search

shiimizu / ComfyUI_smZNodes

Custom nodes for ComfyUI such as CLIP Text Encode++
GNU General Public License v3.0
168 stars 12 forks source link

smZ CLIPTextEncode seem easy to get error when the prompt too long or get Other encoded characters or punctuation #45

Open xueqing0622 opened 6 months ago

xueqing0622 commented 6 months ago

smZ CLIPTextEncode seem easy to get error when the prompt too long or get Other encoded characters or punctuation image

Error occurred when executing smZ CLIPTextEncode:

tuple index out of range

File "F:\ComfyUI\ComfyUI\execution.py", line 154, in recursive_execute output_data, output_ui = get_output_data(obj, input_data_all) File "F:\ComfyUI\ComfyUI\execution.py", line 84, in get_output_data return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True) File "F:\ComfyUI\ComfyUI\execution.py", line 77, in map_node_over_list results.append(getattr(obj, func)(slice_dict(input_data_all, i))) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\nodes.py", line 87, in encode result = run(params) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\smZNodes.py", line 693, in run cond, pooled = clip_clone.encode_from_tokens(tokens, True) File "F:\ComfyUI\ComfyUI\comfy\sd.py", line 131, in encode_from_tokens cond, pooled = self.cond_stage_model.encode_token_weights(tokens) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\smZNodes.py", line 416, in encode_token_weights if multi: schedules = prompt_parser.get_multicond_learned_conditioning(model_hijack.cond_stage_model, texts, steps, None, opts.use_old_scheduling) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\modules\prompt_parser.py", line 270, in get_multicond_learned_conditioning learned_conditioning = get_learned_conditioning(model, prompt_flat_list, steps, hires_steps, use_old_scheduling) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\modules\prompt_parser.py", line 198, in get_learned_conditioning conds = model.forward(texts) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\modules\sd_hijack_clip.py", line 207, in forward z = self.process_tokens(tokens, multipliers) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\modules\sd_hijack_clip.py", line 237, in process_tokens z = self.encode_with_transformers(tokens) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\smZNodes.py", line 223, in encode_with_transformers return self.encode_with_transformerscomfy(tokens, return_pooled) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\smZNodes.py", line 124, in encode_with_transformerscomfy z, pooled = ClipTextEncoderCustom._forward(self.wrapped, tokens_orig) File "F:\ComfyUI\ComfyUI\custom_nodes\ComfyUI_smZNodes\smZNodes.py", line 113, in _forward z, pooled_output = self.forward(tokens) File "F:\ComfyUI\ComfyUI\comfy\sd1_clip.py", line 160, in forward tokens = self.set_up_textual_embeddings(tokens, backup_embeds) File "F:\ComfyUI\ComfyUI\comfy\sd1_clip.py", line 131, in set_up_textual_embeddings if y.shape[0] == current_embeds.weight.shape[1]:

xueqing0622 commented 6 months ago

And these are error prompts samplers: And I use in A1111 mode image

1: 1gir,black hair,Night,Standing in the meadow,Blue stars,looking at viewers, (extremely detailed 8k illustration), (extremely detailed and beautiful background:1.2), ultra detailed painting, professional illustrasion, Ultra-precise depiction, Ultra-detailed depiction, (beautiful and aesthetic:1.2), vivid,intricate, nice hands, perfect hands,

2: highly detailed black sumi-e painting of . in-depth study of perfection, created by a master. best quality, high resolution

shiimizu commented 6 months ago

It seems like an embedding was used? Do you have any embeddings that have the same name as one of the words you used?

xueqing0622 commented 6 months ago

shiimizu, you are right! after I remove all the embedding , It works, How do avoid it happen but I have to keep my embeddings?

shiimizu commented 6 months ago

Make sure your embeddings are renamed to words you don't often use.