|
import PIL |
|
import requests |
|
import torch |
|
import gradio as gr |
|
import random |
|
from PIL import Image |
|
import os |
|
import time |
|
from diffusers import StableDiffusionInstructPix2PixPipeline, EulerAncestralDiscreteScheduler |
|
|
|
|
|
model_id = "timbrooks/instruct-pix2pix" |
|
pipe = StableDiffusionInstructPix2PixPipeline.from_pretrained(model_id, torch_dtype=torch.float16, revision="fp16", safety_checker=None) |
|
pipe.to("cuda") |
|
|
|
pipe.enable_xformers_memory_efficient_attention() |
|
pipe.unet.to(memory_format=torch.channels_last) |
|
|
|
help_text = """ """ |
|
|
|
def previous(image): |
|
return image |
|
|
|
def upload_image(file): |
|
return Image.open(file) |
|
|
|
def upload_button_config(): |
|
return gr.update(visible=False) |
|
|
|
def upload_textbox_config(text_in): |
|
return gr.update(visible=True) |
|
|
|
def dummy_fn(): |
|
return 'dummy' |
|
|
|
def chat(btn_upload, image_in, in_steps, in_guidance_scale, in_img_guidance_scale, image_hid, img_name, counter_out, image_oneup, prompt, history, progress=gr.Progress(track_tqdm=True)): |
|
progress(0, desc="Starting...") |
|
if prompt != '' and prompt.lower() == 'reverse' : |
|
history = history or [] |
|
temp_img_name = img_name[:-4]+str(int(time.time()))+'.png' |
|
image_oneup.save(temp_img_name) |
|
response = 'Reverted to the last image ' + '<img src="/file=' + temp_img_name + '">' |
|
history.append((prompt, response)) |
|
return history, history, image_oneup, temp_img_name, counter_out |
|
if prompt != '' and prompt.lower() == 'restart' : |
|
history = history or [] |
|
temp_img_name = img_name[:-4]+str(int(time.time()))+'.png' |
|
|
|
basewidth = 512 |
|
wpercent = (basewidth/float(image_in.size[0])) |
|
hsize = int((float(image_in.size[1])*float(wpercent))) |
|
image_in = image_in.resize((basewidth,hsize), Image.Resampling.LANCZOS) |
|
image_in.save(temp_img_name) |
|
response = 'Reverted to the last image ' + '<img src="/file=' + temp_img_name + '">' |
|
history.append((prompt, response)) |
|
return history, history, image_in, temp_img_name, counter_out |
|
|
|
add_text_list = ["There you go", "Enjoy your image!", "Nice work! Wonder what you gonna do next!", "Way to go!", "Does this work for you?", "Something like this?"] |
|
if counter_out == 0: |
|
t1 = time.time() |
|
print(f"Time at start = {t1}") |
|
seed = random.randint(0, 1000000) |
|
img_name = f"./edited_image_{seed}.png" |
|
|
|
image_in = Image.open(btn_upload) |
|
|
|
basewidth = 512 |
|
wpercent = (basewidth/float(image_in.size[0])) |
|
hsize = int((float(image_in.size[1])*float(wpercent))) |
|
image_in = image_in.resize((basewidth,hsize), Image.Resampling.LANCZOS) |
|
|
|
|
|
|
|
|
|
|
|
|
|
image_in.save(img_name) |
|
|
|
|
|
|
|
|
|
history = history or [] |
|
response = '<img src="/file=' + img_name + '">' |
|
history.append((prompt, response)) |
|
counter_out += 1 |
|
|
|
t2 = time.time() |
|
print(f"Time at end = {t2}") |
|
time_diff = t2-t1 |
|
print(f"Time taken = {time_diff}") |
|
return history, history, image_in, img_name, counter_out |
|
|
|
elif counter_out == 1: |
|
|
|
edited_image = pipe(prompt, image=image_in, num_inference_steps=int(in_steps), guidance_scale=float(in_guidance_scale), image_guidance_scale=float(in_img_guidance_scale)).images[0] |
|
if os.path.exists(img_name): |
|
os.remove(img_name) |
|
temp_img_name = img_name[:-4]+str(int(time.time()))[-4:] +'.png' |
|
with open(temp_img_name, "wb") as fp: |
|
|
|
edited_image.save(fp) |
|
|
|
saved_image_name1 = fp.name |
|
history = history or [] |
|
response = random.choice(add_text_list) + '<img src="/file=' + saved_image_name1 + '">' |
|
history.append((prompt, response)) |
|
counter_out += 1 |
|
return history, history, edited_image, temp_img_name, counter_out |
|
elif counter_out > 1: |
|
edited_image = pipe(prompt, image=image_hid, num_inference_steps=int(in_steps), guidance_scale=float(in_guidance_scale), image_guidance_scale=float(in_img_guidance_scale)).images[0] |
|
if os.path.exists(img_name): |
|
os.remove(img_name) |
|
temp_img_name = img_name[:-4]+str(int(time.time()))[-4:]+'.png' |
|
|
|
with open(temp_img_name, "wb") as fp: |
|
|
|
edited_image.save(fp) |
|
|
|
saved_image_name2 = fp.name |
|
|
|
history = history or [] |
|
response = random.choice(add_text_list) + '<img src="/file=' + saved_image_name2 + '">' |
|
history.append((prompt, response)) |
|
counter_out += 1 |
|
return history, history, edited_image, temp_img_name, counter_out |
|
|
|
|
|
|
|
with gr.Blocks(css="style.css") as demo: |
|
with gr.Column(elem_id="col-container") as main_col: |
|
gr.HTML("""<div style="text-align: center; max-width: 700px; margin: 0 auto;"> |
|
<div |
|
style=" |
|
display: inline-flex; |
|
align-items: center; |
|
gap: 0.8rem; |
|
font-size: 1.75rem; |
|
" |
|
> |
|
<h1 style="font-weight: 900; margin-bottom: 7px; margin-top: 5px;"> |
|
ChatPix2Pix: Image Editing by Instructions |
|
</h1> |
|
</div> |
|
<p style="margin-bottom: 10px; font-size: 94%"> |
|
For faster inference without waiting in the queue, you may duplicate the space and upgrade to GPU in settings <a href="https://huggingface.co/spaces/ysharma/InstructPix2Pix_Chatbot?duplicate=true"><img style="margin-top: 0em; margin-bottom: 0em" src="https://bit.ly/3gLdBN6" alt="Duplicate Space"></a> |
|
<a href="https://huggingface.co/timbrooks/instruct-pix2pix" target="_blank">Diffusers implementation of instruct-pix2pix</a> - InstructPix2Pix: Learning to Follow Image Editing Instructions! |
|
</p> |
|
</div>""") |
|
|
|
|
|
with gr.Accordion("Advance settings for Training and Inference", open=False): |
|
image_in = gr.Image(visible=False,type='pil', label="Original Image") |
|
gr.Markdown("Advance settings for - Number of Inference steps, Guidanace scale, and Image guidance scale.") |
|
in_steps = gr.Number(label="Enter the number of Inference steps", value = 20) |
|
in_guidance_scale = gr.Slider(1,10, step=0.5, label="Set Guidance scale", value=7.5) |
|
in_img_guidance_scale = gr.Slider(1,10, step=0.5, label="Set Image Guidance scale", value=1.5) |
|
image_hid = gr.Image(type='pil', visible=False) |
|
image_oneup = gr.Image(type='pil', visible=False) |
|
img_name_temp_out = gr.Textbox(visible=False) |
|
counter_out = gr.Number(visible=False, value=0, precision=0) |
|
dummy_num = gr.Number(visible=False) |
|
|
|
|
|
text_in = gr.Textbox(value='', Placeholder="Type your instructions here and press enter", elem_id = "input_prompt", visible=False, label='Great! Now you can edit your image with Instructions') |
|
btn_upload = gr.UploadButton("Upload image", file_types=["image"], file_count="single", elem_id="upload_button") |
|
|
|
chatbot = gr.Chatbot(elem_id = 'chatbot-component') |
|
state_in = gr.State() |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
element_dummy = gr.HTML(visbile = False, elem_id = 'dummy_elem') |
|
|
|
|
|
btn_upload.upload(chat, |
|
[btn_upload, image_in, in_steps, in_guidance_scale, in_img_guidance_scale, image_hid, img_name_temp_out,counter_out, image_oneup, text_in, state_in], |
|
[chatbot, state_in, image_in, img_name_temp_out, counter_out]) |
|
btn_upload.upload(fn = upload_textbox_config, inputs=text_in, outputs = text_in) |
|
|
|
text_in.submit(chat,[btn_upload, image_in, in_steps, in_guidance_scale, in_img_guidance_scale, image_hid, img_name_temp_out,counter_out, image_oneup, text_in, state_in], [chatbot, state_in, image_hid, img_name_temp_out, counter_out]) |
|
text_in.submit(previous, [image_hid], [image_oneup]) |
|
|
|
chatbot.change(fn = upload_button_config, outputs=btn_upload) |
|
text_in.submit(None, [], [], _js = "() => document.getElementById('#chatbot-component').scrollTop = document.getElementById('#chatbot-component').scrollHeight") |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
gr.Markdown(help_text, elem_id = 'help_text') |
|
|
|
|
|
demo.queue(concurrency_count=3) |
|
demo.launch(debug=True) |