Create app.py

app.py

@@ -0,0 +1,77 @@
+import string
+import gradio as gr
+import requests
+import io
+
+url="http://202.85.216.23:5000/vqa_img"
+def inference_chat(input_image,input_text):
+    with io.BytesIO() as buf:
+        input_image.save(buf, 'jpeg')
+        image_bytes = buf.getvalue()
+        files={"img":("input_imge.jpg",image_bytes,'image/png',{})}
+        res=requests.request("post",url=url,data={"input_text":input_text},files=files)
+        return res.json()["answer"]
+    
+with gr.Blocks(
+    css="""
+    .message.svelte-w6rprc.svelte-w6rprc.svelte-w6rprc {font-size: 20px; margin-top: 20px}
+    #component-21 > div.wrap.svelte-w6rprc {height: 600px;}
+    """
+) as iface:
+    state = gr.State([])
+    #caption_output = None
+    #gr.Markdown(title)
+    #gr.Markdown(description)
+    #gr.Markdown(article)
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            image_input = gr.Image(type="pil")
+            with gr.Row():
+                with gr.Column(scale=1):
+                    chat_input = gr.Textbox(lines=1, label="VQA Input(问题输入)")
+                    with gr.Row():
+                        clear_button = gr.Button(value="Clear", interactive=True)
+                        submit_button = gr.Button(
+                            value="Submit", interactive=True, variant="primary"
+                        )
+        with gr.Column():
+            caption_output = gr.Textbox(lines=0, label="VQA Output(模型答案输出)")
+            
+            
+        image_input.change(
+            lambda: ("", "", []),
+            [],
+            [ caption_output, state],
+            queue=False,
+        )
+        chat_input.submit(
+                    inference_chat,
+                    [
+                        image_input,
+                        chat_input,
+                    ],
+                    [ caption_output],
+                )
+        clear_button.click(
+                        lambda: ("", [], []),
+                        [],
+                        [chat_input,  state],
+                        queue=False,
+                    )
+        submit_button.click(
+                        inference_chat,
+                        [
+                            image_input,
+                            chat_input,
+                        ],
+                        [caption_output],
+                    )
+
+   # examples = gr.Examples(
+   #     examples=examples,
+   #     inputs=[image_input, chat_input],
+  #  )
+
+iface.queue(concurrency_count=1, api_open=False, max_size=10)
+iface.launch(enable_queue=True)