Wav2vec2 Base 760 Turkish

🚀 Wav2Vec2-Base-760-Turkish

A pre - trained Turkish model fine - tuned on the Common Voice dataset for automatic speech recognition.

🚀 Quick Start

This is a pre - trained Turkish model ceyda/wav2vec2-base-760. It has been fine - tuned on Turkish using the Common Voice. When using this model, ensure that your speech input is sampled at 16kHz.

✨ Features

• Dataset: Fine - tuned on the Common Voice Turkish dataset.
• Metrics: Evaluated with Word Error Rate (WER).
• License: Licensed under the Apache 2.0 license.

📦 Installation

No specific installation steps are provided in the original document.

💻 Usage Examples

Basic Usage

The model can be used directly (without a language model) as follows:

import torch
import torchaudio
from datasets import load_dataset
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor

test_dataset = load_dataset("common_voice", "tr", split="test[:2%]") 

processor = Wav2Vec2Processor.from_pretrained("ceyda/wav2vec2-base-960-turkish")
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-base-960-turkish")

resampler = torchaudio.transforms.Resample(48_000, 16_000)

# Preprocessing the datasets.
# We need to read the aduio files as arrays
def speech_file_to_array_fn(batch):
    speech_array, sampling_rate = torchaudio.load(batch["path"])
    batch["speech"] = resampler(speech_array).squeeze().numpy()
    return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["speech"][:2], sampling_rate=16_000, return_tensors="pt", padding=True)

with torch.no_grad():
    logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits

predicted_ids = torch.argmax(logits, dim=-1)

print("Prediction:", processor.batch_decode(predicted_ids))
print("Reference:", test_dataset["sentence"][:2])

Advanced Usage

The model can be evaluated as follows on the Turkish test data of Common Voice.

import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re

test_dataset = load_dataset("common_voice", "tr", split="test")
wer = load_metric("wer")

processor = Wav2Vec2Processor.from_pretrained("ceyda/wav2vec2-base-960-turkish")
model = Wav2Vec2ForCTC.from_pretrained("ceyda/wav2vec2-base-960-turkish")
model.to("cuda")

chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\‘\”\'\`…\’»«]'
resampler = torchaudio.transforms.Resample(48_000, 16_000)

# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
    batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
    speech_array, sampling_rate = torchaudio.load(batch["path"])
    batch["speech"] = resampler(speech_array).squeeze().numpy()
    return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)

# Preprocessing the datasets.
# We need to read the aduio files as arrays

#Attention mask is not used because the base-model was not trained with it. reference: https://github.com/huggingface/transformers/blob/403d530eec105c0e229fc2b754afdf77a4439def/src/transformers/models/wav2vec2/tokenization_wav2vec2.py#L305
def evaluate(batch):
    inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)

    with torch.no_grad():
        logits = model(inputs.input_values.to("cuda")).logits

    pred_ids = torch.argmax(logits, dim=-1)
    batch["pred_strings"] = processor.batch_decode(pred_ids,skip_special_tokens=True)
    return batch

result = test_dataset.map(evaluate, batched=True, batch_size=8)

print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))

Test Results:

• WER: 22.602390
• CER: 6.054137

🔧 Technical Details

The Common Voice train, validation datasets were used for training. The script used for training can be found here

📄 License

This model is licensed under the Apache 2.0 license.