1

app.py

@@ -46,13 +46,17 @@ def resize(img_list):
 #clip.write_videofile('/content/gdrive/My Drive/AI/my_vid_20apr.mp4')
 
 
-def merge_audio_video(entities_num, resize_img_list, speech):
+def merge_audio_video(entities_num, resize_img_list, text_input):
     print("** inside merge aud vid **")
     print(type(resize_img_list))
     print(type(resize_img_list[0]))
         
-    #convert flac to mp3 audio format 
+    
+    #Convert text to speech using facebook's latest model from HF hub   
+    speech = text2speech(text_input)
+    print(type(speech))
     wav_audio = AudioSegment.from_file(speech, "flac")  #("/content/gdrive/My Drive/AI/audio1.flac", "flac")
+    #convert flac to mp3 audio format 
     print('flac audio read', type(wav_audio))
     wav_audio.export("audio.mp3", format="mp3")  #("/content/gdrive/My Drive/AI/audio1.mp3", format="mp3")
     print('flac audio converted to mp3 audio' )
@@ -114,8 +118,8 @@ def engine(text_input):
     print('back from resize')
      
     #Convert text to speech using facebook's latest model from HF hub   
-    speech = text2speech(text_input)
-    print('back in engine')
+    #speech = text2speech(text_input)
+    #print('back in engine')
     
     #getting audio clip's duration
     #audio_length = int(WAVE(speech).info.length)
@@ -129,7 +133,7 @@ def engine(text_input):
     #clip = images_to_video(fps, resize_img_list)
      
     #Merge video and audio created above
-    mergedclip = merge_audio_video(entities_num, resize_img_list, speech)
+    mergedclip = merge_audio_video(entities_num, resize_img_list, text_input)
     #{'prompt':text_input,'steps':'50','width':'256','height':'256','images':'1','scale':10}).launch()
     #img_intfc = gr.Interface.load("spaces/multimodalart/latentdiffusion", inputs=[gr.inputs.Textbox(lines=1, label="Input Text"), gr.inputs.Textbox(lines=1, label="Input Text"), gr.inputs.Textbox(lines=1, label="Input Text"), gr.inputs.Textbox(lines=1, label="Input Text"), gr.inputs.Textbox(lines=1, label="Input Text"), gr.inputs.Textbox(lines=1, label="Input Text")],
     #outputs=[gr.outputs.Image(type="pil", label="output image"),gr.outputs.Carousel(label="Individual images",components=["image"]),gr.outputs.Textbox(label="Error")], )