled French version of Whisper-Large-V3, optimized for inference efficiency by reducing decoder layers while maintaining good performance

DistilViT is an image caption generation model based on Vision Transformer (ViT) and distilled GPT-2, capable of converting images into textual descriptions.

A German speech recognition model based on distil-whisper technology, with 756 million parameters, achieving faster inference speeds while maintaining high quality.

Distil-Whisper is a knowledge-distilled version of Whisper large-v3, focusing on English automatic speech recognition, offering faster inference speeds while maintaining accuracy close to the original model.

A distilled model based on BERT-small architecture, specifically designed for Czech-English semantic embedding

A lightweight sentence similarity model optimized for financial domain question matching, compressing 12-layer BERT to 4 layers through distillation technology, significantly improving inference efficiency

An audio classification model based on the wav2vec 2.0 architecture, designed to distinguish between adult and child voices

DistilCamemBERT is a distilled version of the French CamemBERT model, significantly reducing model complexity while maintaining performance through knowledge distillation techniques.

DistilBERT is a distilled version of the BERT base model, maintaining similar performance while being more lightweight and efficient, suitable for natural language processing tasks such as sequence classification and token classification.

This is a lightweight version of distilbert-base-multilingual-cased, focused on bilingual processing of English and German, maintaining the representation capability and accuracy of the original model.

This is a distilled version of distilbert-base-multilingual-cased, specifically optimized for English and Arabic processing while maintaining the original model's accuracy.