65 lines
1.9 KiB
YAML
65 lines
1.9 KiB
YAML
model:
|
|
base_learning_rate: 1.0e-04
|
|
target: easyanimate.vae.ldm.models.omnigen_casual3dcnn.AutoencoderKLMagvit_fromOmnigen
|
|
params:
|
|
slice_compression_vae: true
|
|
mini_batch_encoder: 8
|
|
mini_batch_decoder: 2
|
|
monitor: train/rec_loss
|
|
ckpt_path: models/Diffusion_Transformer/EasyAnimateV2-XL-2-512x512/vae/diffusion_pytorch_model.safetensors
|
|
down_block_types: ("SpatialDownBlock3D", "SpatialTemporalDownBlock3D", "SpatialTemporalDownBlock3D",
|
|
"SpatialTemporalDownBlock3D",)
|
|
up_block_types: ("SpatialUpBlock3D", "SpatialTemporalUpBlock3D", "SpatialTemporalUpBlock3D",
|
|
"SpatialTemporalUpBlock3D",)
|
|
lossconfig:
|
|
target: easyanimate.vae.ldm.modules.losses.LPIPSWithDiscriminator
|
|
params:
|
|
disc_start: 50001
|
|
kl_weight: 1.0e-06
|
|
disc_weight: 0.5
|
|
l2_loss_weight: 0.0
|
|
l1_loss_weight: 1.0
|
|
perceptual_weight: 1.0
|
|
|
|
data:
|
|
target: train_vae.DataModuleFromConfig
|
|
|
|
params:
|
|
batch_size: 1
|
|
wrap: true
|
|
num_workers: 8
|
|
train:
|
|
target: easyanimate.vae.ldm.data.dataset_image_video.CustomSRTrain
|
|
params:
|
|
data_json_path: pretrain.json
|
|
data_root: /your_data_root # This is used in relative path
|
|
size: 256
|
|
degradation: pil_nearest
|
|
video_size: 256
|
|
video_len: 25
|
|
slice_interval: 1
|
|
validation:
|
|
target: easyanimate.vae.ldm.data.dataset_image_video.CustomSRValidation
|
|
params:
|
|
data_json_path: pretrain.json
|
|
data_root: /your_data_root # This is used in relative path
|
|
size: 256
|
|
degradation: pil_nearest
|
|
video_size: 256
|
|
video_len: 25
|
|
slice_interval: 1
|
|
|
|
lightning:
|
|
callbacks:
|
|
image_logger:
|
|
target: train_vae.ImageLogger
|
|
params:
|
|
batch_frequency: 5000
|
|
max_images: 8
|
|
increase_log_steps: True
|
|
|
|
trainer:
|
|
benchmark: True
|
|
accumulate_grad_batches: 1
|
|
gpus: "0"
|
|
num_nodes: 1 |