Update README.md

README.md

@@ -34,12 +34,48 @@ The model can be used directly (without a language model) as follows:
 ```python
 import torch
 import torchaudio
+import pydub 
+from pydub.utils import mediainfo
+import array
+from pydub import AudioSegment
+from pydub.utils import get_array_type
+import numpy as np 
+
 from datasets import load_dataset
 from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
 test_dataset = load_dataset("common_voice", "tr", split="test[:2%]") 
 processor = Wav2Vec2Processor.from_pretrained("gorkemgoknar/wav2vec2-large-xlsr-53-turkish")
 model = Wav2Vec2ForCTC.from_pretrained("gorkemgoknar/wav2vec2-large-xlsr-53-turkish") 
-resampler = torchaudio.transforms.Resample(48_000, 16_000)
+
+
+
+def audio_resampler(batch,new_sample_rate = 16000):
+    
+    ##torchaudio and librosa troublesome to use for mp3 in windows
+    #speech_array, sampling_rate = torchaudio.load(batch["path"])
+    #speech_array, sampling_rate = librosa.load(batch["path"])
+
+    #AudioSegment does the job over ffmpeg(need install)    
+    sound = AudioSegment.from_file(file=batch["path"])
+    sound = sound.set_frame_rate(new_sample_rate)
+
+    left = sound.split_to_mono()[0]
+    bit_depth = left.sample_width * 8
+    array_type = get_array_type(bit_depth)
+
+    numeric_array = np.array(array.array(array_type, left._data) )
+
+    #windows hack as torchaudio cannot read mp3
+    speech_array = torch.FloatTensor(numeric_array)
+    
+    batch["speech"] = numeric_array
+    batch["sampling_rate"] = new_sample_rate
+    batch["target_text"] = batch["sentence"]
+
+    return batch
+  
+resampler = audio_resampler(16000)
+
 # Preprocessing the datasets.
 # We need to read the aduio files as arrays
 def speech_file_to_array_fn(batch):
@@ -69,7 +105,7 @@ model = Wav2Vec2ForCTC.from_pretrained("gorkemgoknar/wav2vec2-large-xlsr-53-turk
 model.to("cuda")
 
 #Note: Not ignoring "'"  on this one 
-chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\”\�]'
+chars_to_ignore_regex = '[\\,\\?\\.\\!\\-\\;\\:\\"\\“\\%\\‘\\”\\�]'
 
 resampler = torchaudio.transforms.Resample(48_000, 16_000)
 # Preprocessing the datasets.