README.md

metadata

language: hu
datasets:
  - common_voice
metrics:
  - wer
  - cer
tags:
  - audio
  - automatic-speech-recognition
  - speech
  - xlsr-fine-tuning-week
license: apache-2.0
model-index:
  - name: XLSR Wav2Vec2 Hungarian by Jonatas Grosman
    results:
      - task:
          name: Speech Recognition
          type: automatic-speech-recognition
        dataset:
          name: Common Voice hu
          type: common_voice
          args: hu
        metrics:
          - name: Test WER
            type: wer
            value: 31.4
          - name: Test CER
            type: cer
            value: 6.2

Wav2Vec2-Large-XLSR-53-Hungarian

Fine-tuned facebook/wav2vec2-large-xlsr-53 on Hungarian using the Common Voice and CSS10. When using this model, make sure that your speech input is sampled at 16kHz.

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:

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

LANG_ID = "hu"
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-hungarian"
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
BÜSZKÉK VAGYUNK A MAGYAR EMBEREK NAGYSZERŰ SZELLEMI ALKOTÁSAIRA.	BÜSZKÉK VAGYUNK A MAGYAR EMBEREK NAGYSZERŰ SZELLEMI ALKOTÁSAIRE
A NEMZETSÉG TAGJAI KÖZÜL EZT TERMESZTIK A LEGSZÉLESEBB KÖRBEN ÍZLETES TERMÉSÉÉRT.	A NEMZETSÉG TAGJAI KÖZÜL ESZSZERMESZTIK A LEGSZELESEBB KÖRBEN IZLETES TERMÉSSÉÉRT
A VÁROSBA VÁGYÓDOTT A LEGJOBBAN, ÉPPEN MERT ODA NEM JUTHATOTT EL SOHA.	A VÁROSBA VÁGYÓDOTT A LEGJOBBAN ÉPPEN MERT ODA NEM JUTHATOTT EL SOHA
SÍRJA MÁRA MEGSEMMISÜLT.	SIMGI A MANDO MEG SEMMICSEN
MINDEN ZENESZÁMOT DRÁGAKŐNEK NEVEZETT.	MINDEN ZENA SZÁMODRAGAKŐNEK NEVEZETT
ÍGY MÚLT EL A DÉLELŐTT.	ÍGY MÚLT EL A DÍN ELŐTT
REMEK POFA!	A REMEG PUFO
SZEMET SZEMÉRT, FOGAT FOGÉRT.	SZEMET SZEMÉRT FOGADD FOGÉRT
BIZTOSAN LAKIK ITT NÉHÁNY ATYÁMFIA.	BIZTOSAN LAKIKÉT NÉHANY ATYAMFIA
A SOROK KÖZÖTT OLVAS.	A SOROG KÖZÖTT OLVAS

Evaluation

The model can be evaluated as follows on the Hungarian 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 = "hu"
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-hungarian"
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-hungarian	31.40%	6.20%
anton-l/wav2vec2-large-xlsr-53-hungarian	42.39%	9.39%
gchhablani/wav2vec2-large-xlsr-hu	46.42%	10.04%
birgermoell/wav2vec2-large-xlsr-hungarian	46.93%	10.31%