# Low error rate

Optical Character Recognition (OCR) model optimized for Portuguese text, fine-tuned based on Microsoft's TrOCR base model

Qaari 0.1 Urdu is a model optimized for optical character recognition (OCR) of Urdu text. It is fine-tuned based on Qwen/Qwen2-VL-2B and has significantly improved Urdu OCR capabilities.

An Arabic OCR model fine-tuned based on Qwen2 VL model, optimized for full-page Arabic text recognition

FireRedASR is a series of open-source industrial-grade automatic speech recognition (ASR) models supporting Mandarin, Chinese dialects, and English. It achieves new state-of-the-art (SOTA) performance on public Mandarin ASR benchmarks while excelling in lyrics recognition.

QLoRA fine-tuned version based on LLaMA 3 Youko, specialized in Japanese visual novel English translation tasks

This model is a fine-tuned Sinhala OCR model based on Ransaka/sinhala-ocr-model, primarily used for recognizing Sinhala text.

ArOCR is an Arabic Optical Character Recognition (OCR) model capable of converting images containing Arabic text into editable text format.

This is an automatic speech recognition (ASR) model fine-tuned for Japanese speech recognition tasks based on the facebook/wav2vec2-large-xlsr-53 model.

An audio classification model based on the XLS-R architecture, designed to distinguish between adult and child speech.