# ONNX Optimization

NVIDIA Parakeet TDT 0.6B V2 is a model based on automatic speech recognition (ASR) tasks, suitable for English speech-to-text tasks.

This is the ONNX format version of the facebook/dinov2-base model, suitable for computer vision tasks.

This is a multilingual detection model based on XLM-RoBERTa, supporting recognition of 20 languages.

Moonshine Tiny is a lightweight automatic speech recognition (ASR) model suitable for embedded devices and edge computing scenarios.

TestSavantAI models are a set of fine-tuned classifiers specifically designed to defend against prompt injection and jailbreak attacks targeting large language models (LLMs).

The TestSavantAI model is a set of classifiers specifically designed to defend against prompt injection and jailbreak attacks in large language models (LLMs). The tiny version is based on the BERT-tiny architecture, balancing security and computational efficiency.

A time series forecasting model based on the PatchTSMixer architecture, developed by IBM, suitable for multivariate time series forecasting tasks.

This is the ONNX format conversion of Microsoft's SpeechT5 text-to-speech (TTS) model, optimized for Transformers.js

An ONNX-optimized Whisper large speech recognition model designed for web deployment

This is an ONNX-optimized version of the facebook/bart-large-cnn model, primarily used for text summarization tasks.

TimeSformer-HR is a high-resolution spatiotemporal Transformer model for video, fine-tuned on the Kinetics-400 dataset, suitable for video action recognition tasks.

The ONNX O4 version of BGE-RERANKER-V2 is an optimized text reordering model that supports relevance scoring for multilingual text pairs.

Depth-Anything-V2-Base is an ONNX-format depth estimation model adapted for Transformers.js, designed for image depth estimation on the web.

This is the ONNX-converted version of the lakshyakh93/deberta_finetuned_pii model, designed to identify Personally Identifiable Information (PII) in text.

MusicGen Small is a Transformer-based music generation model capable of producing high-quality music clips from text descriptions.

Object detection model based on YOLOv9, adapted for Transformers.js, capable of running in a browser

BGE-M3 is an embedding model that supports dense retrieval, lexical matching, and multi-vector interaction, converted to ONNX format for compatibility with frameworks like ONNX Runtime.

This is the ONNX quantized version of the BAAI/bge-m3 model, supporting three functionalities: dense retrieval, multi-vector retrieval, and sparse retrieval, covering over 100 languages.

GTE-Base is a general-purpose text embedding model capable of converting text into high-dimensional vector representations, suitable for text classification and similarity search tasks.

A multilingual detection model based on XLM-RoBERTa, capable of identifying the language category of text.

Chinese CLIP model based on ViT architecture, supporting multimodal understanding of images and text

A voice command recognition model based on the HuBERT architecture, optimized for keyword spotting tasks

This is a multilingual sentence transformer model that maps text to a dense vector space, supporting semantic search and clustering tasks

This is the ONNX format conversion of the papluca/xlm-roberta-base-language-detection model, designed for multilingual text classification tasks, supporting detection in 20 languages.

This is the ONNX-converted version of the deepset/deberta-v3-base-injection model for detecting prompt injection attacks.

Nougat is a vision-based academic document understanding model capable of converting scientific PDF images into Markdown-formatted text.

A classical image super-resolution model based on Swin2SR architecture, capable of upscaling image resolution by 4 times

This is a sentence transformer model that maps sentences and paragraphs into a dense vector space, suitable for tasks such as clustering and semantic search.

Whisper is a pre-trained automatic speech recognition (ASR) and speech translation model, trained on 680,000 hours of labeled data, demonstrating strong generalization capabilities. This repository contains the INT4 weight-only quantized version of the Whisper large v2 model in ONNX format.

Whisper is a pre-trained model for automatic speech recognition (ASR) and speech translation. This repository provides the Whisper large model in ONNX format with INT4 weight quantization, powered by Intel® Neural Compressor and Intel® Transformers Extension.

Quantized (INT8) ONNX variant of BGE-large-en-v1.5 with inference acceleration via DeepSparse

OpenAI's open-source CLIP model, based on Vision Transformer (ViT) architecture, supporting joint understanding of images and text.

Large-scale speech recognition model based on the wav2vec 2.0 architecture, supporting English speech-to-text conversion

CLIP model developed by OpenAI, based on Vision Transformer architecture, supporting joint understanding of images and text

OpenAI's open-source CLIP model, based on Vision Transformer architecture, supporting cross-modal understanding of images and text

This is an ONNX-converted model based on sentence-transformers/all-MiniLM-L6-v2, capable of mapping sentences and paragraphs into a 384-dimensional dense vector space, suitable for tasks like clustering or semantic search.

Image classification model based on Transformer architecture, pre-trained and fine-tuned on ImageNet-21k and ImageNet-1k datasets