metadata

license: mit
base_model: facebook/w2v-bert-2.0
tags:
  - generated_from_trainer
datasets:
  - common_voice_7_0
metrics:
  - wer
model-index:
  - name: w2v-bert-2.0-luganda-CV-train-validation-7.0
    results:
      - task:
          name: Automatic Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: common_voice_7_0
          type: common_voice_7_0
          config: lg
          split: test
          args: lg
        metrics:
          - name: Wer
            type: wer
            value: 0.1933150003273751

w2v-bert-2.0-luganda-CV-train-validation-7.0

This model is a fine-tuned version of facebook/w2v-bert-2.0 on the Luganda mozilla common voices 7.0 dataset. We use the train and validation set for training and the test set for evaluation. When using this dataset, make sure that the audio has a sampling rate of 16kHz.It achieves the following results on the test set:

Loss: 0.2282
Wer: 0.1933

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

The model was trained on version 7 of the Luganda dataset of Mozilla common voices dataset. We used the train and validation dataset for training and the test dataset for validation.

Training procedure

We trained the model on a 32 GB V100 GPU for 10 epochs using a learning rate of 5e-05. We used the AdamW optimizer.

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-05
train_batch_size: 32
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 500
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
1.1859	1.89	300	0.2854	0.2866
0.1137	3.77	600	0.2503	0.2469
0.0712	5.66	900	0.2043	0.2092
0.0446	7.55	1200	0.2156	0.2005
0.0269	9.43	1500	0.2282	0.1933

Framework versions

Transformers 4.38.1
Pytorch 2.2.1+cu121
Datasets 2.17.0
Tokenizers 0.15.2

Usage

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

test_dataset = load_dataset("common_voice", "lg", split="test[:10]") 

processor = Wav2Vec2Processor.from_pretrained("dmusingu/w2v-bert-2.0-luganda-CV-train-validation-7.0") 
model = Wav2Vec2ForCTC.from_pretrained("dmusingu/w2v-bert-2.0-luganda-CV-train-validation-7.0")

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):
    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[:2]["speech"], 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])