error in comfyui when i try to load lora

[nikolaiusa]

Error occurred when executing DiffusersPipelineLoader:

Incorrect path_or_model_id: 'I:\HunyuanDiT\ComfyUI\custom_nodes\comfyui-hydit....\models\loras\last.ckpt'. Please provide either the path to a local folder or the repo_id of a model on the Hub.

File "I:\HunyuanDiT\ComfyUI\execution.py", line 151, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\execution.py", line 81, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\execution.py", line 74, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\custom_nodes\comfyui-hydit\nodes.py", line 109, in create_pipeline gen = End2End(args_hunyuan[0], Path(os.path.join(HUNYUAN_PATH, pipeline_folder_name)), LOAR_PATH=LORA_PATH) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\custom_nodes\comfyui-hydit\hydit\inference_comfyui.py", line 241, in init self.model.load_adapter(LOAR_PATH) File "I:\ANACONDA3\envs\HUNYUANDIT\Lib\site-packages\transformers\integrations\peft.py", line 176, in load_adapter adapter_config_file = find_adapter_config_file( ^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\ANACONDA3\envs\HUNYUANDIT\Lib\site-packages\transformers\utils\peft_utils.py", line 88, in find_adapter_config_file adapter_cached_filename = cached_file( ^^^^^^^^^^^^ File "I:\ANACONDA3\envs\HUNYUANDIT\Lib\site-packages\transformers\utils\hub.py", line 463, in cached_file raise EnvironmentError(

[nikolaiusa]

I made lora in Kohyass but I can’t run it in comfyui

[nikolaiusa]

should I create an adapter_config.json file for my lora? how to do it?

[h3clikejava]

Do you solve it?

[nikolaiusa]

Do you solve it?

no

[xuhuaren]

the released new version could load lora from khoyo, today.

[nikolaiusa]

new error, can you tell me what it could be.

Error occurred when executing CLIPTextEncode:

'tuple' object has no attribute 'pop'

File "I:\HunyuanDiT\ComfyUI\execution.py", line 151, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\execution.py", line 81, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\execution.py", line 74, in map_node_over_list results.append(getattr(obj, func)(**slice_dict(input_data_all, i))) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "I:\HunyuanDiT\ComfyUI\nodes.py", line 59, in encode cond = output.pop("cond") ^^^^^^^^^^

[nikolaiusa]

I did this - the error disappeared git reset --hard 90389b3b8a69c08c3ed0bcc9d87a92246578a8e3

[nikolaiusa]

but in the console there are errors as if the lora is not working

a key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_output_dense.lora_up.weight

etc...

[xuhuaren]

have you replace the lora.py following readme

On Wed, Jul 17, 2024 at 12:21 AM nikolaiusa @.***> wrote:

but in the console there are errors as if the lora is not working

a key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_query.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_output_dense.lora_up.weight

etc...

— Reply to this email directly, view it on GitHub https://github.com/Tencent/HunyuanDiT/issues/146#issuecomment-2231339759, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACZHK7OS5TRLP3UDTTARWBDZMVCBFAVCNFSM6AAAAABKNGKSNOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZRGMZTSNZVHE . You are receiving this because you commented.Message ID: @.***>

[nikolaiusa]

have you replace the lora.py following readme

thx i replaced.

after this comman: python custom_nodes/comfyui-hydit/convert_hunyuan_to_comfyui_lora.py
--lora_path ${HunyuanDiT}/log_EXP/001-lora_porcelain_ema_rank64/checkpoints/0000100.pt/adapter_model.safetensors
--save_lora_path ${ComfyUI}/models/loras/adapter_model_convert.safetensors

this error in console:

a key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_down.weight

Without this command " convert_hunyuan_to_comfyui_lora.py " the lora works - no error in console.

but with strange resultst )

[xuhuaren]

the lora you trained is using khoyos， don’t need this convert script.

On Wed, Jul 17, 2024 at 5:34 PM nikolaiusa @.***> wrote:

have you replace the lora.py following readme

thx i replaced.

after this comman: python custom_nodes/comfyui-hydit/convert_hunyuan_to_comfyui_lora.py --lora_path ${HunyuanDiT}/log_EXP/001-lora_porcelain_ema_rank64/checkpoints/ 0000100.pt/adapter_model.safetensors --save_lora_path ${ComfyUI}/models/loras/adapter_model_convert.safetensors

this error in console:

a key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_attention_self_value.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_intermediate_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_8_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_down.weight lora key not loaded: lora_te1_encoder_layer_9_attention_output_dense.lora_up.weight lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.alpha lora key not loaded: lora_te1_encoder_layer_9_attention_self_key.lora_down.weight

Without this command " convert_hunyuan_to_comfyui_lora.py " the lora works

• no error in console.

but with strange resultst )

— Reply to this email directly, view it on GitHub https://github.com/Tencent/HunyuanDiT/issues/146#issuecomment-2232868103, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACZHK7NI4GJ4FH7Y3R5OLU3ZMY3CJAVCNFSM6AAAAABKNGKSNOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZSHA3DQMJQGM . You are receiving this because you commented.Message ID: @.***>