Upload app.py

app.py

@@ -26,14 +26,14 @@ def measure_performance(func):
     return wrapper
 
 @measure_performance
-def generate_story(theme):
+def generate_story_text(theme):
     prompt = f"Tell me a storie about {theme}"
     story = generate_story_pipe(prompt, max_length=400, num_return_sequences=1, temperature=0.7, do_sample=True, repetition_penalty=3.0, eos_token_id=generate_story_pipe.model.config.eos_token_id, pad_token_id=generate_story_pipe.model.config.pad_token_id)
     generated_story = story[0]["generated_text"]
     return generated_story
 
 @measure_performance
-def gen_tts(text):
+def generate_tts(text):
     audio_arr = generate_audio(text, history_prompt="es_speaker_0")
     audio_arr = (audio_arr * 32767).astype(np.int16)
     return (SAMPLE_RATE, audio_arr)
@@ -53,9 +53,9 @@ def generate_image(prompt):
 
     return image
 
-def generate_story_and_convert_to_audio(theme):
-    story_text = generate_story(theme)
-    return gen_tts(story_text)
+def generate_story(theme):
+    story_text = generate_story_text(theme)
+    return (generate_tts(story_text), generate_image(story_text))
 
 def subtitle():
     gr.Markdown("""
@@ -146,12 +146,11 @@ with gr.Blocks(css=css) as block:
             image_out = gr.Image(label="Generated image", show_label=False)
     run_button = gr.Button("Cuéntame la historia")
     run_button.click(fn=lambda: gr.update(visible=False), inputs=None, outputs=None, queue=False).then(
-        fn=generate_story_and_convert_to_audio, inputs=[story_theme], outputs=[audio_out], queue=True).then(
-        fn=lambda theme: generate_image(theme), inputs=[story_theme], outputs=[image_out], queue=True).then(
+        fn=generate_story, inputs=[story_theme], outputs=[audio_out, image_out], queue=True).then(
         fn=lambda: gr.update(visible=True), inputs=None, outputs=None, queue=False)
 
 
-    gr.Examples(examples=examples, fn=[generate_story_and_convert_to_audio, generate_image], inputs=[story_theme],
+    gr.Examples(examples=examples, fn=generate_story, inputs=[story_theme],
                 outputs=[audio_out, image_out], cache_examples=True)
     end_text()