#https://github.com/huggingface/diffusers/tree/main/examples/dreambooth
#export 
MODEL_NAME="stabilityai/stable-diffusion-2-1-base"
#export 
INSTANCE_DIR="./data_example"
#export 
OUTPUT_DIR="./output_example"

    
from diffusers import StableDiffusionPipeline
from lora_diffusion import monkeypatch_lora, tune_lora_scale
import torch
import os 
import gradio as gr
#os.system('python file.py')
import subprocess
# If your shell script has shebang,
# you can omit shell=True argument.
#subprocess.run("./run_lora_db.sh", shell=True)

#####
model_id = "stabilityai/stable-diffusion-2-1-base"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16).to("cuda")
prompt = "style of sks, baby lion"
torch.manual_seed(1)
#image = pipe(prompt, num_inference_steps=50, guidance_scale= 7).images[0] #no need
#image  # nice. diffusers are cool. #no need
finetuned_lora_weights = "./lora_weight.pt"

#####
#my fine tuned weights
def monkeypatching(alpha): #, prompt, pipe): finetuned_lora_weights
    monkeypatch_lora(pipe.unet, torch.load(finetuned_lora_weights)) #"./lora_weight.pt")) 
    tune_lora_scale(pipe.unet, alpha) #1.00)
    image = pipe(prompt, num_inference_steps=50, guidance_scale=7).images[0]
    image.save("./illust_lora.jpg") #"./contents/illust_lora.jpg")
    return image

def accelerate_train_lora(steps):
    print("***********inside accelerate_train_lora 11111***********")
    #subprocess.run(accelerate launch {"./train_lora_dreambooth.py"} \
    #subprocess.Popen(f'accelerate launch {"./train_lora_dreambooth.py"} \
    os.system( f'accelerate launch {"./train_lora_dreambooth.py"} \
      --pretrained_model_name_or_path={MODEL_NAME}  \
      --instance_data_dir={INSTANCE_DIR} \
      --output_dir={OUTPUT_DIR} \
      --instance_prompt="style of sks" \
      --resolution=512 \
      --train_batch_size=1 \
      --gradient_accumulation_steps=1 \
      --learning_rate=1e-4 \
      --lr_scheduler="constant" \
      --lr_warmup_steps=0 \
      --max_train_steps={int(steps)}') #,shell=True) #30000
    print("***********inside accelerate_train_lora 22222***********")
    return
    
with gr.Blocks() as demo:
    with gr.Row():
        in_images =  gr.File(label="Upload images to fine-tune for LORA", file_count="multiple")
        #in_prompt = gr.Textbox(label="Enter a ")
        in_steps = gr.Number(label="Enter number of steps")
        in_alpha = gr.Slider(0.1,1.0, step=0.01, label="Set Alpha level - higher value has more chances to overfit")
    with gr.Row():
        b1 = gr.Button(value="Train LORA model")
        b2 = gr.Button(value="Inference using LORA model")
    with gr.Row():
        out_image = gr.Image(label="Image generated by LORA model")
        out_file = gr.File(label="Lora trained model weights")
    b1.click(fn = accelerate_train_lora, inputs=in_steps, outputs=out_file)
    b2.click(fn = monkeypatching, inputs=in_alpha, outputs=out_image)

demo.launch(debug=True, show_error=True)