KeyError: 'model.diffusion_model.input_blocks.0.0.weight'
pokameng opened this issue · comments
hello
@justinpinkney
when i load the finetune model sd-clip-vit-l14-img-embed_ema_only.ckpt
it alert: KeyError: 'model.diffusion_model.input_blocks.0.0.weight'
so how to solve this problem?
Thanks!!!
my yaml is:
`model:
base_learning_rate: 1.0e-05
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: rgb_image
image_size: 64
channels: 4
cond_stage_trainable: False # Note: different from the one we trained before
# unet_trainable: attn
# unet_trainable: "attn"
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: [ 1000 ]
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 #ldm.models.autoencoder.VQModelInterface #ldm.models.autoencoder.AutoencoderKL
params:
ckpt_path: /home/dailongquan/110.014/wsm/stable-diffusion-main/models/first_stage_models/kl-f8/model.ckpt
embed_dim: 4
#n_embed: 8192
monitor: val/rec_loss
ddconfig:
double_z: True
z_channels: 4
resolution: 256
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.FrozenCLIPImageEmbedder
data:
target: main.DataModuleFromConfig
params:
batch_size: 2
num_workers: 8
num_val_workers: 0 # Avoid a weird val dataloader issue
train:
target: ldm.data.simple.FangHuaData
# image_key: image
params:
root_dir: /home/share/movie_dataset/fanghua/png
ext: jpg
image_transforms:
- target: torchvision.transforms.Resize
params:
size: 256
interpolation: 3
- target: torchvision.transforms.RandomCrop
params:
size: 256
validation:
target: ldm.data.simple.FangHuaData
# image_key: image
params:
root_dir: /home/share/movie_dataset/fanghua/png
ext: jpg
image_transforms:
- target: torchvision.transforms.Resize
params:
size: 256
interpolation: 3
- target: torchvision.transforms.RandomCrop
params:
size: 256
lightning:
find_unused_parameters: false
modelcheckpoint:
params:
every_n_train_steps: 500
callbacks:
image_logger:
target: main.ImageLogger
params:
batch_frequency: 100
max_images: 8
increase_log_steps: False
log_first_step: True
log_images_kwargs:
use_ema_scope: False
inpaint: False
plot_progressive_rows: False
plot_diffusion_rows: False
N: 8
unconditional_guidance_scale: 3.0
unconditional_guidance_label: [""]
trainer:
benchmark: True
# val_check_interval: 5000000 # really sorry
num_sanity_val_steps: 0
accumulate_grad_batches: 2
`