RuntimeError: The size of tensor a (8) must match the size of tensor b (6) at non-singleton dimension 3

[ContrastingRealities]

I'm having issues training a LoHa, this error always occurs right before training begins

steps: 0%| | 0/7776 [00:00<?, ?it/s] epoch 1/18 Traceback (most recent call last): File "C:\Users\Alan\kohya_ss\train_network.py", line 1033, in trainer.train(args) File "C:\Users\Alan\kohya_ss\train_network.py", line 810, in train noise_pred = self.call_unet( File "C:\Users\Alan\kohya_ss\train_network.py", line 123, in call_unet noise_pred = unet(noisy_latents, timesteps, text_conds).sample File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, kwargs) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl return forward_call(*args, *kwargs) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\accelerate\utils\operations.py", line 680, in forward return model_forward(args, kwargs) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\accelerate\utils\operations.py", line 668, in call return convert_to_fp32(self.model_forward(*args, kwargs)) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\amp\autocast_mode.py", line 16, in decorate_autocast return func(*args, *kwargs) File "C:\Users\Alan\kohya_ss\library\original_unet.py", line 1585, in forward sample, res_samples = downsample_block( File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl return self._call_impl(args, kwargs) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl return forward_call(*args, kwargs) File "C:\Users\Alan\kohya_ss\library\original_unet.py", line 1019, in forward hidden_states = resnet(hidden_states, temb) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1518, in _wrapped_call_impl return self._call_impl(*args, *kwargs) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\torch\nn\modules\module.py", line 1527, in _call_impl return forward_call(args, kwargs) File "C:\Users\Alan\kohya_ss\library\original_unet.py", line 477, in forward output_tensor = input_tensor + hidden_states RuntimeError: The size of tensor a (8) must match the size of tensor b (6) at non-singleton dimension 3 steps: 0%| | 0/7776 [00:01<?, ?it/s] Traceback (most recent call last): File "C:\Users\Alan\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main return _run_code(code, main_globals, None, File "C:\Users\Alan\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code exec(code, run_globals) File "C:\Users\Alan\kohya_ss\venv\Scripts\accelerate.exe__main__.py", line 7, in File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 47, in main args.func(args) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 1017, in launch_command simple_launcher(args) File "C:\Users\Alan\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 637, in simple_launcher raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd) subprocess.CalledProcessError: Command '['C:\Users\Alan\kohya_ss\venv\Scripts\python.exe', './train_network.py', '--enable_bucket', '--min_bucket_reso=256', '--max_bucket_reso=2048', '--pretrained_model_name_or_path=C:/Users/Alan/Downloads/anyloraCheckpoint_bakedvaeBlessedFp16 (1).safetensors', '--train_data_dir=C:/Users/Alan/Downloads/ShadowAndFaceFRevised', '--resolution=512,512', '--output_dir=C:/Users/Alan/Documents/ShadowAndFaceFRevised', '--network_alpha=1', '--save_model_as=safetensors', '--network_module=lycoris.kohya', '--network_args', 'preset=full', 'conv_dim=1', 'conv_alpha=1', 'rank_dropout=0.1', 'module_dropout=0.1', 'use_tucker=False', 'use_scalar=False', 'rank_dropout_scale=False', 'algo=loha', 'train_norm=False', '--network_dropout=0', '--text_encoder_lr=5e-05', '--unet_lr=0.0001', '--network_dim=8', '--output_name=ShadowAndFaceFRevised', '--lr_scheduler_num_cycles=18', '--scale_weight_norms=1', '--learning_rate=0.0001', '--lr_scheduler=cosine', '--lr_warmup_steps=778', '--train_batch_size=2', '--max_train_steps=7776', '--save_every_n_epochs=1', '--mixed_precision=bf16', '--save_precision=bf16', '--caption_extension=txt', '--cache_latents', '--optimizer_type=AdamW8bit', '--max_grad_norm=1', '--max_data_loader_n_workers=0', '--clip_skip=2', '--bucket_reso_steps=64', '--shuffle_caption', '--xformers', '--bucket_no_upscale', '--noise_offset=0.0', '--sample_sampler=euler_a', '--sample_prompts=C:/Users/Alan/Documents/ShadowAndFaceFRevised\sample\prompt.txt', '--sample_every_n_epochs=1']' returned non-zero exit status 1.

[bmaltais]

I suggest you try to reach the author of LyCORIS and raise this issue with him. Either an issue with LyCORIS or sd-scripts...

https://github.com/KohakuBlueleaf/LyCORIS/issues

[ContrastingRealities]

Thank you I will try that