Update app.py

app.py

@@ -57,19 +57,18 @@ SIDEBAR_INFO = f"""
 
 warnings.filterwarnings("ignore")
 
-#def convert_to_wav(filepath):
-    #_,file_ending = os.path.splitext(f'{filepath}')
-    #audio_file = filepath.replace(file_ending, ".wav")
-    #print("starting conversion to wav")
-    #os.system(f'ffmpeg -i "{filepath}" -ar 16000 -ac 1 -c:a pcm_s16le "{audio_file}"')
-    #return audio_file
+def convert_to_wav(filepath):
+    _,file_ending = os.path.splitext(f'{filepath}')
+    audio_file = filepath.replace(file_ending, ".wav")
+    os.system(f'ffmpeg -i "{filepath}" -ar 16000 -ac 1 -c:a pcm_s16le "{audio_file}"')
+    return audio_file
 
 #:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
-def convert_to_wav(audio_file):
-    audio = AudioSegment.from_file(audio_file, format="m4a")
-    wav_file = "temp.wav"
-    audio.export(wav_file, format="wav")
-    return wav_file
+#def convert_to_wav(audio_file):
+    #audio = AudioSegment.from_file(audio_file, format="m4a")
+    #wav_file = "temp.wav"
+    #audio.export(wav_file, format="wav")
+    #return wav_file
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
@@ -81,8 +80,8 @@ pipe = pipeline(
 )
 
 def transcribe_audio(audio_file, batch_size=10):
-    if audio_file.endswith(".m4a"):
-        audio_file = convert_to_wav(audio_file)
+    #if audio_file.endswith(".m4a"):
+        #audio_file = convert_to_wav(audio_file)
 
     start_time = time.time()
 
@@ -96,10 +95,6 @@ def transcribe_audio(audio_file, batch_size=10):
     output_time = end_time - start_time
     word_count = len(text.split())
 
-    result = f"Transcription: {text.strip()}\n\nTime taken: {output_time:.2f} seconds\nNumber of words: {word_count}"
-
-
-    return text.strip(), result
 
     memory = psutil.virtual_memory()
     gpu_utilization, gpu_memory = GPUInfo.gpu_usage()
@@ -107,8 +102,12 @@ def transcribe_audio(audio_file, batch_size=10):
     gpu_memory = gpu_memory[0] if len(gpu_memory) > 0 else 0
     system_info = f"""
     *Memory: {memory.total / (1024 * 1024 * 1024):.2f}GB, used: {memory.percent}%, available: {memory.available / (1024 * 1024 * 1024):.2f}GB.*
-    *Processing time: {time_diff:.5} seconds.*
-    *GPU Utilization: {gpu_utilization}%, GPU Memory: {gpu_memory}"""
+    *Processing time: {output_time:.2f} seconds.*
+    *Number of words: {word_count}*
+    *GPU Utilization: {gpu_utilization}%, GPU Memory: {gpu_memory}*"""
+
+
+    return text.strip(), system_info
 #:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
 
 # Clean/preprocess text
@@ -119,6 +118,7 @@ def clean_text(text):
     return text
 
 nlp = spacy.blank("nb")  # 'nb' ==> codename = Norwegian Bokmål
+nlp.add_pipe('sentencizer')
 spacy_stop_words = spacy.lang.nb.stop_words.STOP_WORDS
 
 def preprocess_text(text):
@@ -224,7 +224,6 @@ iface = gr.Blocks()
 with iface:
     
     gr.Image(LOGO) # LOGO variable as string to gr.Image constructor
-    gr.HTML(SIDEBAR_INFO)
     gr.Markdown(HEADER_INFO)
 
     with gr.Tabs():