changed the process image function

app.py

@@ -76,12 +76,17 @@ def convert_image_to_required_format(image):
     """
     convert image from numpy to base64
     """
-    if type(image) == type(np.array([])):
-        return base64.b64encode(image).decode('utf-8')
+    img = base64.b64encode(image).decode('utf-8')
+    image_name = np.random.randint(0, 10)
+    with open(f'{image_name}.png', 'wb') as f:
+        f.write(base64.b64decode(img))
+    return image_name
+
+    
 
 
 def process_image_with_openai(image):
-    image_data = convert_image_to_required_format(image)
+    image_name = convert_image_to_required_format(image)
     openai_api_key = os.getenv('OPENAI_API_KEY') 
     oai_org = os.getenv('OAI_ORG')
     if openai_api_key is None:
@@ -92,7 +97,7 @@ def process_image_with_openai(image):
         "messages": [
             {
                 "role": "user",
-                "content": image_data
+                "content": image_name
             }
         ],
         "max_tokens": 300

test.py

@@ -0,0 +1,40 @@
+# import gradio library
+import gradio as gr
+import numpy as np
+import base64
+import time
+import gradio_client as grc
+# client = grc.InterfaceDescriptionClient("https://gradio.app")
+# define a function that takes input and returns output
+client = grc.Client("facebook/seamless_m4t")
+
+print(client.view_api(all_endpoints=True))
+job = client.submit(
+        "S2TT",
+        "file",
+        None,
+        "sample_input.mp3",
+        "",
+        "French",
+        "English",
+        api_name="/run",
+    )
+while job.done() == False :
+    time.sleep(1)
+    
+
+gr.Markdown(job.result())
+#     return out
+
+# # print(client.view_api(all_endpoints=True))
+# def convert_image_to_required_format(audio):
+#     sr, data = audio
+#     return str(type(sr))+str(type(data))
+
+# # create a gradio interface
+# iface = gr.Interface(
+#     fn=sound_to_text, 
+#     inputs=[gr.Audio()],
+#     outputs=[gr.Markdown(label="Output Text")],
+#     )
+# iface.launch()