change way of loading image

delete groq prompts and add sources image

use pil for default image

change image path

refine prompt for specifying only front and fix image

add example image

lower weight flat

refine prompt again

refined prompt

guidance scale = 3

change loras weights and prompt refinement

finish gogle gemini implementation

first commit google gemini

put back the example in the prompt

refine prompt

refine prompt and increase guidance scale

fix image width and height

change from diffusion pipeline to flux control

fix type pil

convert image to pil

add hf login

fix typo

fix missing import

fix missing comma

use flux canny lora

remove js injection

try other js injection

fix light forcing

try to force light mode

change background color

try change button style

fix generated_prompt not defined

add button class

remove titles

revert

fix

remove unused functions

refine prompt

fix layout

update app layout

edit prompt

concatenate prompt and image generation

change prompt for llama vision

remove useless stuff

add api key to groq

switch to groq for image description

return markdown text

fix

fix function arguments bug

add prompt generation