Wav2vec2-large-slavic-parlaspeech-hr Open-source System - Empowering Automatic Speech Recognition of Croatian Parliament

Wav2vec2 Large Slavic Parlaspeech Hr

Developed by classla

This is an automatic speech recognition system for Croatian, based on a Slavic language pre-trained model, specifically optimized for parliamentary speech scenarios

Speech Recognition

Transformers

Other#Croatian speech recognition #Parliament scenario optimization #Slavic language pre-training

Downloads 5,768

Release Time : 4/28/2022

Model Overview

The model is fine-tuned from the facebook/wav2vec2-large-slavic-voxpopuli-v2 pre-trained model using 300 hours of Croatian parliamentary speech data from ParlaSpeech-HR v1.0, specifically designed for speech recognition tasks in Croatian parliamentary settings

Model Features

Slavic language pre-training

Fine-tuned from a Slavic language pre-trained model, providing better adaptability for Croatian

Parliamentary speech optimization

Specifically optimized for the acoustic characteristics of Croatian parliamentary scenarios

High performance metrics

Achieves 2.22% character error rate and 6.79% word error rate on test sets

Model Capabilities

Croatian speech recognition

Parliamentary speech transcription

Long audio processing

Use Cases

Government agencies

Parliament meeting minutes

Automatically transcribe Croatian parliamentary meeting content

Efficiently generates meeting transcripts with over 93% accuracy

Academic research

Political discourse analysis

Provide political scientists with textual data of parliamentary speeches

Supports large-scale political discourse analysis research

🚀 wav2vec2-large-slavic-parlaspeech-hr

This model is designed for Croatian Automatic Speech Recognition (ASR). It builds upon the facebook/wav2vec2-large-slavic-voxpopuli-v2 model and has been fine - tuned using 300 hours of recordings and transcripts from the ASR Croatian parliament dataset ParlaSpeech - HR v1.0.

🚀 Quick Start

If you use this model, please cite the following paper:

Nikola Ljubešić, Danijel Koržinek, Peter Rupnik, Ivo - Pavao Jazbec. ParlaSpeech - HR -- a freely available ASR dataset for Croatian bootstrapped from the ParlaMint corpus. http://www.lrec-conf.org/proceedings/lrec2022/workshops/ParlaCLARINIII/pdf/2022.parlaclariniii-1.16.pdf

✨ Features

Metrics

Evaluation is performed on the dev and test portions of the ParlaSpeech - HR v1.0 dataset.

Split	CER	WER
dev	0.0311	0.0921
test	0.0222	0.0679

There are multiple models available, and in terms of CER and WER, the best - performing model is wav2vec2-large-slavic-parlaspeech-hr-lm.

💻 Usage Examples

Basic Usage

from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
import soundfile as sf
import torch
import os
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
# load model and tokenizer
processor = Wav2Vec2Processor.from_pretrained(
    "classla/wav2vec2-large-slavic-parlaspeech-hr")
model = Wav2Vec2ForCTC.from_pretrained("classla/wav2vec2-large-slavic-parlaspeech-hr")
# download the example wav files:
os.system("wget https://huggingface.co/classla/wav2vec2-large-slavic-parlaspeech-hr/raw/main/00020570a.flac.wav")
# read the wav file 
speech, sample_rate = sf.read("00020570a.flac.wav")
input_values = processor(speech, sampling_rate=sample_rate, return_tensors="pt").input_values.to(device)
# remove the raw wav file
os.system("rm 00020570a.flac.wav")
# retrieve logits
logits = model.to(device)(input_values).logits
# take argmax and decode
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.decode(predicted_ids[0]).lower()
# transcription: 'veliki broj poslovnih subjekata posluje sa minusom velik dio'

🔧 Technical Details

Training hyperparameters

In fine - tuning, the following arguments were used:

Property	Details
`per_device_train_batch_size`	16
`gradient_accumulation_steps`	4
`num_train_epochs`	8
`learning_rate`	3e - 4
`warmup_steps`	500

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご