mirror of
https://github.com/ostris/ai-toolkit.git
synced 2026-01-26 16:39:47 +00:00
WIP diffusers pipeline is weird. Starting to hate sdxl
This commit is contained in:
Jaret Burkett
@@ -16,7 +16,8 @@ sys.path.append(REPOS_ROOT)
|
||||
|
||||
process_dict = {
|
||||
'vae': 'TrainVAEProcess',
|
||||
'slider': 'TrainSliderProcess',
|
||||
'slider_dev': 'TrainSliderProcess',
|
||||
'slider': 'TrainSliderProcessOld',
|
||||
'lora_hack': 'TrainLoRAHack',
|
||||
'rescale_sd': 'TrainSDRescaleProcess',
|
||||
}
|
||||
|
||||
@@ -5,6 +5,9 @@ from collections import OrderedDict
|
||||
import os
|
||||
from typing import Optional
|
||||
|
||||
from safetensors.torch import save_file, load_file
|
||||
from tqdm import tqdm
|
||||
|
||||
from toolkit.config_modules import SliderConfig
|
||||
from toolkit.paths import REPOS_ROOT
|
||||
import sys
|
||||
@@ -35,28 +38,35 @@ def flush():
|
||||
class EncodedPromptPair:
|
||||
def __init__(
|
||||
self,
|
||||
target_class,
|
||||
positive,
|
||||
negative,
|
||||
positive_target,
|
||||
positive_target_with_neutral,
|
||||
negative_target,
|
||||
negative_target_with_neutral,
|
||||
neutral,
|
||||
width=512,
|
||||
height=512,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=1.0,
|
||||
weight=1.0
|
||||
both_targets,
|
||||
empty_prompt
|
||||
):
|
||||
self.target_class = target_class
|
||||
self.positive = positive
|
||||
self.negative = negative
|
||||
self.positive_target = positive_target
|
||||
self.positive_target_with_neutral = positive_target_with_neutral
|
||||
self.negative_target = negative_target
|
||||
self.negative_target_with_neutral = negative_target_with_neutral
|
||||
self.neutral = neutral
|
||||
self.width = width
|
||||
self.height = height
|
||||
self.action: int = action
|
||||
self.multiplier = multiplier
|
||||
self.weight = weight
|
||||
self.empty_prompt = empty_prompt
|
||||
self.both_targets = both_targets
|
||||
|
||||
# simulate torch to for tensors
|
||||
def to(self, *args, **kwargs):
|
||||
self.positive_target = self.positive_target.to(*args, **kwargs)
|
||||
self.positive_target_with_neutral = self.positive_target_with_neutral.to(*args, **kwargs)
|
||||
self.negative_target = self.negative_target.to(*args, **kwargs)
|
||||
self.negative_target_with_neutral = self.negative_target_with_neutral.to(*args, **kwargs)
|
||||
self.neutral = self.neutral.to(*args, **kwargs)
|
||||
self.empty_prompt = self.empty_prompt.to(*args, **kwargs)
|
||||
self.both_targets = self.both_targets.to(*args, **kwargs)
|
||||
return self
|
||||
|
||||
|
||||
class PromptEmbedsCache: # 使いまわしたいので
|
||||
class PromptEmbedsCache:
|
||||
prompts: dict[str, PromptEmbeds] = {}
|
||||
|
||||
def __setitem__(self, __name: str, __value: PromptEmbeds) -> None:
|
||||
@@ -84,6 +94,7 @@ class EncodedAnchor:
|
||||
class TrainSliderProcess(BaseSDTrainProcess):
|
||||
def __init__(self, process_id: int, job, config: OrderedDict):
|
||||
super().__init__(process_id, job, config)
|
||||
self.prompt_txt_list = None
|
||||
self.step_num = 0
|
||||
self.start_step = 0
|
||||
self.device = self.get_conf('device', self.job.device)
|
||||
@@ -97,115 +108,95 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
pass
|
||||
|
||||
def hook_before_train_loop(self):
|
||||
self.print(f"Loading prompt file from {self.slider_config.prompt_file}")
|
||||
|
||||
# read line by line from file
|
||||
with open(self.slider_config.prompt_file, 'r') as f:
|
||||
self.prompt_txt_list = f.readlines()
|
||||
# clean empty lines
|
||||
self.prompt_txt_list = [line.strip() for line in self.prompt_txt_list if len(line.strip()) > 0]
|
||||
|
||||
self.print(f"Loaded {len(self.prompt_txt_list)} prompts. Encoding them..")
|
||||
|
||||
cache = PromptEmbedsCache()
|
||||
prompt_pairs: list[EncodedPromptPair] = []
|
||||
|
||||
# get encoded latents for our prompts
|
||||
with torch.no_grad():
|
||||
neutral = ""
|
||||
for target in self.slider_config.targets:
|
||||
# build the cache
|
||||
for prompt in [
|
||||
target.target_class,
|
||||
target.positive,
|
||||
target.negative,
|
||||
neutral # empty neutral
|
||||
]:
|
||||
if cache[prompt] is None:
|
||||
cache[prompt] = self.sd.encode_prompt(prompt)
|
||||
for resolution in self.slider_config.resolutions:
|
||||
width, height = resolution
|
||||
only_erase = len(target.positive.strip()) == 0
|
||||
only_enhance = len(target.negative.strip()) == 0
|
||||
if self.slider_config.prompt_tensors is not None:
|
||||
# check to see if it exists
|
||||
if os.path.exists(self.slider_config.prompt_tensors):
|
||||
# load it.
|
||||
self.print(f"Loading prompt tensors from {self.slider_config.prompt_tensors}")
|
||||
prompt_tensors = load_file(self.slider_config.prompt_tensors, device='cpu')
|
||||
# add them to the cache
|
||||
for prompt_txt, prompt_tensor in prompt_tensors.items():
|
||||
if prompt_txt.startswith("te:"):
|
||||
prompt = prompt_txt[3:]
|
||||
# text_embeds
|
||||
text_embeds = prompt_tensor
|
||||
pooled_embeds = None
|
||||
# find pool embeds
|
||||
if f"pe:{prompt}" in prompt_tensors:
|
||||
pooled_embeds = prompt_tensors[f"pe:{prompt}"]
|
||||
|
||||
both = not only_erase and not only_enhance
|
||||
# make it
|
||||
prompt_embeds = PromptEmbeds([text_embeds, pooled_embeds])
|
||||
cache[prompt] = prompt_embeds.to(device='cpu', dtype=torch.float32)
|
||||
|
||||
if only_erase and only_enhance:
|
||||
raise ValueError("target must have at least one of positive or negative or both")
|
||||
# for slider we need to have an enhancer, an eraser, and then
|
||||
# an inverse with negative weights to balance the network
|
||||
# if we don't do this, we will get different contrast and focus.
|
||||
# we only perform actions of enhancing and erasing on the negative
|
||||
# todo work on way to do all of this in one shot
|
||||
if len(cache.prompts) == 0:
|
||||
print("Prompt tensors not found. Encoding prompts..")
|
||||
empty_prompt = ""
|
||||
# encode empty_prompt
|
||||
cache[empty_prompt] = self.sd.encode_prompt(empty_prompt)
|
||||
|
||||
if both or only_erase:
|
||||
prompt_pairs += [
|
||||
# erase standard
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.positive],
|
||||
negative=cache[target.negative],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=target.multiplier,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
if both or only_enhance:
|
||||
prompt_pairs += [
|
||||
# enhance standard, swap pos neg
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.negative],
|
||||
negative=cache[target.positive],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ENHANCE_NEGATIVE,
|
||||
multiplier=target.multiplier,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
if both:
|
||||
prompt_pairs += [
|
||||
# erase inverted
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.negative],
|
||||
negative=cache[target.positive],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=target.multiplier * -1.0,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
prompt_pairs += [
|
||||
# enhance inverted
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.positive],
|
||||
negative=cache[target.negative],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ENHANCE_NEGATIVE,
|
||||
multiplier=target.multiplier * -1.0,
|
||||
weight=target.weight
|
||||
),
|
||||
for neutral in tqdm(self.prompt_txt_list, desc="Encoding prompts", leave=False):
|
||||
for target in self.slider_config.targets:
|
||||
prompt_list = [
|
||||
f"{target.positive}", # positive_target
|
||||
f"{target.positive} {neutral}", # positive_target with neutral
|
||||
f"{target.negative}", # negative_target
|
||||
f"{target.negative} {neutral}", # negative_target with neutral
|
||||
f"{neutral}", # neutral
|
||||
f"{target.positive} {target.negative}", # both targets
|
||||
f"{target.negative} {target.positive}", # both targets
|
||||
]
|
||||
for p in prompt_list:
|
||||
# build the cache
|
||||
if cache[p] is None:
|
||||
cache[p] = self.sd.encode_prompt(p).to(device="cpu", dtype=torch.float32)
|
||||
|
||||
# setup anchors
|
||||
anchor_pairs = []
|
||||
for anchor in self.slider_config.anchors:
|
||||
# build the cache
|
||||
for prompt in [
|
||||
anchor.prompt,
|
||||
anchor.neg_prompt # empty neutral
|
||||
]:
|
||||
if cache[prompt] == None:
|
||||
cache[prompt] = self.sd.encode_prompt(prompt)
|
||||
if self.slider_config.prompt_tensors:
|
||||
print(f"Saving prompt tensors to {self.slider_config.prompt_tensors}")
|
||||
state_dict = {}
|
||||
for prompt_txt, prompt_embeds in cache.prompts.items():
|
||||
state_dict[f"te:{prompt_txt}"] = prompt_embeds.text_embeds.to("cpu",
|
||||
dtype=get_torch_dtype('fp16'))
|
||||
if prompt_embeds.pooled_embeds is not None:
|
||||
state_dict[f"pe:{prompt_txt}"] = prompt_embeds.pooled_embeds.to("cpu",
|
||||
dtype=get_torch_dtype(
|
||||
'fp16'))
|
||||
save_file(state_dict, self.slider_config.prompt_tensors)
|
||||
|
||||
anchor_pairs += [
|
||||
EncodedAnchor(
|
||||
prompt=cache[anchor.prompt],
|
||||
neg_prompt=cache[anchor.neg_prompt],
|
||||
multiplier=anchor.multiplier
|
||||
)
|
||||
]
|
||||
self.print("Encoding complete. Building prompt pairs..")
|
||||
for neutral in self.prompt_txt_list:
|
||||
for target in self.slider_config.targets:
|
||||
both_prompts_list = [
|
||||
f"{target.positive} {target.negative}",
|
||||
f"{target.negative} {target.positive}",
|
||||
]
|
||||
# randomly pick one of the both prompts to prevent bias
|
||||
both_prompts = both_prompts_list[torch.randint(0, 2, (1,)).item()]
|
||||
|
||||
prompt_pair = EncodedPromptPair(
|
||||
positive_target=cache[f"{target.positive}"],
|
||||
positive_target_with_neutral=cache[f"{target.positive} {neutral}"],
|
||||
negative_target=cache[f"{target.negative}"],
|
||||
negative_target_with_neutral=cache[f"{target.negative} {neutral}"],
|
||||
neutral=cache[neutral],
|
||||
both_targets=cache[both_prompts],
|
||||
empty_prompt=cache[""],
|
||||
).to(device="cpu", dtype=torch.float32)
|
||||
self.prompt_pairs.append(prompt_pair)
|
||||
|
||||
# move to cpu to save vram
|
||||
# We don't need text encoder anymore, but keep it on cpu for sampling
|
||||
@@ -216,8 +207,7 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
else:
|
||||
self.sd.text_encoder.to("cpu")
|
||||
self.prompt_cache = cache
|
||||
self.prompt_pairs = prompt_pairs
|
||||
self.anchor_pairs = anchor_pairs
|
||||
|
||||
flush()
|
||||
# end hook_before_train_loop
|
||||
|
||||
@@ -228,15 +218,13 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
prompt_pair: EncodedPromptPair = self.prompt_pairs[
|
||||
torch.randint(0, len(self.prompt_pairs), (1,)).item()
|
||||
]
|
||||
# move to device and dtype
|
||||
prompt_pair.to(self.device_torch, dtype=dtype)
|
||||
|
||||
height = prompt_pair.height
|
||||
width = prompt_pair.width
|
||||
target_class = prompt_pair.target_class
|
||||
neutral = prompt_pair.neutral
|
||||
negative = prompt_pair.negative
|
||||
positive = prompt_pair.positive
|
||||
weight = prompt_pair.weight
|
||||
multiplier = prompt_pair.multiplier
|
||||
# get a random resolution
|
||||
height, width = self.slider_config.resolutions[
|
||||
torch.randint(0, len(self.slider_config.resolutions), (1,)).item()
|
||||
]
|
||||
|
||||
unet = self.sd.unet
|
||||
noise_scheduler = self.sd.noise_scheduler
|
||||
@@ -244,21 +232,6 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
lr_scheduler = self.lr_scheduler
|
||||
loss_function = torch.nn.MSELoss()
|
||||
|
||||
def get_noise_pred(p, n, gs, cts, dn):
|
||||
return self.predict_noise(
|
||||
latents=dn,
|
||||
text_embeddings=train_tools.concat_prompt_embeddings(
|
||||
p, # unconditional
|
||||
n, # positive
|
||||
self.train_config.batch_size,
|
||||
),
|
||||
timestep=cts,
|
||||
guidance_scale=gs,
|
||||
)
|
||||
|
||||
# set network multiplier
|
||||
self.network.multiplier = multiplier
|
||||
|
||||
with torch.no_grad():
|
||||
self.sd.noise_scheduler.set_timesteps(
|
||||
self.train_config.max_denoising_steps, device=self.device_torch
|
||||
@@ -281,99 +254,154 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
latents = noise * self.sd.noise_scheduler.init_noise_sigma
|
||||
latents = latents.to(self.device_torch, dtype=dtype)
|
||||
|
||||
with self.network:
|
||||
assert self.network.is_active
|
||||
self.network.multiplier = multiplier
|
||||
denoised_latents = self.diffuse_some_steps(
|
||||
latents, # pass simple noise latents
|
||||
train_tools.concat_prompt_embeddings(
|
||||
positive, # unconditional
|
||||
target_class, # target
|
||||
self.train_config.batch_size,
|
||||
),
|
||||
start_timesteps=0,
|
||||
total_timesteps=timesteps_to,
|
||||
guidance_scale=3,
|
||||
)
|
||||
denoised_fraction = timesteps_to / (self.train_config.max_denoising_steps + 1)
|
||||
self.sd.pipeline.to(self.device_torch)
|
||||
torch.set_default_device(self.device_torch)
|
||||
self.sd.pipeline.set_progress_bar_config(disable=True)
|
||||
|
||||
# get generate semi denoised latents without network
|
||||
# only neutrap in positive and both targets in negative
|
||||
assert not self.network.is_active
|
||||
# denoised_latents = self.sd.pipeline(
|
||||
# num_inference_steps=self.train_config.max_denoising_steps,
|
||||
# denoising_end=denoised_fraction,
|
||||
# latents=latents,
|
||||
# prompt_embeds=prompt_pair.neutral.text_embeds,
|
||||
# negative_prompt_embeds=prompt_pair.both_targets.text_embeds,
|
||||
# pooled_prompt_embeds=prompt_pair.neutral.pooled_embeds,
|
||||
# negative_pooled_prompt_embeds=prompt_pair.both_targets.pooled_embeds,
|
||||
# output_type="latent",
|
||||
# num_images_per_prompt=self.train_config.batch_size,
|
||||
# guidance_scale=3,
|
||||
# ).images.to(self.device_torch, dtype=dtype)
|
||||
|
||||
noise_scheduler.set_timesteps(1000)
|
||||
|
||||
current_timestep = noise_scheduler.timesteps[
|
||||
int(timesteps_to * 1000 / self.train_config.max_denoising_steps)
|
||||
]
|
||||
denoised_latents = noise
|
||||
|
||||
positive_latents = get_noise_pred(
|
||||
positive, negative, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
neutral_latents = get_noise_pred(
|
||||
positive, neutral, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
unconditional_latents = get_noise_pred(
|
||||
positive, positive, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
anchor_loss = None
|
||||
if len(self.anchor_pairs) > 0:
|
||||
# get a random anchor pair
|
||||
anchor: EncodedAnchor = self.anchor_pairs[
|
||||
torch.randint(0, len(self.anchor_pairs), (1,)).item()
|
||||
]
|
||||
with torch.no_grad():
|
||||
anchor_target_noise = get_noise_pred(
|
||||
anchor.prompt, anchor.neg_prompt, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
with self.network:
|
||||
# anchor whatever weight prompt pair is using
|
||||
pos_nem_mult = 1.0 if prompt_pair.multiplier > 0 else -1.0
|
||||
self.network.multiplier = anchor.multiplier * pos_nem_mult
|
||||
|
||||
anchor_pred_noise = get_noise_pred(
|
||||
anchor.prompt, anchor.neg_prompt, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
self.network.multiplier = prompt_pair.multiplier
|
||||
|
||||
with self.network:
|
||||
self.network.multiplier = prompt_pair.multiplier
|
||||
target_latents = get_noise_pred(
|
||||
positive, target_class, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
# if self.logging_config.verbose:
|
||||
# self.print("target_latents:", target_latents[0, 0, :5, :5])
|
||||
|
||||
positive_latents.requires_grad = False
|
||||
neutral_latents.requires_grad = False
|
||||
unconditional_latents.requires_grad = False
|
||||
if len(self.anchor_pairs) > 0:
|
||||
anchor_target_noise.requires_grad = False
|
||||
anchor_loss = loss_function(
|
||||
anchor_target_noise,
|
||||
anchor_pred_noise,
|
||||
# neutral prediction
|
||||
neutral_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
latents=denoised_latents,
|
||||
prompt_embeds=prompt_pair.neutral.text_embeds,
|
||||
negative_prompt_embeds=prompt_pair.empty_prompt.text_embeds,
|
||||
pooled_prompt_embeds=prompt_pair.neutral.pooled_embeds,
|
||||
negative_pooled_prompt_embeds=prompt_pair.both_targets.pooled_embeds,
|
||||
timestep=current_timestep,
|
||||
guidance_scale=1,
|
||||
num_images_per_prompt=self.train_config.batch_size,
|
||||
num_inference_steps=1000,
|
||||
)
|
||||
erase = prompt_pair.action == ACTION_TYPES_SLIDER.ERASE_NEGATIVE
|
||||
guidance_scale = 1.0
|
||||
|
||||
offset = guidance_scale * (positive_latents - unconditional_latents)
|
||||
# with self.network:
|
||||
# assert self.network.is_active
|
||||
# self.network.multiplier = 1.0
|
||||
#
|
||||
# positive_pos_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
# latents=denoised_latents,
|
||||
# prompt_embeds=prompt_pair.positive_target_with_neutral.text_embeds,
|
||||
# negative_prompt_embeds=prompt_pair.negative_target.text_embeds,
|
||||
# pooled_prompt_embeds=prompt_pair.positive_target_with_neutral.pooled_embeds,
|
||||
# negative_pooled_prompt_embeds=prompt_pair.negative_target.pooled_embeds,
|
||||
# timestep=current_timestep,
|
||||
# guidance_scale=1,
|
||||
# num_images_per_prompt=self.train_config.batch_size,
|
||||
# num_inference_steps=1000
|
||||
# )
|
||||
#
|
||||
# self.network.multiplier = -1.0
|
||||
#
|
||||
# negative_neg_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
# latents=denoised_latents,
|
||||
# prompt_embeds=prompt_pair.negative_target_with_neutral.text_embeds,
|
||||
# negative_prompt_embeds=prompt_pair.positive_target.text_embeds,
|
||||
# pooled_prompt_embeds=prompt_pair.negative_target_with_neutral.pooled_embeds,
|
||||
# negative_pooled_prompt_embeds=prompt_pair.positive_target.pooled_embeds,
|
||||
# timestep=current_timestep,
|
||||
# guidance_scale=1,
|
||||
# num_images_per_prompt=self.train_config.batch_size,
|
||||
# num_inference_steps=1000
|
||||
# )
|
||||
|
||||
offset_neutral = neutral_latents
|
||||
if erase:
|
||||
offset_neutral -= offset
|
||||
else:
|
||||
# enhance
|
||||
offset_neutral += offset
|
||||
# start grads
|
||||
self.optimizer.zero_grad()
|
||||
|
||||
loss = loss_function(
|
||||
target_latents,
|
||||
offset_neutral,
|
||||
) * weight
|
||||
multiplier = 5.0
|
||||
|
||||
loss_slide = loss.item()
|
||||
# predict postiitive
|
||||
with self.network:
|
||||
assert self.network.is_active
|
||||
self.network.multiplier = multiplier * 1.0
|
||||
|
||||
if anchor_loss is not None:
|
||||
loss += anchor_loss
|
||||
# positive_pos_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
# latents=denoised_latents,
|
||||
# prompt_embeds=prompt_pair.positive_target_with_neutral.text_embeds,
|
||||
# negative_prompt_embeds=prompt_pair.negative_target.text_embeds,
|
||||
# pooled_prompt_embeds=prompt_pair.positive_target_with_neutral.pooled_embeds,
|
||||
# negative_pooled_prompt_embeds=prompt_pair.negative_target.pooled_embeds,
|
||||
# timestep=current_timestep,
|
||||
# guidance_scale=1,
|
||||
# num_images_per_prompt=self.train_config.batch_size,
|
||||
# num_inference_steps=self.train_config.max_denoising_steps,
|
||||
# )
|
||||
|
||||
negative_pos_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
latents=denoised_latents,
|
||||
prompt_embeds=prompt_pair.negative_target_with_neutral.text_embeds,
|
||||
negative_prompt_embeds=prompt_pair.positive_target.text_embeds,
|
||||
pooled_prompt_embeds=prompt_pair.negative_target_with_neutral.pooled_embeds,
|
||||
negative_pooled_prompt_embeds=prompt_pair.positive_target.pooled_embeds,
|
||||
timestep=current_timestep,
|
||||
guidance_scale=1,
|
||||
num_images_per_prompt=self.train_config.batch_size,
|
||||
num_inference_steps=1000,
|
||||
)
|
||||
|
||||
self.network.multiplier = multiplier * -1.0
|
||||
|
||||
positive_neg_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
latents=denoised_latents,
|
||||
prompt_embeds=prompt_pair.positive_target_with_neutral.text_embeds,
|
||||
negative_prompt_embeds=prompt_pair.negative_target.text_embeds,
|
||||
pooled_prompt_embeds=prompt_pair.positive_target_with_neutral.pooled_embeds,
|
||||
negative_pooled_prompt_embeds=prompt_pair.negative_target.pooled_embeds,
|
||||
timestep=current_timestep,
|
||||
guidance_scale=1,
|
||||
num_images_per_prompt=self.train_config.batch_size,
|
||||
num_inference_steps=1000,
|
||||
)
|
||||
|
||||
# negative_neg_noise_prediction = self.sd.pipeline.predict_noise(
|
||||
# latents=denoised_latents,
|
||||
# prompt_embeds=prompt_pair.negative_target_with_neutral.text_embeds,
|
||||
# negative_prompt_embeds=prompt_pair.positive_target.text_embeds,
|
||||
# pooled_prompt_embeds=prompt_pair.negative_target_with_neutral.pooled_embeds,
|
||||
# negative_pooled_prompt_embeds=prompt_pair.positive_target.pooled_embeds,
|
||||
# timestep=current_timestep,
|
||||
# guidance_scale=1,
|
||||
# num_images_per_prompt=self.train_config.batch_size,
|
||||
# num_inference_steps=self.train_config.max_denoising_steps,
|
||||
# )
|
||||
|
||||
self.network.multiplier = 1.0
|
||||
|
||||
neutral_noise_prediction.requires_grad = False
|
||||
# positive_pos_noise_prediction.requires_grad = False
|
||||
# negative_neg_noise_prediction.requires_grad = False
|
||||
|
||||
# calculate loss
|
||||
loss_shrink_pos_neg = loss_function(
|
||||
negative_pos_noise_prediction,
|
||||
neutral_noise_prediction,
|
||||
)
|
||||
|
||||
loss_shrink_neg_pos = loss_function(
|
||||
positive_neg_noise_prediction,
|
||||
negative_pos_noise_prediction,
|
||||
)
|
||||
|
||||
loss = loss_shrink_pos_neg + loss_shrink_neg_pos
|
||||
|
||||
loss_float = loss.item()
|
||||
|
||||
@@ -384,12 +412,14 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
lr_scheduler.step()
|
||||
|
||||
del (
|
||||
positive_latents,
|
||||
neutral_latents,
|
||||
unconditional_latents,
|
||||
target_latents,
|
||||
denoised_latents,
|
||||
positive_neg_noise_prediction,
|
||||
negative_pos_noise_prediction,
|
||||
neutral_noise_prediction,
|
||||
latents,
|
||||
)
|
||||
# move back to cpu
|
||||
prompt_pair.to("cpu")
|
||||
flush()
|
||||
|
||||
# reset network
|
||||
@@ -398,9 +428,6 @@ class TrainSliderProcess(BaseSDTrainProcess):
|
||||
loss_dict = OrderedDict(
|
||||
{'loss': loss_float},
|
||||
)
|
||||
if anchor_loss is not None:
|
||||
loss_dict['sl_l'] = loss_slide
|
||||
loss_dict['an_l'] = anchor_loss.item()
|
||||
|
||||
return loss_dict
|
||||
# end hook_train_loop
|
||||
|
||||
406
jobs/process/TrainSliderProcessOld.py
Normal file
406
jobs/process/TrainSliderProcessOld.py
Normal file
@@ -0,0 +1,406 @@
|
||||
# ref:
|
||||
# - https://github.com/p1atdev/LECO/blob/main/train_lora.py
|
||||
import time
|
||||
from collections import OrderedDict
|
||||
import os
|
||||
from typing import Optional
|
||||
|
||||
from toolkit.config_modules import SliderConfig
|
||||
from toolkit.paths import REPOS_ROOT
|
||||
import sys
|
||||
|
||||
from toolkit.stable_diffusion_model import PromptEmbeds
|
||||
|
||||
sys.path.append(REPOS_ROOT)
|
||||
sys.path.append(os.path.join(REPOS_ROOT, 'leco'))
|
||||
from toolkit.train_tools import get_torch_dtype, apply_noise_offset
|
||||
import gc
|
||||
from toolkit import train_tools
|
||||
|
||||
import torch
|
||||
from leco import train_util, model_util
|
||||
from .BaseSDTrainProcess import BaseSDTrainProcess, StableDiffusion
|
||||
|
||||
|
||||
class ACTION_TYPES_SLIDER:
|
||||
ERASE_NEGATIVE = 0
|
||||
ENHANCE_NEGATIVE = 1
|
||||
|
||||
|
||||
def flush():
|
||||
torch.cuda.empty_cache()
|
||||
gc.collect()
|
||||
|
||||
|
||||
class EncodedPromptPair:
|
||||
def __init__(
|
||||
self,
|
||||
target_class,
|
||||
positive,
|
||||
negative,
|
||||
neutral,
|
||||
width=512,
|
||||
height=512,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=1.0,
|
||||
weight=1.0
|
||||
):
|
||||
self.target_class = target_class
|
||||
self.positive = positive
|
||||
self.negative = negative
|
||||
self.neutral = neutral
|
||||
self.width = width
|
||||
self.height = height
|
||||
self.action: int = action
|
||||
self.multiplier = multiplier
|
||||
self.weight = weight
|
||||
|
||||
|
||||
class PromptEmbedsCache: # 使いまわしたいので
|
||||
prompts: dict[str, PromptEmbeds] = {}
|
||||
|
||||
def __setitem__(self, __name: str, __value: PromptEmbeds) -> None:
|
||||
self.prompts[__name] = __value
|
||||
|
||||
def __getitem__(self, __name: str) -> Optional[PromptEmbeds]:
|
||||
if __name in self.prompts:
|
||||
return self.prompts[__name]
|
||||
else:
|
||||
return None
|
||||
|
||||
|
||||
class EncodedAnchor:
|
||||
def __init__(
|
||||
self,
|
||||
prompt,
|
||||
neg_prompt,
|
||||
multiplier=1.0
|
||||
):
|
||||
self.prompt = prompt
|
||||
self.neg_prompt = neg_prompt
|
||||
self.multiplier = multiplier
|
||||
|
||||
|
||||
class TrainSliderProcessOld(BaseSDTrainProcess):
|
||||
def __init__(self, process_id: int, job, config: OrderedDict):
|
||||
super().__init__(process_id, job, config)
|
||||
self.step_num = 0
|
||||
self.start_step = 0
|
||||
self.device = self.get_conf('device', self.job.device)
|
||||
self.device_torch = torch.device(self.device)
|
||||
self.slider_config = SliderConfig(**self.get_conf('slider', {}))
|
||||
self.prompt_cache = PromptEmbedsCache()
|
||||
self.prompt_pairs: list[EncodedPromptPair] = []
|
||||
self.anchor_pairs: list[EncodedAnchor] = []
|
||||
|
||||
def before_model_load(self):
|
||||
pass
|
||||
|
||||
def hook_before_train_loop(self):
|
||||
cache = PromptEmbedsCache()
|
||||
prompt_pairs: list[EncodedPromptPair] = []
|
||||
|
||||
# get encoded latents for our prompts
|
||||
with torch.no_grad():
|
||||
neutral = ""
|
||||
for target in self.slider_config.targets:
|
||||
# build the cache
|
||||
for prompt in [
|
||||
target.target_class,
|
||||
target.positive,
|
||||
target.negative,
|
||||
neutral # empty neutral
|
||||
]:
|
||||
if cache[prompt] is None:
|
||||
cache[prompt] = self.sd.encode_prompt(prompt)
|
||||
for resolution in self.slider_config.resolutions:
|
||||
width, height = resolution
|
||||
only_erase = len(target.positive.strip()) == 0
|
||||
only_enhance = len(target.negative.strip()) == 0
|
||||
|
||||
both = not only_erase and not only_enhance
|
||||
|
||||
if only_erase and only_enhance:
|
||||
raise ValueError("target must have at least one of positive or negative or both")
|
||||
# for slider we need to have an enhancer, an eraser, and then
|
||||
# an inverse with negative weights to balance the network
|
||||
# if we don't do this, we will get different contrast and focus.
|
||||
# we only perform actions of enhancing and erasing on the negative
|
||||
# todo work on way to do all of this in one shot
|
||||
|
||||
if both or only_erase:
|
||||
prompt_pairs += [
|
||||
# erase standard
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.positive],
|
||||
negative=cache[target.negative],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=target.multiplier,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
if both or only_enhance:
|
||||
prompt_pairs += [
|
||||
# enhance standard, swap pos neg
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.negative],
|
||||
negative=cache[target.positive],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ENHANCE_NEGATIVE,
|
||||
multiplier=target.multiplier,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
if both:
|
||||
prompt_pairs += [
|
||||
# erase inverted
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.negative],
|
||||
negative=cache[target.positive],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ERASE_NEGATIVE,
|
||||
multiplier=target.multiplier * -1.0,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
prompt_pairs += [
|
||||
# enhance inverted
|
||||
EncodedPromptPair(
|
||||
target_class=cache[target.target_class],
|
||||
positive=cache[target.positive],
|
||||
negative=cache[target.negative],
|
||||
neutral=cache[neutral],
|
||||
width=width,
|
||||
height=height,
|
||||
action=ACTION_TYPES_SLIDER.ENHANCE_NEGATIVE,
|
||||
multiplier=target.multiplier * -1.0,
|
||||
weight=target.weight
|
||||
),
|
||||
]
|
||||
|
||||
# setup anchors
|
||||
anchor_pairs = []
|
||||
for anchor in self.slider_config.anchors:
|
||||
# build the cache
|
||||
for prompt in [
|
||||
anchor.prompt,
|
||||
anchor.neg_prompt # empty neutral
|
||||
]:
|
||||
if cache[prompt] == None:
|
||||
cache[prompt] = self.sd.encode_prompt(prompt)
|
||||
|
||||
anchor_pairs += [
|
||||
EncodedAnchor(
|
||||
prompt=cache[anchor.prompt],
|
||||
neg_prompt=cache[anchor.neg_prompt],
|
||||
multiplier=anchor.multiplier
|
||||
)
|
||||
]
|
||||
|
||||
# move to cpu to save vram
|
||||
# We don't need text encoder anymore, but keep it on cpu for sampling
|
||||
# if text encoder is list
|
||||
if isinstance(self.sd.text_encoder, list):
|
||||
for encoder in self.sd.text_encoder:
|
||||
encoder.to("cpu")
|
||||
else:
|
||||
self.sd.text_encoder.to("cpu")
|
||||
self.prompt_cache = cache
|
||||
self.prompt_pairs = prompt_pairs
|
||||
self.anchor_pairs = anchor_pairs
|
||||
flush()
|
||||
# end hook_before_train_loop
|
||||
|
||||
def hook_train_loop(self):
|
||||
dtype = get_torch_dtype(self.train_config.dtype)
|
||||
|
||||
# get a random pair
|
||||
prompt_pair: EncodedPromptPair = self.prompt_pairs[
|
||||
torch.randint(0, len(self.prompt_pairs), (1,)).item()
|
||||
]
|
||||
|
||||
height = prompt_pair.height
|
||||
width = prompt_pair.width
|
||||
target_class = prompt_pair.target_class
|
||||
neutral = prompt_pair.neutral
|
||||
negative = prompt_pair.negative
|
||||
positive = prompt_pair.positive
|
||||
weight = prompt_pair.weight
|
||||
multiplier = prompt_pair.multiplier
|
||||
|
||||
unet = self.sd.unet
|
||||
noise_scheduler = self.sd.noise_scheduler
|
||||
optimizer = self.optimizer
|
||||
lr_scheduler = self.lr_scheduler
|
||||
loss_function = torch.nn.MSELoss()
|
||||
|
||||
def get_noise_pred(p, n, gs, cts, dn):
|
||||
return self.predict_noise(
|
||||
latents=dn,
|
||||
text_embeddings=train_tools.concat_prompt_embeddings(
|
||||
p, # unconditional
|
||||
n, # positive
|
||||
self.train_config.batch_size,
|
||||
),
|
||||
timestep=cts,
|
||||
guidance_scale=gs,
|
||||
)
|
||||
|
||||
# set network multiplier
|
||||
self.network.multiplier = multiplier
|
||||
|
||||
with torch.no_grad():
|
||||
self.sd.noise_scheduler.set_timesteps(
|
||||
self.train_config.max_denoising_steps, device=self.device_torch
|
||||
)
|
||||
|
||||
self.optimizer.zero_grad()
|
||||
|
||||
# ger a random number of steps
|
||||
timesteps_to = torch.randint(
|
||||
1, self.train_config.max_denoising_steps, (1,)
|
||||
).item()
|
||||
|
||||
# get noise
|
||||
noise = self.get_latent_noise(
|
||||
pixel_height=height,
|
||||
pixel_width=width,
|
||||
).to(self.device_torch, dtype=dtype)
|
||||
|
||||
# get latents
|
||||
latents = noise * self.sd.noise_scheduler.init_noise_sigma
|
||||
latents = latents.to(self.device_torch, dtype=dtype)
|
||||
|
||||
with self.network:
|
||||
assert self.network.is_active
|
||||
self.network.multiplier = multiplier
|
||||
denoised_latents = self.diffuse_some_steps(
|
||||
latents, # pass simple noise latents
|
||||
train_tools.concat_prompt_embeddings(
|
||||
positive, # unconditional
|
||||
target_class, # target
|
||||
self.train_config.batch_size,
|
||||
),
|
||||
start_timesteps=0,
|
||||
total_timesteps=timesteps_to,
|
||||
guidance_scale=3,
|
||||
)
|
||||
|
||||
noise_scheduler.set_timesteps(1000)
|
||||
|
||||
current_timestep = noise_scheduler.timesteps[
|
||||
int(timesteps_to * 1000 / self.train_config.max_denoising_steps)
|
||||
]
|
||||
|
||||
positive_latents = get_noise_pred(
|
||||
positive, negative, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
neutral_latents = get_noise_pred(
|
||||
positive, neutral, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
unconditional_latents = get_noise_pred(
|
||||
positive, positive, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
anchor_loss = None
|
||||
if len(self.anchor_pairs) > 0:
|
||||
# get a random anchor pair
|
||||
anchor: EncodedAnchor = self.anchor_pairs[
|
||||
torch.randint(0, len(self.anchor_pairs), (1,)).item()
|
||||
]
|
||||
with torch.no_grad():
|
||||
anchor_target_noise = get_noise_pred(
|
||||
anchor.prompt, anchor.neg_prompt, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
with self.network:
|
||||
# anchor whatever weight prompt pair is using
|
||||
pos_nem_mult = 1.0 if prompt_pair.multiplier > 0 else -1.0
|
||||
self.network.multiplier = anchor.multiplier * pos_nem_mult
|
||||
|
||||
anchor_pred_noise = get_noise_pred(
|
||||
anchor.prompt, anchor.neg_prompt, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
self.network.multiplier = prompt_pair.multiplier
|
||||
|
||||
with self.network:
|
||||
self.network.multiplier = prompt_pair.multiplier
|
||||
target_latents = get_noise_pred(
|
||||
positive, target_class, 1, current_timestep, denoised_latents
|
||||
).to("cpu", dtype=torch.float32)
|
||||
|
||||
# if self.logging_config.verbose:
|
||||
# self.print("target_latents:", target_latents[0, 0, :5, :5])
|
||||
|
||||
positive_latents.requires_grad = False
|
||||
neutral_latents.requires_grad = False
|
||||
unconditional_latents.requires_grad = False
|
||||
if len(self.anchor_pairs) > 0:
|
||||
anchor_target_noise.requires_grad = False
|
||||
anchor_loss = loss_function(
|
||||
anchor_target_noise,
|
||||
anchor_pred_noise,
|
||||
)
|
||||
erase = prompt_pair.action == ACTION_TYPES_SLIDER.ERASE_NEGATIVE
|
||||
guidance_scale = 1.0
|
||||
|
||||
offset = guidance_scale * (positive_latents - unconditional_latents)
|
||||
|
||||
offset_neutral = neutral_latents
|
||||
if erase:
|
||||
offset_neutral -= offset
|
||||
else:
|
||||
# enhance
|
||||
offset_neutral += offset
|
||||
|
||||
loss = loss_function(
|
||||
target_latents,
|
||||
offset_neutral,
|
||||
) * weight
|
||||
|
||||
loss_slide = loss.item()
|
||||
|
||||
if anchor_loss is not None:
|
||||
loss += anchor_loss
|
||||
|
||||
loss_float = loss.item()
|
||||
|
||||
loss = loss.to(self.device_torch)
|
||||
|
||||
loss.backward()
|
||||
optimizer.step()
|
||||
lr_scheduler.step()
|
||||
|
||||
del (
|
||||
positive_latents,
|
||||
neutral_latents,
|
||||
unconditional_latents,
|
||||
target_latents,
|
||||
latents,
|
||||
)
|
||||
flush()
|
||||
|
||||
# reset network
|
||||
self.network.multiplier = 1.0
|
||||
|
||||
loss_dict = OrderedDict(
|
||||
{'loss': loss_float},
|
||||
)
|
||||
if anchor_loss is not None:
|
||||
loss_dict['sl_l'] = loss_slide
|
||||
loss_dict['an_l'] = anchor_loss.item()
|
||||
|
||||
return loss_dict
|
||||
# end hook_train_loop
|
||||
@@ -6,5 +6,6 @@ from .BaseTrainProcess import BaseTrainProcess
|
||||
from .TrainVAEProcess import TrainVAEProcess
|
||||
from .BaseMergeProcess import BaseMergeProcess
|
||||
from .TrainSliderProcess import TrainSliderProcess
|
||||
from .TrainSliderProcessOld import TrainSliderProcessOld
|
||||
from .TrainLoRAHack import TrainLoRAHack
|
||||
from .TrainSDRescaleProcess import TrainSDRescaleProcess
|
||||
@@ -99,3 +99,5 @@ class SliderConfig:
|
||||
anchors = [SliderConfigAnchors(**anchor) for anchor in anchors]
|
||||
self.anchors: List[SliderConfigAnchors] = anchors
|
||||
self.resolutions: List[List[int]] = kwargs.get('resolutions', [[512, 512]])
|
||||
self.prompt_file: str = kwargs.get('prompt_file', '')
|
||||
self.prompt_tensors: str = kwargs.get('prompt_tensors', '')
|
||||
|
||||
@@ -30,10 +30,10 @@ class PromptEmbeds:
|
||||
self.text_embeds = args
|
||||
self.pooled_embeds = None
|
||||
|
||||
def to(self, **kwargs):
|
||||
self.text_embeds = self.text_embeds.to(**kwargs)
|
||||
def to(self, *args, **kwargs):
|
||||
self.text_embeds = self.text_embeds.to(*args, **kwargs)
|
||||
if self.pooled_embeds is not None:
|
||||
self.pooled_embeds = self.pooled_embeds.to(**kwargs)
|
||||
self.pooled_embeds = self.pooled_embeds.to(*args, **kwargs)
|
||||
return self
|
||||
|
||||
|
||||
|
||||
Reference in New Issue
Block a user