updated logging

gradio_asr_en_libri100_word_vs_bpe.py

@@ -9,7 +9,6 @@ from difflib import Differ
 from glob import glob
 from loguru import logger
 import librosa
-import scipy.signal as sps
 import gradio as gr
 
 from espnet2.bin.asr_inference import Speech2Text
@@ -45,20 +44,10 @@ def predict(wav_file):
     logger.info('wav file loaded')
     # Load audio    
     speech, rate = librosa.load(wav_file, sr=16000)
-    # rate, speech = wav_file
-    # if len(speech.shape) == 2:
-    #     speech = speech.T[0]
-    # wav = wavio.read(wav_file)
-    # speech = wav.data
-    # rate = wav.rate
-    logger.info(f'speech.shape: {speech.shape}')
-    logger.info(f'speech.dtype: {speech.dtype}')
-    logger.info(f'speech rate: {rate}')
-    # speech = speech.astype('float32')
-    
-    # number_of_samples = round(len(speech_raw) * float(16000) / rate)
-    # speech = sps.resample(speech_raw, number_of_samples)
-    # speech = speech_raw.astype('float32')
+
+    logger.info(f'--- speech.shape: {speech.shape}')
+    logger.info(f'--- speech.dtype: {speech.dtype}')
+    logger.info(f'--- speech rate: {rate}')
 
     # Run inference
     W = model_word(speech)[0]