jonatasgrosman/wav2vec2-large-xlsr-53-persian

Fine-tuned XLSR-53 large model for speech recognition in Persian

Fine-tuned facebook/wav2vec2-large-xlsr-53 on Persian using the train and validation splits of Common Voice 6.1. When using this model, make sure that your speech input is sampled at 16kHz.

This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud :)

The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint

Usage

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

Using the HuggingSound library:

from huggingsound import SpeechRecognitionModel

model = SpeechRecognitionModel("jonatasgrosman/wav2vec2-large-xlsr-53-persian")
audio_paths = ["/path/to/file.mp3", "/path/to/another_file.wav"]

transcriptions = model.transcribe(audio_paths)

Writing your own inference script:

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

LANG_ID = "fa"
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-persian"
SAMPLES = 5

test_dataset = load_dataset("common_voice", LANG_ID, split=f"test[:{SAMPLES}]")

processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)

# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
    speech_array, sampling_rate = librosa.load(batch["path"], sr=16_000)
    batch["speech"] = speech_array
    batch["sentence"] = batch["sentence"].upper()
    return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)
inputs = processor(test_dataset["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)
predicted_sentences = processor.batch_decode(predicted_ids)

for i, predicted_sentence in enumerate(predicted_sentences):
    print("-" * 100)
    print("Reference:", test_dataset[i]["sentence"])
    print("Prediction:", predicted_sentence)

Reference	Prediction
از مهمونداری کنار بکشم	از مهمانداری کنار بکشم
برو از مهرداد بپرس.	برو از ماقدعاد به پرس
خب ، تو چیكار می كنی؟	خوب تو چیکار می کنی
مسقط پایتخت عمان در عربی به معنای محل سقوط است	مسقط پایتخت عمان در عربی به بعنای محل سقوط است
آه، نه اصلاُ!	اهنه اصلا
توانست	توانست
قصیده فن شعر میگوید ای دوستان	قصیده فن شعر میگوید ایدوستون
دو استایل متفاوت دارین	دوبوست داریل و متفاوت بری
دو روز قبل از کریسمس ؟	اون مفتود پش پشش
ساعت های کاری چیست؟	این توری که موشیکل خب

Evaluation

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

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

LANG_ID = "fa"
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-persian"
DEVICE = "cuda"

CHARS_TO_IGNORE = [",", "?", "¿", ".", "!", "¡", ";", "；", ":", '""', "%", '"', "�", "ʿ", "·", "჻", "~", "՞",
                   "؟", "،", "।", "॥", "«", "»", "„", "“", "”", "「", "」", "‘", "’", "《", "》", "(", ")", "[", "]",
                   "{", "}", "=", "`", "_", "+", "<", ">", "…", "–", "°", "´", "ʾ", "‹", "›", "©", "®", "—", "→", "。",
                   "、", "﹂", "﹁", "‧", "～", "﹏", "，", "｛", "｝", "（", "）", "［", "］", "【", "】", "‥", "〽",
                   "『", "』", "〝", "〟", "⟨", "⟩", "〜", "：", "！", "？", "♪", "؛", "/", "\\", "º", "−", "^", "ʻ", "ˆ"]

test_dataset = load_dataset("common_voice", LANG_ID, split="test")

wer = load_metric("wer.py") # https://github.com/jonatasgrosman/wav2vec2-sprint/blob/main/wer.py
cer = load_metric("cer.py") # https://github.com/jonatasgrosman/wav2vec2-sprint/blob/main/cer.py

chars_to_ignore_regex = f"[{re.escape(''.join(CHARS_TO_IGNORE))}]"

processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)
model.to(DEVICE)

# Preprocessing the datasets.
# We need to read the audio files as arrays
def speech_file_to_array_fn(batch):
    with warnings.catch_warnings():
        warnings.simplefilter("ignore")
        speech_array, sampling_rate = librosa.load(batch["path"], sr=16_000)
    batch["speech"] = speech_array
    batch["sentence"] = re.sub(chars_to_ignore_regex, "", batch["sentence"]).upper()
    return batch

test_dataset = test_dataset.map(speech_file_to_array_fn)

# Preprocessing the datasets.
# We need to read the audio 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(DEVICE), attention_mask=inputs.attention_mask.to(DEVICE)).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)

predictions = [x.upper() for x in result["pred_strings"]]
references = [x.upper() for x in result["sentence"]]

print(f"WER: {wer.compute(predictions=predictions, references=references, chunk_size=1000) * 100}")
print(f"CER: {cer.compute(predictions=predictions, references=references, chunk_size=1000) * 100}")

Test Result:

In the table below I report the Word Error Rate (WER) and the Character Error Rate (CER) of the model. I ran the evaluation script described above on other models as well (on 2021-04-22). Note that the table below may show different results from those already reported, this may have been caused due to some specificity of the other evaluation scripts used.

Model	WER	CER
jonatasgrosman/wav2vec2-large-xlsr-53-persian	30.12%	7.37%
m3hrdadfi/wav2vec2-large-xlsr-persian-v2	33.85%	8.79%
m3hrdadfi/wav2vec2-large-xlsr-persian	34.37%	8.98%

Citation

If you want to cite this model you can use this:

@misc{grosman2021xlsr53-large-persian,
  title={Fine-tuned {XLSR}-53 large model for speech recognition in {P}ersian},
  author={Grosman, Jonatas},
  howpublished={\url{https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-persian}},
  year={2021}
}