Optimized the prompt word module of the plugin

[dkqjhoifh]

Krita is the best free ai plugin I've ever used, but its prompts are still only available in English, which is a pain in the way I keep going to the web page to translate prompts during the drawing process, so why not use Argos Translate CLIP Text Encode Node instead of clip prompt? That way, the whole world won't have to worry about prompts and translations.

[dkqjhoifh]

[dkqjhoifh]

I tried to manually modify the node of the clip prompt in the krita workflow, although it works in comfyui, but it doesn't seem to take effect in krita, krita still only recognizes English.

[dkqjhoifh]

Which krita workflow are you modifying and can you tell me the file path?

[Drakosha405]

Line 243 in comfy_workflow.py

dog from German

cat from Russian

[XylitolJ]

I tried to manually modify the node of the clip prompt in the krita workflow, although it works in comfyui, but it doesn't seem to take effect in krita, krita still only recognizes English.

I tried to insert this node into the workflow, I didn't find any problems, I changed it to German for verification, and everything is fine in the process logs. Line 243 in comfy_workflow.py return self.add("ArgosTranslateCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="ru")

Unfortunately mine doesn't work, could you please take a look ?
I've already install Alexpet Node:

then insert line like this:

but it appear an error when execute

My Krita version: 1.19.0

in client.log:

`
2024-07-02 15:00:16,701 INFO Found LoRA model hyper for SD 1.5: Hyper-SD15-8steps-CFG-lora.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model hyper for SD XL: Hyper-SDXL-8steps-CFG-lora.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model lcm for SD 1.5: lcm-lora-sdv1-5.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model lcm for SD XL: lcm-lora-sdxl.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model face for SD 1.5: ipadapter\ip-adapter-faceid-plus_sd15_lora.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model face for SD XL: ip-adapter-faceid-plusv2_sdxl_lora.safetensors
2024-07-02 15:00:16,701 INFO Found LoRA model lightning for SD XL: sdxl_lightning_8step_lora.safetensors
2024-07-02 15:00:17,733 INFO SD 1.5: supported
2024-07-02 15:00:17,733 INFO SD XL: supported
2024-07-02 15:00:17,733 INFO SD 3: supported
2024-07-02 15:00:33,584 ERROR Job 6b6215e3-2075-4468-80f5-08a6e4ec1d67 failed: 'ru'
[' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 151, in recursive_execute\n output_data, output_ui = get_output_data(obj, input_data_all)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 81, in get_output_data\n return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 74, in map_node_over_list\n results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_Custom_Nodes_AlekPet\ArgosTranslateNode\argos_translate_node.py", line 182, in argos_translate_text\n self.langs_support = ALL_CODES[from_translate]["targets"]\n ~~~~~~~~~^^^^^^^^^^^^^^^^\n']
2024-07-02 15:01:01,617 ERROR Job 7f85f10b-d3cd-4c47-a014-dc7746aed7e2 failed: 'ru'
[' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 151, in recursive_execute\n output_data, output_ui = get_output_data(obj, input_data_all)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 81, in get_output_data\n return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 74, in map_node_over_list\n results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_Custom_Nodes_AlekPet\ArgosTranslateNode\argos_translate_node.py", line 182, in argos_translate_text\n self.langs_support = ALL_CODES[from_translate]["targets"]\n ~~~~~~~~~^^^^^^^^^^^^^^^^\n']
2024-07-02 15:01:06,234 ERROR Job 4d98eac8-ff8e-4212-b49d-152aa3560f4e failed: 'ru'
[' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 151, in recursive_execute\n output_data, output_ui = get_output_data(obj, input_data_all)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 81, in get_output_data\n return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 74, in map_node_over_list\n results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_Custom_Nodes_AlekPet\ArgosTranslateNode\argos_translate_node.py", line 182, in argos_translate_text\n self.langs_support = ALL_CODES[from_translate]["targets"]\n ~~~~~~~~~^^^^^^^^^^^^^^^^\n']
2024-07-02 15:01:07,134 ERROR Job 5434b0a4-4711-41d7-9d17-bb39d234cae4 failed: 'ru'
[' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 151, in recursive_execute\n output_data, output_ui = get_output_data(obj, input_data_all)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 81, in get_output_data\n return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 74, in map_node_over_list\n results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_Custom_Nodes_AlekPet\ArgosTranslateNode\argos_translate_node.py", line 182, in argos_translate_text\n self.langs_support = ALL_CODES[from_translate]["targets"]\n ~~~~~~~~~^^^^^^^^^^^^^^^^\n']
2024-07-02 15:01:07,601 ERROR Job 7315a369-be80-4521-91b9-8fbdc6caa148 failed: 'ru'
[' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 151, in recursive_execute\n output_data, output_ui = get_output_data(obj, input_data_all)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 81, in get_output_data\n return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\execution.py", line 74, in map_node_over_list\n results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n', ' File "D:\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI_Custom_Nodes_AlekPet\ArgosTranslateNode\argos_translate_node.py", line 182, in argos_translate_text\n self.langs_support = ALL_CODES[from_translate]["targets"]\n ~~~~~~~~~^^^^^^^^^^^^^^^^\n']

`

[Drakosha405]

Unfortunately mine doesn't work, could you please take a look ?

I'm not familiar enough with the code to determine what caused the error. as you can see, everything worked for many people.
I was just experimenting with it, nothing more.
try DeepTranslatorCLIPTextEncodeNode maybe its work for u.
return self.add("DeepTranslatorCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="russian", auth_data=None)

[XylitolJ]

I tried with DeepTranslator and it work! thank you

[dkqjhoifh]

Thank you very much, it worked!! I can finally type in Chinese!!

[lovemakerz]

Hello, awesome,
You can explain for beginners ?

Only edit this line ?
or another file / script is needed ?
( i'm french )

Thanks

[lovemakerz]

Thanks a lot, i have try, and working fine !!!

[Drakosha405]

Hello Drakosha405, I referred to your method to modify the Deep translation, but there may be errors. I hope you can help me see how to modify the code when you have time. Thank you

return self.add("DeepTranslatorCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="chinese simplified", service="MyMemoryTranslator [free]")

I started it, and it seems to translate from the desired language, but when loading the workflow, the value from_translate = auto is always displayed for verification.
return self.add("DeepTranslatorCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="russian", auth_data=None, service="MyMemoryTranslator [free]")

but the check showed that in auto mode there is no translation of the mixed prompt, and when you select the desired one, everything is translated.

Prompt: cyberpunk, кирпичный дом (cyberpunk, bricks house) with same seed.
Prompt: girl, кошачьи уши (girl, cat ears) with same seed.

from_translate="auto"

from_translate="russian"

[Marksusu]

有这个翻译节点实在太好用了

[SyamsQ]

Hello Drakosha405, I referred to your method to modify the Deep translation, but there may be errors. I hope you can help me see how to modify the code when you have time. Thank you
return self.add("DeepTranslatorCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="chinese simplified", service="MyMemoryTranslator [free]")

I started it, and it seems to translate from the desired language, but when loading the workflow, the value from_translate = auto is always displayed for verification. return self.add("DeepTranslatorCLIPTextEncodeNode", 1, clip=clip, text=text, from_translate="russian", auth_data=None, service="MyMemoryTranslator [free]")

but the check showed that in auto mode there is no translation of the mixed prompt, and when you select the desired one, everything is translated.

Prompt: cyberpunk, кирпичный дом (cyberpunk, bricks house) with same seed. Prompt: girl, кошачьи уши (girl, cat ears) with same seed.

from_translate="auto"

from_translate="russian"

do we need something to download for get this translate feature?

[Drakosha405]

do we need something to download for get this translate feature?

u need this install

[Drakosha405]

be careful when using DeepTranslatorCLIPTextEncodeNode encountered this. And it broke everything
Error: Server Error: You made too many requests to the server.According to google, you are allowed to make 5 requests per secondand up to 200k requests per day. You can wait and try again later oryou can try the translate_batch function

I had to remove service="MyMemoryTranslator [free]" ... let's see if there will be such a mistake again.

[syl1130]

only remove , service="MyMemoryTranslator [free]" ?

[Drakosha405]

only remove , service="MyMemoryTranslator [free]" ?

yes

[xingewh]

Please note that when using the Argos translation node, if you input English, it is highly likely that they will output incorrect results. Moreover, the actual translation results are not visible on the Krita AI interface.

[Marksusu]

的确，应该有一个单独的“提示词翻译、生成、修改面板”（可以结合某些借助 ollama 的提示词生成工具）── 仅仅用于调整英文提示词的草稿，直到修改满意了，再应用到正反提示词节点。

[xingewh]

的确，应该有一个单独的“提示词翻译、生成、修改面板”（可以结合某些借助 ollama 的提示词生成工具）── 仅仅用于调整英文提示词的草稿，直到修改满意了，再应用到正反提示词节点。

是的，现在的Argos节点在翻译长文本的时候，结果有时不准确。我在翻译节点的那个项目里和作者就这一问题进行了沟通。这个不准备可能是因为ArgosTranslateCLIPTextEncodeNode使用的是本地模型进行翻译，模型规模较小，可能存在问题。DeepTranslatorCLIPTextEncodeNode就好很多，它是在线调用谷歌翻译等网站的API来实现的，但在国内需要使用全局代理才可以，并且就算是使用了全局代理，我把它加到了kirta的代码中也会出现与网络相关的报错。

为了准确的使用结果，还是使用原来的英文输入吧。可以在comfyui的上建一个合适的翻译工作流，然后复制过来。

[Marksusu]

的确，应该有一个单独的“提示词翻译、生成、修改面板”（可以结合某些借助 ollama 的提示词生成工具）── 仅仅用于调整英文提示词的草稿，直到修改满意了，再应用到正反提示词节点。

是的，现在的Argos节点在翻译长文本的时候，结果有时不准确。我在翻译节点的那个项目里和作者就这一问题进行了沟通。这个不准备可能是因为ArgosTranslateCLIPTextEncodeNode使用的是本地模型进行翻译，模型规模较小，可能存在问题。DeepTranslatorCLIPTextEncodeNode就好很多，它是在线调用谷歌翻译等网站的API来实现的，但在国内需要使用全局代理才可以，并且就算是使用了全局代理，我把它加到了kirta的代码中也会出现与网络相关的报错。

为了准确地使用结果，还是使用原来的英文输入吧。可以在comfyui的上建一个合适的翻译工作流，然后复制过来。

https://github.com/kingzcheung/ComfyUI_kkTranslator_nodes
用这个改一下是可以的，kkTranslator 是用的本地模型，能忠实地翻译逐个短语，比 Ollama 稳定（Ollama 里有些模型会胡言乱语，但好处是你可以指派它去编写提示词，任其发挥，也很有意思：https://github.com/ScreamingHawk/comfyui-ollama-prompt-encode，也值得一试。）。
我还发现，如果先在终端打开 ComfyUI， 再连接 Krita 就能在终端看到每次的翻译结果，这真是太好玩了。

[Drakosha405]

in 1.20.0 version.
now we need edit 355 line in comfy_workflow.py