wuziheng
commented
10 months ago Thank you for your use and efforts. Regarding this model, it seems that this model cannot be trained using 'kohya,' and it appears that some keys are not aligned. You can try using 'kohya' training to address this issue. At the same time, we are also upgrading the training code to support higher versions of diffusers, but it may not necessarily resolve your specific problem.
new_checkpoint["encoder.conv_in.weight"] = vae_state_dict["encoder.conv_in.weight"] KeyError: 'encoder.conv_in.weight'"
I have had great success using your extension to train faces for several models. But there is one that can't seem to be used for training. It can be found here https://civitai.com/models/142364?modelVersionId=157856. I tried converting it to fp16, but that didn't make a difference.
The error is as follows:
Traceback (most recent call last): File "C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya\train_lora.py", line 1478, in
main()
File "C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya\utils\gpu_info.py", line 195, in wrapper
result = func(*args, **kwargs)
File "C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya\train_lora.py", line 861, in main
text_encoder, vae, unet = load_models_from_stable_diffusion_checkpoint(False, args.pretrained_model_ckpt)
File "C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya\utils\model_utils.py", line 842, in load_models_from_stable_diffusion_checkpoint
converted_vae_checkpoint = convert_ldm_vae_checkpoint(state_dict, vae_config)
File "C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya\utils\model_utils.py", line 370, in convert_ldm_vae_checkpoint
new_checkpoint["encoder.conv_in.weight"] = vae_state_dict["encoder.conv_in.weight"]
KeyError: 'encoder.conv_in.weight'
Traceback (most recent call last):
File "runpy.py", line 196, in _run_module_as_main
File "runpy.py", line 86, in _run_code
File "C:\sd.webui\system\python\lib\site-packages\accelerate\commands\launch.py", line 989, in
main()
File "C:\sd.webui\system\python\lib\site-packages\accelerate\commands\launch.py", line 985, in main
launch_command(args)
File "C:\sd.webui\system\python\lib\site-packages\accelerate\commands\launch.py", line 979, in launch_command
simple_launcher(args)
File "C:\sd.webui\system\python\lib\site-packages\accelerate\commands\launch.py", line 628, in simple_launcher
raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['C:\sd.webui\system\python\python.exe', 'C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya/train_lora.py', '--pretrained_model_name_or_path=extensions\sd-webui-EasyPhoto\models\stable-diffusion-v1-5', '--pretrained_model_ckpt=models\Stable-diffusion\lxlmodel_v10_fp16.safetensors', '--train_data_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\processed_images', '--caption_column=text', '--resolution=512', '--random_flip', '--train_batch_size=1', '--gradient_accumulation_steps=4', '--dataloader_num_workers=0', '--max_train_steps=800', '--checkpointing_steps=100', '--learning_rate=0.0001', '--lr_scheduler=constant', '--lr_warmup_steps=0', '--train_text_encoder', '--seed=42', '--rank=128', '--network_alpha=64', '--validation_prompt=easyphoto_face, easyphoto, 1person', '--validation_steps=100', '--output_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\user_weights', '--logging_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\user_weights', '--enable_xformers_memory_efficient_attention', '--mixed_precision=fp16', '--template_dir=extensions\sd-webui-EasyPhoto\models\training_templates', '--template_mask', '--merge_best_lora_based_face_id', '--merge_best_lora_name=swah-lxlmodel', '--cache_log_file=C:\sd.webui\webui\outputs/easyphoto-tmp/train_kohya_log.txt', '--validation']' returned non-zero exit status 1.
Error executing the command: Command '['C:\sd.webui\system\python\python.exe', '-m', 'accelerate.commands.launch', '--mixed_precision=fp16', '--main_process_port=3456', 'C:\sd.webui\webui\extensions\sd-webui-EasyPhoto\scripts\train_kohya/train_lora.py', '--pretrained_model_name_or_path=extensions\sd-webui-EasyPhoto\models\stable-diffusion-v1-5', '--pretrained_model_ckpt=models\Stable-diffusion\lxlmodel_v10_fp16.safetensors', '--train_data_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\processed_images', '--caption_column=text', '--resolution=512', '--random_flip', '--train_batch_size=1', '--gradient_accumulation_steps=4', '--dataloader_num_workers=0', '--max_train_steps=800', '--checkpointing_steps=100', '--learning_rate=0.0001', '--lr_scheduler=constant', '--lr_warmup_steps=0', '--train_text_encoder', '--seed=42', '--rank=128', '--network_alpha=64', '--validation_prompt=easyphoto_face, easyphoto, 1person', '--validation_steps=100', '--output_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\user_weights', '--logging_dir=outputs\easyphoto-user-id-infos\swah-lxlmodel\user_weights', '--enable_xformers_memory_efficient_attention', '--mixed_precision=fp16', '--template_dir=extensions\sd-webui-EasyPhoto\models\training_templates', '--template_mask', '--merge_best_lora_based_face_id', '--merge_best_lora_name=swah-lxlmodel', '--cache_log_file=C:\sd.webui\webui\outputs/easyphoto-tmp/train_kohya_log.txt', '--validation']' returned non-zero exit status 1.
Using already loaded model lxlmodel_v10_fp16.safetensors [00a41efe34]: done in 1.1s (send model to device: 1.1s)