stable-diffusion-finetune/configs/stable-diffusion/txt2img-multinode-clip-encoder-f16-768.yaml

131 lines
3.5 KiB
YAML
Raw Normal View History

2022-06-16 15:12:54 +00:00
model:
base_learning_rate: 1.0e-04
target: ldm.models.diffusion.ddpm.LatentDiffusion
params:
linear_start: 0.001
linear_end: 0.015
num_timesteps_cond: 1
log_every_t: 200
timesteps: 1000
first_stage_key: "jpg"
cond_stage_key: "txt"
image_size: 48
channels: 16
cond_stage_trainable: false # Note: different from the one we trained before
conditioning_key: crossattn
monitor: val/loss_simple_ema
scale_factor: 0.22765929 # magic number
2022-06-17 23:05:03 +00:00
ckpt_path: "/home/mchorse/stable-diffusion-ckpts/256f16-2022-06-15-216k-pruned.ckpt"
2022-06-16 15:12:54 +00:00
scheduler_config: # 10000 warmup steps
target: ldm.lr_scheduler.LambdaLinearScheduler
params:
warm_up_steps: [ 10000 ]
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: 48 # not really needed
in_channels: 16
out_channels: 16
model_channels: 320 # TODO: scale model here
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: 16
monitor: val/rec_loss
ddconfig:
double_z: True
z_channels: 16
resolution: 256
in_channels: 3
out_ch: 3
ch: 128
ch_mult: [ 1,1,2,2,4 ] # num_down = len(ch_mult)-1
num_res_blocks: 2
attn_resolutions: [ 16 ]
dropout: 0.0
lossconfig:
target: torch.nn.Identity
cond_stage_config:
target: ldm.modules.encoders.modules.FrozenCLIPEmbedder
data:
target: ldm.data.laion.WebDataModuleFromConfig
params:
tar_base: "pipe:aws s3 cp s3://s-datasets/laion5b/laion2B-data/"
2022-06-18 07:30:59 +00:00
batch_size: 6
2022-06-16 15:12:54 +00:00
num_workers: 4
multinode: True
min_size: 384 # TODO: experiment. Note: for 2B, images are stored at max 384 resolution
train:
shards: '{000000..231317}.tar -'
shuffle: 10000
image_key: jpg
image_transforms:
- target: torchvision.transforms.Resize
params:
size: 768
interpolation: 3
- target: torchvision.transforms.RandomCrop
params:
size: 768
# NOTE use enough shards to avoid empty validation loops in workers
validation:
shards: '{231318..231349}.tar -'
shuffle: 0
image_key: jpg
image_transforms:
- target: torchvision.transforms.Resize
params:
size: 768
interpolation: 3
- target: torchvision.transforms.CenterCrop
params:
size: 768
lightning:
2022-06-17 23:20:03 +00:00
find_unused_parameters: False
2022-06-16 15:12:54 +00:00
callbacks:
image_logger:
target: main.ImageLogger
params:
batch_frequency: 5000
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
unconditional_guidance_scale: 3.0
unconditional_guidance_label: [""]
trainer:
benchmark: True
val_check_interval: 5000000
num_sanity_val_steps: 0
accumulate_grad_batches: 2