2022-12-20 14:15:17 +00:00
|
|
|
import gradio as gr
|
|
|
|
import json
|
|
|
|
import math
|
|
|
|
import os
|
|
|
|
import subprocess
|
|
|
|
import pathlib
|
|
|
|
import shutil
|
2022-12-21 14:05:06 +00:00
|
|
|
import argparse
|
2022-12-20 14:15:17 +00:00
|
|
|
from library.common_gui import (
|
|
|
|
get_folder_path,
|
|
|
|
get_file_path,
|
2023-01-01 19:14:58 +00:00
|
|
|
get_any_file_path,
|
2022-12-20 14:15:17 +00:00
|
|
|
get_saveasfile_path,
|
|
|
|
)
|
2022-12-22 16:51:34 +00:00
|
|
|
from library.utilities import utilities_tab
|
2022-12-20 14:15:17 +00:00
|
|
|
|
|
|
|
folder_symbol = '\U0001f4c2' # 📂
|
|
|
|
refresh_symbol = '\U0001f504' # 🔄
|
|
|
|
save_style_symbol = '\U0001f4be' # 💾
|
|
|
|
document_symbol = '\U0001F4C4' # 📄
|
|
|
|
|
|
|
|
|
|
|
|
def save_configuration(
|
|
|
|
save_as,
|
|
|
|
file_path,
|
|
|
|
pretrained_model_name_or_path,
|
|
|
|
v2,
|
|
|
|
v_parameterization,
|
|
|
|
train_dir,
|
|
|
|
image_folder,
|
|
|
|
output_dir,
|
|
|
|
logging_dir,
|
|
|
|
max_resolution,
|
2023-01-02 18:07:17 +00:00
|
|
|
min_bucket_reso,
|
|
|
|
max_bucket_reso,
|
|
|
|
batch_size,
|
|
|
|
flip_aug,
|
|
|
|
caption_metadata_filename,
|
|
|
|
latent_metadata_filename,
|
|
|
|
full_path,
|
2022-12-20 14:15:17 +00:00
|
|
|
learning_rate,
|
|
|
|
lr_scheduler,
|
|
|
|
lr_warmup,
|
|
|
|
dataset_repeats,
|
|
|
|
train_batch_size,
|
|
|
|
epoch,
|
|
|
|
save_every_n_epochs,
|
|
|
|
mixed_precision,
|
|
|
|
save_precision,
|
|
|
|
seed,
|
|
|
|
num_cpu_threads_per_process,
|
|
|
|
train_text_encoder,
|
|
|
|
create_caption,
|
2023-01-02 18:07:17 +00:00
|
|
|
create_buckets,
|
2022-12-20 14:15:17 +00:00
|
|
|
save_model_as,
|
|
|
|
caption_extension,
|
2023-01-02 18:07:17 +00:00
|
|
|
use_8bit_adam,
|
|
|
|
xformers,
|
2023-01-06 00:16:13 +00:00
|
|
|
clip_skip,
|
2023-01-09 00:31:44 +00:00
|
|
|
save_state,
|
|
|
|
resume,
|
2023-01-09 01:55:41 +00:00
|
|
|
gradient_checkpointing,
|
|
|
|
gradient_accumulation_steps,
|
|
|
|
mem_eff_attn,
|
|
|
|
shuffle_caption,
|
2022-12-20 14:15:17 +00:00
|
|
|
):
|
|
|
|
original_file_path = file_path
|
|
|
|
|
|
|
|
save_as_bool = True if save_as.get('label') == 'True' else False
|
|
|
|
|
|
|
|
if save_as_bool:
|
|
|
|
print('Save as...')
|
|
|
|
file_path = get_saveasfile_path(file_path)
|
|
|
|
else:
|
|
|
|
print('Save...')
|
|
|
|
if file_path == None or file_path == '':
|
|
|
|
file_path = get_saveasfile_path(file_path)
|
|
|
|
|
|
|
|
# print(file_path)
|
|
|
|
|
|
|
|
if file_path == None:
|
|
|
|
return original_file_path
|
|
|
|
|
|
|
|
# Return the values of the variables as a dictionary
|
|
|
|
variables = {
|
|
|
|
'pretrained_model_name_or_path': pretrained_model_name_or_path,
|
|
|
|
'v2': v2,
|
|
|
|
'v_parameterization': v_parameterization,
|
|
|
|
'train_dir': train_dir,
|
|
|
|
'image_folder': image_folder,
|
|
|
|
'output_dir': output_dir,
|
|
|
|
'logging_dir': logging_dir,
|
|
|
|
'max_resolution': max_resolution,
|
2023-01-02 18:07:17 +00:00
|
|
|
'min_bucket_reso': min_bucket_reso,
|
|
|
|
'max_bucket_reso': max_bucket_reso,
|
|
|
|
'batch_size': batch_size,
|
|
|
|
'flip_aug': flip_aug,
|
|
|
|
'caption_metadata_filename': caption_metadata_filename,
|
|
|
|
'latent_metadata_filename': latent_metadata_filename,
|
|
|
|
'full_path': full_path,
|
2022-12-20 14:15:17 +00:00
|
|
|
'learning_rate': learning_rate,
|
|
|
|
'lr_scheduler': lr_scheduler,
|
|
|
|
'lr_warmup': lr_warmup,
|
|
|
|
'dataset_repeats': dataset_repeats,
|
|
|
|
'train_batch_size': train_batch_size,
|
|
|
|
'epoch': epoch,
|
|
|
|
'save_every_n_epochs': save_every_n_epochs,
|
|
|
|
'mixed_precision': mixed_precision,
|
|
|
|
'save_precision': save_precision,
|
|
|
|
'seed': seed,
|
|
|
|
'num_cpu_threads_per_process': num_cpu_threads_per_process,
|
|
|
|
'train_text_encoder': train_text_encoder,
|
|
|
|
'create_buckets': create_buckets,
|
|
|
|
'create_caption': create_caption,
|
|
|
|
'save_model_as': save_model_as,
|
|
|
|
'caption_extension': caption_extension,
|
2023-01-02 18:07:17 +00:00
|
|
|
'use_8bit_adam': use_8bit_adam,
|
|
|
|
'xformers': xformers,
|
2023-01-06 00:16:13 +00:00
|
|
|
'clip_skip': clip_skip,
|
2023-01-09 00:31:44 +00:00
|
|
|
'save_state': save_state,
|
|
|
|
'resume': resume,
|
2023-01-09 01:55:41 +00:00
|
|
|
'gradient_checkpointing': gradient_checkpointing,
|
|
|
|
'gradient_accumulation_steps': gradient_accumulation_steps,
|
|
|
|
'mem_eff_attn': mem_eff_attn,
|
|
|
|
'shuffle_caption': shuffle_caption,
|
2022-12-20 14:15:17 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
# Save the data to the selected file
|
2022-12-21 14:05:06 +00:00
|
|
|
with open(file_path, 'w') as file:
|
|
|
|
json.dump(variables, file)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
|
|
|
return file_path
|
|
|
|
|
|
|
|
|
|
|
|
def open_config_file(
|
|
|
|
file_path,
|
|
|
|
pretrained_model_name_or_path,
|
|
|
|
v2,
|
|
|
|
v_parameterization,
|
|
|
|
train_dir,
|
|
|
|
image_folder,
|
|
|
|
output_dir,
|
|
|
|
logging_dir,
|
|
|
|
max_resolution,
|
2023-01-02 18:07:17 +00:00
|
|
|
min_bucket_reso,
|
|
|
|
max_bucket_reso,
|
|
|
|
batch_size,
|
|
|
|
flip_aug,
|
|
|
|
caption_metadata_filename,
|
|
|
|
latent_metadata_filename,
|
|
|
|
full_path,
|
2022-12-20 14:15:17 +00:00
|
|
|
learning_rate,
|
|
|
|
lr_scheduler,
|
|
|
|
lr_warmup,
|
|
|
|
dataset_repeats,
|
|
|
|
train_batch_size,
|
|
|
|
epoch,
|
|
|
|
save_every_n_epochs,
|
|
|
|
mixed_precision,
|
|
|
|
save_precision,
|
|
|
|
seed,
|
|
|
|
num_cpu_threads_per_process,
|
|
|
|
train_text_encoder,
|
|
|
|
create_caption,
|
2023-01-02 18:07:17 +00:00
|
|
|
create_buckets,
|
2022-12-20 14:15:17 +00:00
|
|
|
save_model_as,
|
|
|
|
caption_extension,
|
2023-01-02 18:07:17 +00:00
|
|
|
use_8bit_adam,
|
|
|
|
xformers,
|
2023-01-06 00:16:13 +00:00
|
|
|
clip_skip,
|
2023-01-09 00:31:44 +00:00
|
|
|
save_state,
|
|
|
|
resume,
|
2023-01-09 01:55:41 +00:00
|
|
|
gradient_checkpointing,
|
|
|
|
gradient_accumulation_steps,
|
|
|
|
mem_eff_attn,
|
|
|
|
shuffle_caption,
|
2022-12-20 14:15:17 +00:00
|
|
|
):
|
|
|
|
original_file_path = file_path
|
|
|
|
file_path = get_file_path(file_path)
|
|
|
|
|
|
|
|
if file_path != '' and file_path != None:
|
2023-01-09 00:31:44 +00:00
|
|
|
print(f'Loading config file {file_path}')
|
2022-12-20 14:15:17 +00:00
|
|
|
# load variables from JSON file
|
|
|
|
with open(file_path, 'r') as f:
|
|
|
|
my_data = json.load(f)
|
|
|
|
else:
|
|
|
|
file_path = original_file_path # In case a file_path was provided and the user decide to cancel the open action
|
|
|
|
my_data = {}
|
|
|
|
|
|
|
|
# Return the values of the variables as a dictionary
|
|
|
|
return (
|
|
|
|
file_path,
|
|
|
|
my_data.get(
|
|
|
|
'pretrained_model_name_or_path', pretrained_model_name_or_path
|
|
|
|
),
|
|
|
|
my_data.get('v2', v2),
|
|
|
|
my_data.get('v_parameterization', v_parameterization),
|
|
|
|
my_data.get('train_dir', train_dir),
|
|
|
|
my_data.get('image_folder', image_folder),
|
|
|
|
my_data.get('output_dir', output_dir),
|
|
|
|
my_data.get('logging_dir', logging_dir),
|
|
|
|
my_data.get('max_resolution', max_resolution),
|
2023-01-02 18:07:17 +00:00
|
|
|
my_data.get('min_bucket_reso', min_bucket_reso),
|
|
|
|
my_data.get('max_bucket_reso', max_bucket_reso),
|
|
|
|
my_data.get('batch_size', batch_size),
|
|
|
|
my_data.get('flip_aug', flip_aug),
|
|
|
|
my_data.get('caption_metadata_filename', caption_metadata_filename),
|
|
|
|
my_data.get('latent_metadata_filename', latent_metadata_filename),
|
|
|
|
my_data.get('full_path', full_path),
|
2022-12-20 14:15:17 +00:00
|
|
|
my_data.get('learning_rate', learning_rate),
|
|
|
|
my_data.get('lr_scheduler', lr_scheduler),
|
|
|
|
my_data.get('lr_warmup', lr_warmup),
|
|
|
|
my_data.get('dataset_repeats', dataset_repeats),
|
|
|
|
my_data.get('train_batch_size', train_batch_size),
|
|
|
|
my_data.get('epoch', epoch),
|
|
|
|
my_data.get('save_every_n_epochs', save_every_n_epochs),
|
|
|
|
my_data.get('mixed_precision', mixed_precision),
|
|
|
|
my_data.get('save_precision', save_precision),
|
|
|
|
my_data.get('seed', seed),
|
|
|
|
my_data.get(
|
|
|
|
'num_cpu_threads_per_process', num_cpu_threads_per_process
|
|
|
|
),
|
|
|
|
my_data.get('train_text_encoder', train_text_encoder),
|
|
|
|
my_data.get('create_buckets', create_buckets),
|
|
|
|
my_data.get('create_caption', create_caption),
|
|
|
|
my_data.get('save_model_as', save_model_as),
|
|
|
|
my_data.get('caption_extension', caption_extension),
|
2023-01-02 18:07:17 +00:00
|
|
|
my_data.get('use_8bit_adam', use_8bit_adam),
|
|
|
|
my_data.get('xformers', xformers),
|
2023-01-06 00:16:13 +00:00
|
|
|
my_data.get('clip_skip', clip_skip),
|
2023-01-09 00:31:44 +00:00
|
|
|
my_data.get('save_state', save_state),
|
|
|
|
my_data.get('resume', resume),
|
2023-01-09 01:55:41 +00:00
|
|
|
my_data.get('gradient_checkpointing', gradient_checkpointing),
|
|
|
|
my_data.get('gradient_accumulation_steps', gradient_accumulation_steps),
|
|
|
|
my_data.get('mem_eff_attn', mem_eff_attn),
|
|
|
|
my_data.get('shuffle_caption', shuffle_caption),
|
2022-12-20 14:15:17 +00:00
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
def train_model(
|
|
|
|
pretrained_model_name_or_path,
|
|
|
|
v2,
|
|
|
|
v_parameterization,
|
|
|
|
train_dir,
|
|
|
|
image_folder,
|
|
|
|
output_dir,
|
|
|
|
logging_dir,
|
|
|
|
max_resolution,
|
2023-01-02 18:07:17 +00:00
|
|
|
min_bucket_reso,
|
|
|
|
max_bucket_reso,
|
|
|
|
batch_size,
|
|
|
|
flip_aug,
|
|
|
|
caption_metadata_filename,
|
|
|
|
latent_metadata_filename,
|
|
|
|
full_path,
|
2022-12-20 14:15:17 +00:00
|
|
|
learning_rate,
|
|
|
|
lr_scheduler,
|
|
|
|
lr_warmup,
|
|
|
|
dataset_repeats,
|
|
|
|
train_batch_size,
|
|
|
|
epoch,
|
|
|
|
save_every_n_epochs,
|
|
|
|
mixed_precision,
|
|
|
|
save_precision,
|
|
|
|
seed,
|
|
|
|
num_cpu_threads_per_process,
|
|
|
|
train_text_encoder,
|
2023-01-02 18:07:17 +00:00
|
|
|
generate_caption_database,
|
|
|
|
generate_image_buckets,
|
2022-12-20 14:15:17 +00:00
|
|
|
save_model_as,
|
|
|
|
caption_extension,
|
2023-01-02 18:07:17 +00:00
|
|
|
use_8bit_adam,
|
|
|
|
xformers,
|
2023-01-06 00:16:13 +00:00
|
|
|
clip_skip,
|
2023-01-09 00:31:44 +00:00
|
|
|
save_state,
|
|
|
|
resume,
|
2023-01-09 01:55:41 +00:00
|
|
|
gradient_checkpointing,
|
|
|
|
gradient_accumulation_steps,
|
|
|
|
mem_eff_attn,
|
|
|
|
shuffle_caption,
|
2022-12-20 14:15:17 +00:00
|
|
|
):
|
|
|
|
def save_inference_file(output_dir, v2, v_parameterization):
|
|
|
|
# Copy inference model for v2 if required
|
|
|
|
if v2 and v_parameterization:
|
|
|
|
print(f'Saving v2-inference-v.yaml as {output_dir}/last.yaml')
|
|
|
|
shutil.copy(
|
|
|
|
f'./v2_inference/v2-inference-v.yaml',
|
|
|
|
f'{output_dir}/last.yaml',
|
|
|
|
)
|
|
|
|
elif v2:
|
|
|
|
print(f'Saving v2-inference.yaml as {output_dir}/last.yaml')
|
|
|
|
shutil.copy(
|
|
|
|
f'./v2_inference/v2-inference.yaml',
|
|
|
|
f'{output_dir}/last.yaml',
|
|
|
|
)
|
|
|
|
|
|
|
|
# create caption json file
|
|
|
|
if generate_caption_database:
|
|
|
|
if not os.path.exists(train_dir):
|
|
|
|
os.mkdir(train_dir)
|
|
|
|
|
|
|
|
run_cmd = (
|
|
|
|
f'./venv/Scripts/python.exe finetune/merge_captions_to_metadata.py'
|
|
|
|
)
|
|
|
|
if caption_extension == '':
|
|
|
|
run_cmd += f' --caption_extension=".txt"'
|
|
|
|
else:
|
|
|
|
run_cmd += f' --caption_extension={caption_extension}'
|
2023-01-06 23:25:55 +00:00
|
|
|
run_cmd += f' "{image_folder}"'
|
|
|
|
run_cmd += f' "{train_dir}/{caption_metadata_filename}"'
|
2023-01-02 18:07:17 +00:00
|
|
|
if full_path:
|
|
|
|
run_cmd += f' --full_path'
|
2022-12-20 14:15:17 +00:00
|
|
|
|
|
|
|
print(run_cmd)
|
|
|
|
|
|
|
|
# Run the command
|
|
|
|
subprocess.run(run_cmd)
|
|
|
|
|
|
|
|
# create images buckets
|
|
|
|
if generate_image_buckets:
|
2023-01-02 18:07:17 +00:00
|
|
|
run_cmd = (
|
|
|
|
f'./venv/Scripts/python.exe finetune/prepare_buckets_latents.py'
|
|
|
|
)
|
2023-01-06 23:25:55 +00:00
|
|
|
run_cmd += f' "{image_folder}"'
|
|
|
|
run_cmd += f' "{train_dir}/{caption_metadata_filename}"'
|
|
|
|
run_cmd += f' "{train_dir}/{latent_metadata_filename}"'
|
|
|
|
run_cmd += f' "{pretrained_model_name_or_path}"'
|
2023-01-02 18:07:17 +00:00
|
|
|
run_cmd += f' --batch_size={batch_size}'
|
|
|
|
run_cmd += f' --max_resolution={max_resolution}'
|
|
|
|
run_cmd += f' --min_bucket_reso={min_bucket_reso}'
|
|
|
|
run_cmd += f' --max_bucket_reso={max_bucket_reso}'
|
|
|
|
run_cmd += f' --mixed_precision={mixed_precision}'
|
|
|
|
if flip_aug:
|
|
|
|
run_cmd += f' --flip_aug'
|
|
|
|
if full_path:
|
|
|
|
run_cmd += f' --full_path'
|
|
|
|
|
|
|
|
print(run_cmd)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
|
|
|
# Run the command
|
2023-01-02 18:07:17 +00:00
|
|
|
subprocess.run(run_cmd)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 18:14:46 +00:00
|
|
|
image_num = len(
|
|
|
|
[f for f in os.listdir(image_folder) if f.endswith('.npz')]
|
|
|
|
)
|
|
|
|
print(f'image_num = {image_num}')
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 18:14:46 +00:00
|
|
|
repeats = int(image_num) * int(dataset_repeats)
|
|
|
|
print(f'repeats = {str(repeats)}')
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 18:14:46 +00:00
|
|
|
# calculate max_train_steps
|
|
|
|
max_train_steps = int(
|
|
|
|
math.ceil(float(repeats) / int(train_batch_size) * int(epoch))
|
|
|
|
)
|
2023-01-02 18:07:17 +00:00
|
|
|
|
|
|
|
# Divide by two because flip augmentation create two copied of the source images
|
|
|
|
if flip_aug:
|
|
|
|
max_train_steps = int(math.ceil(float(max_train_steps) / 2))
|
|
|
|
|
2022-12-22 18:14:46 +00:00
|
|
|
print(f'max_train_steps = {max_train_steps}')
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2023-01-02 18:07:17 +00:00
|
|
|
lr_warmup_steps = round(float(int(lr_warmup) * int(max_train_steps) / 100))
|
2022-12-22 18:14:46 +00:00
|
|
|
print(f'lr_warmup_steps = {lr_warmup_steps}')
|
|
|
|
|
|
|
|
run_cmd = f'accelerate launch --num_cpu_threads_per_process={num_cpu_threads_per_process} "./fine_tune.py"'
|
|
|
|
if v2:
|
|
|
|
run_cmd += ' --v2'
|
|
|
|
if v_parameterization:
|
|
|
|
run_cmd += ' --v_parameterization'
|
|
|
|
if train_text_encoder:
|
|
|
|
run_cmd += ' --train_text_encoder'
|
2023-01-02 18:07:17 +00:00
|
|
|
if use_8bit_adam:
|
|
|
|
run_cmd += f' --use_8bit_adam'
|
|
|
|
if xformers:
|
|
|
|
run_cmd += f' --xformers'
|
2023-01-09 01:55:41 +00:00
|
|
|
if gradient_checkpointing:
|
|
|
|
run_cmd += ' --gradient_checkpointing'
|
|
|
|
if mem_eff_attn:
|
|
|
|
run_cmd += ' --mem_eff_attn'
|
|
|
|
if shuffle_caption:
|
|
|
|
run_cmd += ' --shuffle_caption'
|
2022-12-22 18:14:46 +00:00
|
|
|
run_cmd += (
|
2023-01-06 23:25:55 +00:00
|
|
|
f' --pretrained_model_name_or_path="{pretrained_model_name_or_path}"'
|
2022-12-22 18:14:46 +00:00
|
|
|
)
|
2023-01-06 23:25:55 +00:00
|
|
|
run_cmd += f' --in_json="{train_dir}/{latent_metadata_filename}"'
|
|
|
|
run_cmd += f' --train_data_dir="{image_folder}"'
|
|
|
|
run_cmd += f' --output_dir="{output_dir}"'
|
2022-12-22 18:14:46 +00:00
|
|
|
if not logging_dir == '':
|
2023-01-06 23:25:55 +00:00
|
|
|
run_cmd += f' --logging_dir="{logging_dir}"'
|
2022-12-22 18:14:46 +00:00
|
|
|
run_cmd += f' --train_batch_size={train_batch_size}'
|
|
|
|
run_cmd += f' --dataset_repeats={dataset_repeats}'
|
|
|
|
run_cmd += f' --learning_rate={learning_rate}'
|
|
|
|
run_cmd += f' --lr_scheduler={lr_scheduler}'
|
|
|
|
run_cmd += f' --lr_warmup_steps={lr_warmup_steps}'
|
|
|
|
run_cmd += f' --max_train_steps={max_train_steps}'
|
|
|
|
run_cmd += f' --mixed_precision={mixed_precision}'
|
|
|
|
run_cmd += f' --save_every_n_epochs={save_every_n_epochs}'
|
|
|
|
run_cmd += f' --seed={seed}'
|
|
|
|
run_cmd += f' --save_precision={save_precision}'
|
|
|
|
if not save_model_as == 'same as source model':
|
|
|
|
run_cmd += f' --save_model_as={save_model_as}'
|
2023-01-06 12:13:12 +00:00
|
|
|
if int(clip_skip) > 1:
|
|
|
|
run_cmd += f' --clip_skip={str(clip_skip)}'
|
2023-01-09 01:55:41 +00:00
|
|
|
if int(gradient_accumulation_steps) > 1:
|
|
|
|
run_cmd += f' --gradient_accumulation_steps={int(gradient_accumulation_steps)}'
|
2023-01-09 00:31:44 +00:00
|
|
|
if save_state:
|
|
|
|
run_cmd += ' --save_state'
|
|
|
|
if not resume == '':
|
|
|
|
run_cmd += f' --resume={resume}'
|
2022-12-22 18:14:46 +00:00
|
|
|
|
|
|
|
print(run_cmd)
|
|
|
|
# Run the command
|
|
|
|
subprocess.run(run_cmd)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
|
|
|
# check if output_dir/last is a folder... therefore it is a diffuser model
|
|
|
|
last_dir = pathlib.Path(f'{output_dir}/last')
|
|
|
|
|
|
|
|
if not last_dir.is_dir():
|
|
|
|
# Copy inference model for v2 if required
|
|
|
|
save_inference_file(output_dir, v2, v_parameterization)
|
|
|
|
|
|
|
|
|
|
|
|
def set_pretrained_model_name_or_path_input(value, v2, v_parameterization):
|
|
|
|
# define a list of substrings to search for
|
|
|
|
substrings_v2 = [
|
|
|
|
'stabilityai/stable-diffusion-2-1-base',
|
|
|
|
'stabilityai/stable-diffusion-2-base',
|
|
|
|
]
|
|
|
|
|
|
|
|
# check if $v2 and $v_parameterization are empty and if $pretrained_model_name_or_path contains any of the substrings in the v2 list
|
|
|
|
if str(value) in substrings_v2:
|
|
|
|
print('SD v2 model detected. Setting --v2 parameter')
|
|
|
|
v2 = True
|
|
|
|
v_parameterization = False
|
|
|
|
|
|
|
|
return value, v2, v_parameterization
|
|
|
|
|
|
|
|
# define a list of substrings to search for v-objective
|
|
|
|
substrings_v_parameterization = [
|
|
|
|
'stabilityai/stable-diffusion-2-1',
|
|
|
|
'stabilityai/stable-diffusion-2',
|
|
|
|
]
|
|
|
|
|
|
|
|
# check if $v2 and $v_parameterization are empty and if $pretrained_model_name_or_path contains any of the substrings in the v_parameterization list
|
|
|
|
if str(value) in substrings_v_parameterization:
|
|
|
|
print(
|
|
|
|
'SD v2 v_parameterization detected. Setting --v2 parameter and --v_parameterization'
|
|
|
|
)
|
|
|
|
v2 = True
|
|
|
|
v_parameterization = True
|
|
|
|
|
|
|
|
return value, v2, v_parameterization
|
|
|
|
|
|
|
|
# define a list of substrings to v1.x
|
|
|
|
substrings_v1_model = [
|
|
|
|
'CompVis/stable-diffusion-v1-4',
|
|
|
|
'runwayml/stable-diffusion-v1-5',
|
|
|
|
]
|
|
|
|
|
|
|
|
if str(value) in substrings_v1_model:
|
|
|
|
v2 = False
|
|
|
|
v_parameterization = False
|
|
|
|
|
|
|
|
return value, v2, v_parameterization
|
|
|
|
|
|
|
|
if value == 'custom':
|
|
|
|
value = ''
|
|
|
|
v2 = False
|
|
|
|
v_parameterization = False
|
|
|
|
|
|
|
|
return value, v2, v_parameterization
|
|
|
|
|
|
|
|
|
|
|
|
def remove_doublequote(file_path):
|
|
|
|
if file_path != None:
|
|
|
|
file_path = file_path.replace('"', '')
|
|
|
|
|
|
|
|
return file_path
|
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
|
2022-12-21 14:05:06 +00:00
|
|
|
def UI(username, password):
|
|
|
|
|
|
|
|
css = ''
|
|
|
|
|
|
|
|
if os.path.exists('./style.css'):
|
|
|
|
with open(os.path.join('./style.css'), 'r', encoding='utf8') as file:
|
|
|
|
print('Load CSS...')
|
|
|
|
css += file.read() + '\n'
|
|
|
|
|
|
|
|
interface = gr.Blocks(css=css)
|
|
|
|
|
|
|
|
with interface:
|
2023-01-02 18:07:17 +00:00
|
|
|
with gr.Tab('Finetune'):
|
2022-12-22 16:51:34 +00:00
|
|
|
finetune_tab()
|
2023-01-02 18:07:17 +00:00
|
|
|
with gr.Tab('Utilities'):
|
2022-12-22 16:51:34 +00:00
|
|
|
utilities_tab(enable_dreambooth_tab=False)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
# Show the interface
|
|
|
|
if not username == '':
|
|
|
|
interface.launch(auth=(username, password))
|
|
|
|
else:
|
|
|
|
interface.launch()
|
|
|
|
|
2023-01-02 18:07:17 +00:00
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
def finetune_tab():
|
|
|
|
dummy_ft_true = gr.Label(value=True, visible=False)
|
|
|
|
dummy_ft_false = gr.Label(value=False, visible=False)
|
2023-01-02 18:07:17 +00:00
|
|
|
gr.Markdown('Train a custom model using kohya finetune python code...')
|
2022-12-23 01:18:51 +00:00
|
|
|
with gr.Accordion('Configuration file', open=False):
|
2022-12-22 16:51:34 +00:00
|
|
|
with gr.Row():
|
|
|
|
button_open_config = gr.Button(
|
|
|
|
f'Open {folder_symbol}', elem_id='open_folder'
|
|
|
|
)
|
|
|
|
button_save_config = gr.Button(
|
|
|
|
f'Save {save_style_symbol}', elem_id='open_folder'
|
|
|
|
)
|
|
|
|
button_save_as_config = gr.Button(
|
|
|
|
f'Save as... {save_style_symbol}',
|
|
|
|
elem_id='open_folder',
|
|
|
|
)
|
|
|
|
config_file_name = gr.Textbox(
|
|
|
|
label='', placeholder='type file path or use buttons...'
|
|
|
|
)
|
|
|
|
config_file_name.change(
|
|
|
|
remove_doublequote,
|
|
|
|
inputs=[config_file_name],
|
|
|
|
outputs=[config_file_name],
|
|
|
|
)
|
|
|
|
with gr.Tab('Source model'):
|
|
|
|
# Define the input elements
|
|
|
|
with gr.Row():
|
|
|
|
pretrained_model_name_or_path_input = gr.Textbox(
|
|
|
|
label='Pretrained model name or path',
|
|
|
|
placeholder='enter the path to custom model or name of pretrained model',
|
|
|
|
)
|
|
|
|
pretrained_model_name_or_path_file = gr.Button(
|
|
|
|
document_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
|
|
|
pretrained_model_name_or_path_file.click(
|
2023-01-01 19:14:58 +00:00
|
|
|
get_any_file_path,
|
2022-12-22 16:51:34 +00:00
|
|
|
inputs=pretrained_model_name_or_path_input,
|
|
|
|
outputs=pretrained_model_name_or_path_input,
|
|
|
|
)
|
|
|
|
pretrained_model_name_or_path_folder = gr.Button(
|
|
|
|
folder_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
|
|
|
pretrained_model_name_or_path_folder.click(
|
|
|
|
get_folder_path,
|
|
|
|
inputs=pretrained_model_name_or_path_input,
|
|
|
|
outputs=pretrained_model_name_or_path_input,
|
|
|
|
)
|
|
|
|
model_list = gr.Dropdown(
|
|
|
|
label='(Optional) Model Quick Pick',
|
|
|
|
choices=[
|
|
|
|
'custom',
|
|
|
|
'stabilityai/stable-diffusion-2-1-base',
|
|
|
|
'stabilityai/stable-diffusion-2-base',
|
|
|
|
'stabilityai/stable-diffusion-2-1',
|
|
|
|
'stabilityai/stable-diffusion-2',
|
|
|
|
'runwayml/stable-diffusion-v1-5',
|
|
|
|
'CompVis/stable-diffusion-v1-4',
|
2022-12-21 14:05:06 +00:00
|
|
|
],
|
2022-12-22 16:51:34 +00:00
|
|
|
)
|
|
|
|
save_model_as_dropdown = gr.Dropdown(
|
|
|
|
label='Save trained model as',
|
|
|
|
choices=[
|
|
|
|
'same as source model',
|
|
|
|
'ckpt',
|
|
|
|
'diffusers',
|
|
|
|
'diffusers_safetensors',
|
|
|
|
'safetensors',
|
2022-12-20 14:15:17 +00:00
|
|
|
],
|
2022-12-22 16:51:34 +00:00
|
|
|
value='same as source model',
|
2022-12-20 14:15:17 +00:00
|
|
|
)
|
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
with gr.Row():
|
|
|
|
v2_input = gr.Checkbox(label='v2', value=True)
|
|
|
|
v_parameterization_input = gr.Checkbox(
|
|
|
|
label='v_parameterization', value=False
|
|
|
|
)
|
|
|
|
model_list.change(
|
|
|
|
set_pretrained_model_name_or_path_input,
|
|
|
|
inputs=[model_list, v2_input, v_parameterization_input],
|
|
|
|
outputs=[
|
|
|
|
pretrained_model_name_or_path_input,
|
|
|
|
v2_input,
|
|
|
|
v_parameterization_input,
|
|
|
|
],
|
|
|
|
)
|
2022-12-23 01:18:51 +00:00
|
|
|
with gr.Tab('Folders'):
|
2022-12-22 16:51:34 +00:00
|
|
|
with gr.Row():
|
|
|
|
train_dir_input = gr.Textbox(
|
|
|
|
label='Training config folder',
|
|
|
|
placeholder='folder where the training configuration files will be saved',
|
|
|
|
)
|
|
|
|
train_dir_folder = gr.Button(
|
|
|
|
folder_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
2023-01-02 18:07:17 +00:00
|
|
|
train_dir_folder.click(get_folder_path, outputs=train_dir_input)
|
2022-12-22 16:51:34 +00:00
|
|
|
|
|
|
|
image_folder_input = gr.Textbox(
|
|
|
|
label='Training Image folder',
|
|
|
|
placeholder='folder where the training images are located',
|
|
|
|
)
|
|
|
|
image_folder_input_folder = gr.Button(
|
|
|
|
folder_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
|
|
|
image_folder_input_folder.click(
|
|
|
|
get_folder_path, outputs=image_folder_input
|
|
|
|
)
|
|
|
|
with gr.Row():
|
|
|
|
output_dir_input = gr.Textbox(
|
|
|
|
label='Output folder',
|
|
|
|
placeholder='folder where the model will be saved',
|
|
|
|
)
|
|
|
|
output_dir_input_folder = gr.Button(
|
|
|
|
folder_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
|
|
|
output_dir_input_folder.click(
|
|
|
|
get_folder_path, outputs=output_dir_input
|
|
|
|
)
|
|
|
|
|
|
|
|
logging_dir_input = gr.Textbox(
|
|
|
|
label='Logging folder',
|
|
|
|
placeholder='Optional: enable logging and output TensorBoard log to this folder',
|
|
|
|
)
|
|
|
|
logging_dir_input_folder = gr.Button(
|
|
|
|
folder_symbol, elem_id='open_folder_small'
|
|
|
|
)
|
|
|
|
logging_dir_input_folder.click(
|
|
|
|
get_folder_path, outputs=logging_dir_input
|
|
|
|
)
|
|
|
|
train_dir_input.change(
|
|
|
|
remove_doublequote,
|
|
|
|
inputs=[train_dir_input],
|
|
|
|
outputs=[train_dir_input],
|
|
|
|
)
|
|
|
|
image_folder_input.change(
|
|
|
|
remove_doublequote,
|
|
|
|
inputs=[image_folder_input],
|
|
|
|
outputs=[image_folder_input],
|
|
|
|
)
|
|
|
|
output_dir_input.change(
|
|
|
|
remove_doublequote,
|
|
|
|
inputs=[output_dir_input],
|
|
|
|
outputs=[output_dir_input],
|
|
|
|
)
|
2023-01-02 18:07:17 +00:00
|
|
|
with gr.Tab('Dataset preparation'):
|
|
|
|
with gr.Row():
|
|
|
|
max_resolution_input = gr.Textbox(
|
|
|
|
label='Resolution (width,height)', value='512,512'
|
|
|
|
)
|
|
|
|
min_bucket_reso = gr.Textbox(
|
|
|
|
label='Min bucket resolution', value='256'
|
|
|
|
)
|
|
|
|
max_bucket_reso = gr.Textbox(
|
|
|
|
label='Max bucket resolution', value='1024'
|
|
|
|
)
|
|
|
|
batch_size = gr.Textbox(label='Batch size', value='1')
|
|
|
|
with gr.Accordion('Advanced parameters', open=False):
|
|
|
|
with gr.Row():
|
|
|
|
caption_metadata_filename = gr.Textbox(
|
|
|
|
label='Caption metadata filename', value='meta_cap.json'
|
|
|
|
)
|
|
|
|
latent_metadata_filename = gr.Textbox(
|
|
|
|
label='Latent metadata filename', value='meta_lat.json'
|
|
|
|
)
|
|
|
|
full_path = gr.Checkbox(label='Use full path', value=True)
|
|
|
|
flip_aug = gr.Checkbox(label='Flip augmentation', value=False)
|
2022-12-22 16:51:34 +00:00
|
|
|
with gr.Tab('Training parameters'):
|
|
|
|
with gr.Row():
|
2023-01-02 18:07:17 +00:00
|
|
|
learning_rate_input = gr.Textbox(label='Learning rate', value=1e-6)
|
2022-12-22 16:51:34 +00:00
|
|
|
lr_scheduler_input = gr.Dropdown(
|
|
|
|
label='LR Scheduler',
|
|
|
|
choices=[
|
|
|
|
'constant',
|
|
|
|
'constant_with_warmup',
|
|
|
|
'cosine',
|
|
|
|
'cosine_with_restarts',
|
|
|
|
'linear',
|
|
|
|
'polynomial',
|
2022-12-21 14:05:06 +00:00
|
|
|
],
|
2022-12-22 16:51:34 +00:00
|
|
|
value='constant',
|
|
|
|
)
|
|
|
|
lr_warmup_input = gr.Textbox(label='LR warmup', value=0)
|
|
|
|
with gr.Row():
|
|
|
|
dataset_repeats_input = gr.Textbox(
|
|
|
|
label='Dataset repeats', value=40
|
|
|
|
)
|
|
|
|
train_batch_size_input = gr.Slider(
|
|
|
|
minimum=1,
|
|
|
|
maximum=32,
|
|
|
|
label='Train batch size',
|
|
|
|
value=1,
|
|
|
|
step=1,
|
|
|
|
)
|
|
|
|
epoch_input = gr.Textbox(label='Epoch', value=1)
|
|
|
|
save_every_n_epochs_input = gr.Textbox(
|
|
|
|
label='Save every N epochs', value=1
|
|
|
|
)
|
|
|
|
with gr.Row():
|
|
|
|
mixed_precision_input = gr.Dropdown(
|
|
|
|
label='Mixed precision',
|
|
|
|
choices=[
|
|
|
|
'no',
|
|
|
|
'fp16',
|
|
|
|
'bf16',
|
|
|
|
],
|
|
|
|
value='fp16',
|
|
|
|
)
|
|
|
|
save_precision_input = gr.Dropdown(
|
|
|
|
label='Save precision',
|
|
|
|
choices=[
|
|
|
|
'float',
|
|
|
|
'fp16',
|
|
|
|
'bf16',
|
2022-12-21 14:05:06 +00:00
|
|
|
],
|
2022-12-22 16:51:34 +00:00
|
|
|
value='fp16',
|
|
|
|
)
|
|
|
|
num_cpu_threads_per_process_input = gr.Slider(
|
|
|
|
minimum=1,
|
|
|
|
maximum=os.cpu_count(),
|
|
|
|
step=1,
|
|
|
|
label='Number of CPU threads per process',
|
|
|
|
value=os.cpu_count(),
|
|
|
|
)
|
|
|
|
seed_input = gr.Textbox(label='Seed', value=1234)
|
|
|
|
with gr.Row():
|
|
|
|
caption_extention_input = gr.Textbox(
|
|
|
|
label='Caption Extension',
|
|
|
|
placeholder='(Optional) Extension for caption files. default: .txt',
|
2022-12-21 14:05:06 +00:00
|
|
|
)
|
2022-12-22 16:51:34 +00:00
|
|
|
train_text_encoder_input = gr.Checkbox(
|
|
|
|
label='Train text encoder', value=True
|
|
|
|
)
|
2023-01-02 18:07:17 +00:00
|
|
|
with gr.Accordion('Advanced parameters', open=False):
|
|
|
|
with gr.Row():
|
|
|
|
use_8bit_adam = gr.Checkbox(label='Use 8bit adam', value=True)
|
|
|
|
xformers = gr.Checkbox(label='Use xformers', value=True)
|
2023-01-06 00:16:13 +00:00
|
|
|
clip_skip = gr.Slider(
|
|
|
|
label='Clip skip', value='1', minimum=1, maximum=12, step=1
|
|
|
|
)
|
2023-01-09 01:55:41 +00:00
|
|
|
mem_eff_attn = gr.Checkbox(
|
|
|
|
label='Memory efficient attention', value=False
|
|
|
|
)
|
|
|
|
shuffle_caption = gr.Checkbox(
|
|
|
|
label='Shuffle caption', value=False
|
|
|
|
)
|
2023-01-09 00:31:44 +00:00
|
|
|
with gr.Row():
|
|
|
|
save_state = gr.Checkbox(
|
|
|
|
label='Save training state', value=False
|
|
|
|
)
|
|
|
|
resume = gr.Textbox(
|
|
|
|
label='Resume from saved training state',
|
|
|
|
placeholder='path to "last-state" state folder to resume from',
|
|
|
|
)
|
|
|
|
resume_button = gr.Button('📂', elem_id='open_folder_small')
|
|
|
|
resume_button.click(get_folder_path, outputs=resume)
|
2023-01-09 01:55:41 +00:00
|
|
|
gradient_checkpointing = gr.Checkbox(
|
|
|
|
label='Gradient checkpointing', value=False
|
|
|
|
)
|
|
|
|
gradient_accumulation_steps = gr.Number(
|
|
|
|
label='Gradient accumulate steps', value='1'
|
|
|
|
)
|
2022-12-22 16:51:34 +00:00
|
|
|
with gr.Box():
|
|
|
|
with gr.Row():
|
|
|
|
create_caption = gr.Checkbox(
|
2023-01-02 18:07:17 +00:00
|
|
|
label='Generate caption metadata', value=True
|
2022-12-22 16:51:34 +00:00
|
|
|
)
|
|
|
|
create_buckets = gr.Checkbox(
|
2023-01-02 18:07:17 +00:00
|
|
|
label='Generate image buckets metadata', value=True
|
2022-12-22 16:51:34 +00:00
|
|
|
)
|
|
|
|
|
2022-12-22 18:14:46 +00:00
|
|
|
button_run = gr.Button('Train model')
|
2022-12-22 16:51:34 +00:00
|
|
|
|
2023-01-02 18:07:17 +00:00
|
|
|
settings_list = [
|
|
|
|
pretrained_model_name_or_path_input,
|
|
|
|
v2_input,
|
|
|
|
v_parameterization_input,
|
|
|
|
train_dir_input,
|
|
|
|
image_folder_input,
|
|
|
|
output_dir_input,
|
|
|
|
logging_dir_input,
|
|
|
|
max_resolution_input,
|
|
|
|
min_bucket_reso,
|
|
|
|
max_bucket_reso,
|
|
|
|
batch_size,
|
|
|
|
flip_aug,
|
|
|
|
caption_metadata_filename,
|
|
|
|
latent_metadata_filename,
|
|
|
|
full_path,
|
|
|
|
learning_rate_input,
|
|
|
|
lr_scheduler_input,
|
|
|
|
lr_warmup_input,
|
|
|
|
dataset_repeats_input,
|
|
|
|
train_batch_size_input,
|
|
|
|
epoch_input,
|
|
|
|
save_every_n_epochs_input,
|
|
|
|
mixed_precision_input,
|
|
|
|
save_precision_input,
|
|
|
|
seed_input,
|
|
|
|
num_cpu_threads_per_process_input,
|
|
|
|
train_text_encoder_input,
|
|
|
|
create_caption,
|
|
|
|
create_buckets,
|
|
|
|
save_model_as_dropdown,
|
|
|
|
caption_extention_input,
|
|
|
|
use_8bit_adam,
|
|
|
|
xformers,
|
2023-01-06 00:16:13 +00:00
|
|
|
clip_skip,
|
2023-01-09 00:31:44 +00:00
|
|
|
save_state,
|
|
|
|
resume,
|
2023-01-09 01:55:41 +00:00
|
|
|
gradient_checkpointing,
|
|
|
|
gradient_accumulation_steps,
|
|
|
|
mem_eff_attn,
|
|
|
|
shuffle_caption,
|
2023-01-02 18:07:17 +00:00
|
|
|
]
|
|
|
|
|
|
|
|
button_run.click(train_model, inputs=settings_list)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
button_open_config.click(
|
|
|
|
open_config_file,
|
2023-01-02 18:07:17 +00:00
|
|
|
inputs=[config_file_name] + settings_list,
|
|
|
|
outputs=[config_file_name] + settings_list,
|
2022-12-22 16:51:34 +00:00
|
|
|
)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
button_save_config.click(
|
|
|
|
save_configuration,
|
2023-01-02 18:07:17 +00:00
|
|
|
inputs=[dummy_ft_false, config_file_name] + settings_list,
|
2022-12-22 16:51:34 +00:00
|
|
|
outputs=[config_file_name],
|
|
|
|
)
|
2022-12-20 14:15:17 +00:00
|
|
|
|
2022-12-22 16:51:34 +00:00
|
|
|
button_save_as_config.click(
|
|
|
|
save_configuration,
|
2023-01-02 18:07:17 +00:00
|
|
|
inputs=[dummy_ft_true, config_file_name] + settings_list,
|
2022-12-22 16:51:34 +00:00
|
|
|
outputs=[config_file_name],
|
|
|
|
)
|
2022-12-21 14:05:06 +00:00
|
|
|
|
|
|
|
|
|
|
|
if __name__ == '__main__':
|
2022-12-22 16:51:34 +00:00
|
|
|
# torch.cuda.set_per_process_memory_fraction(0.48)
|
|
|
|
parser = argparse.ArgumentParser()
|
|
|
|
parser.add_argument(
|
|
|
|
'--username', type=str, default='', help='Username for authentication'
|
|
|
|
)
|
|
|
|
parser.add_argument(
|
|
|
|
'--password', type=str, default='', help='Password for authentication'
|
|
|
|
)
|
|
|
|
|
|
|
|
args = parser.parse_args()
|
|
|
|
|
|
|
|
UI(username=args.username, password=args.password)
|