• To improve security, we will soon start forcing password resets for any account that uses a weak password on the next login. If you have a weak password or a defunct email, please update it now to prevent future disruption.

Tool RPGM DazedMTLTool - A tool that provides quality MTL translations using ChatGPT

dazedanon

Engaged Member
Modder
Uploader
Donor
Jul 24, 2017
2,090
19,651
is there a way do translate jp novels from syosetu with this? there are a bunch of favs there that i read with google translate, im wondering if this will give a better reading experience.
Maybe it depends on if the scenario text can be accessed in the files.
Hi, can i request something?


I don't know if you can create a excel file for example and store the original text and the translate text


For example:

View attachment 3043987

Thanks
For the Japanese language? The csv script should tl any JA on the 1st column and put he tl on column 2.
 

trumpthatbitch

Active Member
Jun 23, 2017
911
1,402
Lol lets put a ton of eroge into chatgpt.

Wonder what the results will be to the main AI driven dialogue.
 

ell_123

New Member
Nov 26, 2017
11
8
Hi Dazed,

I was wandering if you can help me out a bit. I extracted scripts (attached) from Kirikiri game via Translator++ and tried to translate them with your tool. However when I try to use it, Estimation shows 0 tokens and Translation shows error since there are only None values. I cannot figure out why csv is not getting parsed.

I also tried one of the json files from rpgm game and that one worked fine.
 

dazedanon

Engaged Member
Modder
Uploader
Donor
Jul 24, 2017
2,090
19,651
Hi Dazed,

I was wandering if you can help me out a bit. I extracted scripts (attached) from Kirikiri game via Translator++ and tried to translate them with your tool. However when I try to use it, Estimation shows 0 tokens and Translation shows error since there are only None values. I cannot figure out why csv is not getting parsed.

I also tried one of the json files from rpgm game and that one worked fine.
Fixed now. Thanks for reporting it.
 
  • Like
Reactions: ell_123

ell_123

New Member
Nov 26, 2017
11
8
I think there is a small bug in csv.py, line 215. When script fails and selected option is csv from Translator++ there is no 'text' variable. Replacing with 'jaString' variable gave me proper message.

Another thing, how do you handle cases when translation stopped midway? It seems that if I rerun the script translation starts from the beginning. For now I'm manually removing already translated lines and storing translated file separately, but I'm wondering if I'm doing something wrong.
 

dazedanon

Engaged Member
Modder
Uploader
Donor
Jul 24, 2017
2,090
19,651
I think there is a small bug in csv.py, line 215. When script fails and selected option is csv from Translator++ there is no 'text' variable. Replacing with 'jaString' variable gave me proper message.

Another thing, how do you handle cases when translation stopped midway? It seems that if I rerun the script translation starts from the beginning. For now I'm manually removing already translated lines and storing translated file separately, but I'm wondering if I'm doing something wrong.
I probably need to revamp the script a bit. Let me see what I can do. Probably best to join the Discord so I can help you out a bit easier. OpenAI sometimes randomly times out, it happens.
 
Last edited:

dazedanon

Engaged Member
Modder
Uploader
Donor
Jul 24, 2017
2,090
19,651
I'm curious for future reference, how did you go about fixing clipping issues?
I edited the ks files directly using the tyrano module in my tool. It has its own wordwrapping so running it through the tool fixes it.
 

tophunk4

New Member
Feb 10, 2018
2
0
Is it possible to substitute ChatGPT with a locally run LLM model in Oogabooga?

I can imagine it can save some money for someone with a decent GPU that can run some custom models
 

dazedanon

Engaged Member
Modder
Uploader
Donor
Jul 24, 2017
2,090
19,651
Is it possible to substitute ChatGPT with a locally run LLM model in Oogabooga?

I can imagine it can save some money for someone with a decent GPU that can run some custom models
Maybe but GPTs translation success mainly comes from it being trained in a large variety of different languages. Might not get the same quality from most local LLMs.