# ONNX optimization

This is a quantized model based on ONNX, which is the uint8 quantized version of Qwen/Qwen3-Embedding-0.6B. It reduces the model size while maintaining retrieval performance.

ONNX format version of DINOv2-small, suitable for vision tasks

This is the ONNX version of the AST (Audio Spectrogram Transformer) model, designed specifically for audio classification tasks and fine-tuned on the AudioSet dataset.

RF-DETR Large is an object detection model based on the DETR architecture, designed for efficient detection of objects in images.

A compact zero-shot classification model based on DeBERTa architecture, optimized for long-text natural language inference tasks, converted to ONNX format for web compatibility

YOLOv10 is a real-time end-to-end object detection model with efficient latency-accuracy and size-accuracy trade-offs.

YOLOv9-C is an object detection model based on the YOLOv9 architecture, suitable for real-time detection of multiple objects in images.

ONNX version of the Dutch sentence transformer based on the RobBERT model, mapping text to a 768-dimensional vector space, suitable for semantic search and clustering tasks

Whisper-large-v3 is an open-source automatic speech recognition (ASR) model by OpenAI, supporting speech-to-text tasks in multiple languages.

Natural language inference model based on DeBERTa-v3-large architecture, ONNX weight version adapted for Transformers.js

Natural language inference model based on DeBERTa-v3 architecture, ONNX weight version adapted for Transformers.js

A zero-shot classification model based on DeBERTa-v3, supporting multi-task natural language inference (MNLI), FEVER, and ANLI tasks, converted to ONNX format for compatibility with Transformers.js

This is a multilingual sentence embedding model that maps text to a 512-dimensional vector space, suitable for semantic search and clustering tasks.

A lightweight version of DistilBERT specifically designed for Multi-Genre Natural Language Inference (MNLI) tasks, based on the BERT architecture but with fewer parameters and higher efficiency.

A lightweight text summarization model based on the BART architecture, compressing the original model size through knowledge distillation while retaining core summarization capabilities

This is a sentence transformer model based on the MiniLM-L6 architecture, capable of mapping text to a 384-dimensional vector space, suitable for semantic search and text similarity calculation tasks.

A translation model specifically designed for converting Japanese text in the game 'Final Fantasy XIV' into Korean