Wav2Vec2-Large-XLSR-53-Swedish

Fine-tuned facebook/wav2vec2-large-xlsr-53 on Swedish using the Common Voice and parts for the NST Swedish ASR Database. When using this model, make sure that your speech input is sampled at 16kHz.

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", "sv-SE", split="test[:2%]") #TODO: replace {lang_id} in your language code here. Make sure the code is one of the *ISO codes* of [this](https://huggingface.co/languages) site.

processor = Wav2Vec2Processor.from_pretrained("vasilis/wav2vec2-large-xlsr-53-swedish") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`
model = Wav2Vec2ForCTC.from_pretrained("vasilis/wav2vec2-large-xlsr-53-swedish") #TODO: replace {model_id} with your model id. The model id consists of {your_username}/{your_modelname}, *e.g.* `elgeish/wav2vec2-large-xlsr-53-arabic`

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])

Evaluation

The model can be evaluated as follows on the Swedish 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", "sv-SE", split="test")
wer = load_metric("wer")

processor = Wav2Vec2Processor.from_pretrained("vasilis/wav2vec2-large-xlsr-53-swedish")
model = Wav2Vec2ForCTC.from_pretrained("vasilis/wav2vec2-large-xlsr-53-swedish")
model.to("cuda")

chars_to_ignore_regex = "[\,\?\.\!\-\;\:\"\“\%\‘\”\�\']"  # TODO: adapt this list to include all special characters you removed from the data

resampler = {
    48_000: torchaudio.transforms.Resample(48_000, 16_000),
    44100: torchaudio.transforms.Resample(44100, 16_000),
    32000: torchaudio.transforms.Resample(32000, 16_000)
}

# Preprocessing the datasets.
# We need to read the aduio 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[sampling_rate](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
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"), attention_mask=inputs.attention_mask.to("cuda")).logits
    pred_ids = torch.argmax(logits, dim=-1)
    batch["pred_strings"] = processor.batch_decode(pred_ids)
    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"])))
print("CER: {:2f}".format(100 * wer.compute(predictions=[" ".join(list(entry)) for entry in result["pred_strings"]], references=[" ".join(list(entry)) for entry in result["sentence"]])))

Test Result: 14.695793 %

Training

As first step used Common Voice train dataset and parts from NST as can be found here. Part of NST where removed using this mask

mask = [(5 < len(x.split()) < 20) and np.average([len(entry) for entry in x.split()]) > 5 for x in dataset['transcript'].tolist()]

After training like this for 20000 steps the model was finetuned on all of nst data using the mask

mask = [(1 < len(x.split()) < 25) and np.average([len(entry) for entry in x.split()]) > 3 for x in dataset['transcript'].tolist()]

and all of common voice for 100000 more steps approximately 16 epochs.