waifu-diffusion/configs/stable-diffusion/v1-4-finetune-test.yaml

118 lines
2.9 KiB
YAML

model:
base_learning_rate: 7.5e-06
target: ldm.models.diffusion.ddpm.LatentDiffusion
params:
linear_start: 0.00085
linear_end: 0.0120
num_timesteps_cond: 1
log_every_t: 200
timesteps: 1000
first_stage_key: image
cond_stage_key: caption
image_size: 64
channels: 4
cond_stage_trainable: false # Note: different from the one we trained before
conditioning_key: crossattn
monitor: val/loss_simple_ema
scale_factor: 0.18215
scheduler_config: # 10000 warmup steps
target: ldm.lr_scheduler.LambdaLinearScheduler
params:
warm_up_steps: [ 1 ] # NOTE for resuming. use 10000 if starting from scratch
cycle_lengths: [ 10000000000000 ] # incredibly large number to prevent corner cases
f_start: [ 1.e-6 ]
f_max: [ 1. ]
f_min: [ 1. ]
unet_config:
target: ldm.modules.diffusionmodules.openaimodel.UNetModel
params:
image_size: 32 # unused
in_channels: 4
out_channels: 4
model_channels: 320
attention_resolutions: [ 4, 2, 1 ]
num_res_blocks: 2
channel_mult: [ 1, 2, 4, 4 ]
num_heads: 8
use_spatial_transformer: True
transformer_depth: 1
context_dim: 768
use_checkpoint: True
legacy: False
first_stage_config:
target: ldm.models.autoencoder.AutoencoderKL
params:
embed_dim: 4
monitor: val/rec_loss
ddconfig:
double_z: true
z_channels: 4
resolution: 512
in_channels: 3
out_ch: 3
ch: 128
ch_mult:
- 1
- 2
- 4
- 4
num_res_blocks: 2
attn_resolutions: []
dropout: 0.0
lossconfig:
target: torch.nn.Identity
cond_stage_config:
target: ldm.modules.encoders.modules.FrozenCLIPEmbedder
params:
penultimate: true # use 2nd last layer - https://arxiv.org/pdf/2205.11487.pdf D.1
data:
target: main.DataModuleFromConfig
params:
batch_size: 4
num_workers: 4
wrap: false
train:
target: ldm.data.local.LocalBase
params:
size: 512
mode: "train"
ucg: 0.1 # unconditional guidance training
validation:
target: ldm.data.local.LocalBase
params:
size: 512
mode: "val"
val_split: 64
ucg: 0.1
lightning:
modelcheckpoint:
params:
every_n_train_steps: 500
callbacks:
image_logger:
target: main.ImageLogger
params:
batch_frequency: 500
max_images: 4
increase_log_steps: False
log_first_step: False
log_images_kwargs:
use_ema_scope: False
inpaint: False
plot_progressive_rows: False
plot_diffusion_rows: False
N: 4
ddim_steps: 50
trainer:
benchmark: True
val_check_interval: 5000000
num_sanity_val_steps: 0
accumulate_grad_batches: 1