The model is a fine-tuned version of wav2vec2, specifically adapted using the OpenSLR 41 dataset, which is focused on the Javanese language domain. This adaptation enables the model to effectively recognize and process spoken Javanese, leveraging the robust capabilities of the wav2vec2 architecture combined with domain-specific training data.

Intended uses & limitations

This model is intended for transcribing spoken Javanese language from audio recordings. It achieves a Word Error Rate (WER) of 15%, indicating that while the model performs reasonably well, it still produces significant transcription errors. Users should be aware that the accuracy may vary, particularly in cases with challenging audio conditions or less common dialects. Additionally, this model requires input audio at a sample rate of 16kHz, which may limit its applicability for recordings at different sample rates or lower quality audio files.

Training and evaluation data

The model use OpenSLR41 datasets, and split into 2 section (training and testing), then the model is trained using 1xA100 GPU with a training duration of 4-5 hours.

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 0.0001
train_batch_size: 8
eval_batch_size: 8
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 1000
num_epochs: 75
mixed_precision_training: Native AMP

Log Data | Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.6739	2.8329	2000	0.5742	0.4900
0.4767	5.6657	4000	0.4601	0.4101
0.3889	8.4986	6000	0.3921	0.3329
0.3381	11.3314	8000	0.3323	0.3081
0.2842	14.1643	10000	0.3467	0.3081
0.2505	16.9972	12000	0.3186	0.2833
0.2158	19.8300	14000	0.3003	0.2522
0.1885	22.6629	16000	0.2877	0.2405
0.1695	25.4958	18000	0.3089	0.2405
0.1494	28.3286	20000	0.2924	0.2254
0.1331	31.1615	22000	0.2796	0.2068
0.1293	33.9943	24000	0.2734	0.1895
0.1083	36.8272	26000	0.2844	0.1826
0.0955	39.6601	28000	0.2665	0.1744
0.085	42.4929	30000	0.2772	0.1695
0.0799	45.3258	32000	0.2747	0.1654
0.072	48.1586	34000	0.2746	0.1558
0.0934	50.9915	36000	0.2979	0.1764
0.0912	53.8244	38000	0.2914	0.1778
0.0812	56.6572	40000	0.2762	0.1785
0.0779	59.4901	42000	0.2752	0.1688
0.0718	62.3229	44000	0.2623	0.1633
0.0656	65.1558	46000	0.2704	0.1647
0.0606	67.9887	48000	0.2632	0.1571
0.0564	70.8215	50000	0.2711	0.1551
0.0562	73.6544	52000	0.2727	0.1523

How to run (Gradio Web)

python

import torch
import torchaudio
import gradio as gr
import numpy as np
from transformers import pipeline, AutoProcessor, AutoModelForCTC

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Load the model and processor
MODEL_NAME = "<fill this to your model>"
processor = AutoProcessor.from_pretrained(MODEL_NAME)
model = AutoModelForCTC.from_pretrained(MODEL_NAME)

# Move model to GPU
model.to(device)

# Create the pipeline with the model and processor
transcriber = pipeline("automatic-speech-recognition", model=model, tokenizer=processor.tokenizer, feature_extractor=processor.feature_extractor, device=device)

def transcribe(audio):
    sr, y = audio
    y = y.astype(np.float32)
    y /= np.max(np.abs(y))

    return transcriber({"sampling_rate": sr, "raw": y})["text"]

demo = gr.Interface(
    transcribe,
    gr.Audio(sources=["upload"]),
    "text",
)

demo.launch(share=True)

How to run

python

import torch
import torchaudio
import gradio as gr
import numpy as np
from transformers import pipeline, AutoProcessor, AutoModelForCTC

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

# Load the model and processor
MODEL_NAME = "<fill this to actual model>"
processor = AutoProcessor.from_pretrained(MODEL_NAME)
model = AutoModelForCTC.from_pretrained(MODEL_NAME)

# Move model to GPU
model.to(device)

# Load audio file
AUDIO_PATH = "<replace 'path_to_audio_file.wav' with the actual path to your audio file>"
audio_input, sample_rate = torchaudio.load(AUDIO_PATH)

# Ensure the audio is mono (1 channel)
if audio_input.shape[0] > 1:
    audio_input = torch.mean(audio_input, dim=0, keepdim=True)

# Resample audio if necessary
if sample_rate != 16000:
    resampler = torchaudio.transforms.Resample(orig_freq=sample_rate, new_freq=16000)
    audio_input = resampler(audio_input)

# Process the audio input
input_values = processor(audio_input.squeeze(), sampling_rate=16000, return_tensors="pt").input_values

# Move input values to GPU
input_values = input_values.to(device)

# Perform inference
with torch.no_grad():
    logits = model(input_values).logits

# Decode the logits to text
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)[0]

print("Transcription:", transcription)

Framework versions

Transformers 4.44.0
Pytorch 2.2.1+cu118
Datasets 2.20.0
Tokenizers 0.19.1

johaness14/wav2vec2-jv-large-openslr

wav2vec2-jv-large-openslr

Model description