Upload 3 files

.gitignore

@@ -0,0 +1,2 @@
+.env
+myenv

app.py

@@ -0,0 +1,100 @@
+import gradio as gr
+from dotenv import load_dotenv
+import gradio as gr
+import openai
+import os
+import soundfile
+from espnet2.bin.tts_inference import Text2Speech
+
+
+
+
+def tts(text):
+    
+    text2speech = Text2Speech.from_pretrained("kan-bayashi/ljspeech_vits")
+    speech = text2speech(text)["wav"]
+    speech = speech.numpy()
+
+    # Save the waverform
+    soundfile.write('example_TTS.wav', speech, 22050, 'PCM_24')
+    return os.path.join(os.path.dirname(__file__), "example_TTS.wav")
+
+
+def transcribe(audio):
+    # print(audio)
+    # global conv
+    file = open(audio, "rb")
+    transcription = openai.Audio.transcribe("whisper-1", file=file)
+    req = transcription['text']
+    # conv.append(req)
+    return req
+    
+
+def conversation(audio):
+    # global conv
+    req = transcribe(audio)
+    completion = openai.ChatCompletion.create(
+        model="gpt-3.5-turbo",
+        messages=[
+            {"role": "user", "content": req}]
+            # {"role": "user", "content" : line} for line in conv]
+    )
+    req2 = completion['choices'][0]['message']['content']
+    fin_text = 'You: ' + req + '\n' + 'AI: ' + req2.strip()
+    tts_data = tts(req2)
+    return fin_text, tts_data
+
+def generate_image(audio):
+    text = transcribe(audio)
+    response = openai.Image.create(
+    prompt=text,
+    n=1,
+    size="1024x1024"
+    )
+    return response['data'][0]['url']
+
+
+with gr.Blocks() as demo:
+    # global conv
+    load_dotenv()
+    openai.api_key=os.getenv('api_key')
+    conv = []
+    with gr.Tab("Start a conversation"):
+        with gr.Row():
+            audio_input_conv=gr.Audio(source="microphone", type="filepath")
+            text_output_conv=gr.Textbox(lines=10)
+        audio_output_conv = gr.Audio()
+        with gr.Row():
+            clear_button_conv = gr.Button("Clear")
+            submit_button_conv = gr.Button("Submit")
+
+    with gr.Tab("Generate image"):
+        with gr.Row():
+            audio_input_img=gr.Audio(source="microphone", type="filepath")
+            image_output_img = gr.Image()
+        with gr.Row():
+            clear_button_img = gr.Button("Clear")
+            submit_button_img = gr.Button("Submit")
+        # image_button = gr.Button("Flip")
+
+    # with gr.Accordion("Open for More!"):
+    #     gr.Markdown("Look at me...")
+
+    
+    with gr.Accordion("How to use"):
+        gr.Markdown("Record and submit your voice to talk to AI or to generate an image!!")
+
+    submit_button_conv.click(conversation, inputs=audio_input_conv, outputs=[text_output_conv, audio_output_conv])
+    clear_button_conv.click(lambda: None, None, audio_input_conv, queue=False)
+    submit_button_img.click(generate_image, inputs=audio_input_img, outputs=image_output_img)
+    clear_button_img.click(lambda: None, None, audio_input_img, queue=False)
+    clear_button_img.click(lambda: None, None, image_output_img, queue=False)
+
+
+# def same_auth(username, password):
+#     return username == password
+
+
+# demo.launch(auth=same_auth)
+
+demo.launch()

requirements.txt

@@ -0,0 +1,4 @@
+gradio==3.20.1
+openai==0.27.0
+python-dotenv==1.0.0
+soundfile==0.12.1