chris/MiniGPT-4
1
0
Fork 0
mirror of https://github.com/Vision-CAIR/MiniGPT-4.git synced 2025-04-07 19:40:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

modify the code so to enable demo to accept multiple image input

Browse Source
This commit is contained in:
maple.liang 2023-05-17 17:15:15 +08:00
parent 7fc77104c5
commit 5b0a99e95d
1 changed files with 41 additions and 19 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

60
demo.py
View File

@ -68,21 +68,32 @@ print('Initialization Finished')
# Gradio Setting # Gradio Setting
# ======================================== # ========================================
def gradio_reset(chat_state, img_list): def gradio_reset():
if chat_state is not None: # reset chatbot, image, text_input, upload_button, chat_state, img_list, img_emb_list, gallery
chat_state.messages = [] return None, \
if img_list is not None: gr.update(value=None, interactive=True), \
img_list = [] gr.update(placeholder='Please upload your image first', interactive=False), \
return None, gr.update(value=None, interactive=True), gr.update(placeholder='Please upload your image first', interactive=False),gr.update(value="Upload & Start Chat", interactive=True), chat_state, img_list gr.update(value="Upload & Start Chat", interactive=True), \
CONV_VISION.copy(), \
[], \
[], \
[]
def upload_img(gr_img, chat_state): def upload_img(gr_img, chat_state, img_list, img_emb_list):
if gr_img is None: if gr_img is None:
return None, None, gr.update(interactive=True), chat_state, None return None, None, gr.update(interactive=True), chat_state, img_list, img_emb_list
chat_state = CONV_VISION.copy() img_list.append(gr_img)
img_list = [] # upload an image to the chat
llm_message = chat.upload_img(gr_img, chat_state, img_list) chat.upload_img(gr_img, chat_state, img_emb_list)
return gr.update(interactive=False), gr.update(interactive=True, placeholder='Type and press Enter'), gr.update(value="Start Chatting", interactive=False), chat_state, img_list # update image, text_input, upload_button, chat_state, gallery, img_emb_list
return gr.update(value=None, interactive=False), \
gr.update(interactive=True, placeholder='Type and press Enter'), \
gr.update(value="Send more images after sending a message", interactive=False), \
chat_state, \
img_list, \
img_emb_list
def gradio_ask(user_message, chatbot, chat_state): def gradio_ask(user_message, chatbot, chat_state):
if len(user_message) == 0: if len(user_message) == 0:
@ -100,7 +111,11 @@ def gradio_answer(chatbot, chat_state, img_list, num_beams, temperature):
max_new_tokens=300, max_new_tokens=300,
max_length=2000)[0] max_length=2000)[0]
chatbot[-1][1] = llm_message chatbot[-1][1] = llm_message
return chatbot, chat_state # update chatbot, chat_state, image, upload_button
return chatbot, \
chat_state, \
gr.update(interactive=True), \
gr.update(value="Send more image", interactive=True)
title = """<h1 align="center">Demo of MiniGPT-4</h1>""" title = """<h1 align="center">Demo of MiniGPT-4</h1>"""
@ -140,19 +155,26 @@ with gr.Blocks() as demo:
) )
with gr.Column(): with gr.Column():
chat_state = gr.State() chat_state = gr.State(CONV_VISION.copy())
img_list = gr.State() img_list = gr.State([])
img_emb_list = gr.State([])
gallery = gr.Gallery(label="Uploaded Images", show_label=True) \
.style(rows=[1], object_fit="scale-down", height="500px", preview=True)
chatbot = gr.Chatbot(label='MiniGPT-4') chatbot = gr.Chatbot(label='MiniGPT-4')
text_input = gr.Textbox(label='User', placeholder='Please upload your image first', interactive=False) text_input = gr.Textbox(label='User', placeholder='Please upload your image first', interactive=False)
upload_button.click(upload_img, [image, chat_state], upload_button.click(upload_img, [image, chat_state, img_list, img_emb_list],
[image, text_input, upload_button, chat_state, img_list]) [image, text_input, upload_button, chat_state, gallery, img_emb_list])
text_input \ text_input \
.submit(gradio_ask, [text_input, chatbot, chat_state], [text_input, chatbot, chat_state]) \ .submit(gradio_ask, [text_input, chatbot, chat_state], [text_input, chatbot, chat_state]) \
.then(gradio_answer, [chatbot, chat_state, img_list, num_beams, temperature], [chatbot, chat_state]) .then(gradio_answer,
[chatbot, chat_state, img_emb_list, num_beams, temperature],
[chatbot, chat_state, image, upload_button])
clear.click(gradio_reset, [chat_state, img_list], [chatbot, image, text_input, upload_button, chat_state, img_list], clear.click(gradio_reset,
None,
[chatbot, image, text_input, upload_button, chat_state, img_list, img_emb_list, gallery],
queue=False) queue=False)
demo.launch(share=True, enable_queue=True) demo.launch(share=True, enable_queue=True)