Tool RPGM SLR - Offline JP to EN Translation for RPG Maker VX, VX Ace, MV, MZ, and Pictures

R

RenderedFatality

Member
Dec 1, 2018
117
107
Theres some fancier things you can do with that particular model to speed it up, ik_llama.cpp and a specific argument

--override-tensor "([0-2]).ffn_.*_exps.=CUDA0" --override-tensor "([3-9]|[1-9][0-9]+).ffn_.*_exps.=CPU" or something like that, + tuning --ngl argument. + --fmoe on ik_llama.cpp

I was taking a better portion of an hour earlier capturing responses and creating a tiny dataset to finetune the 4b and 14b earlier, and the 4b is fairly promising, accidently baked it for 8 epochs on 405 examples so is it overfitted? absolutely. but it definitely wasnt bad.
 
Last edited:
R

RenderedFatality

Member
Dec 1, 2018
117
107
Put together a few fine tunes on some of the Qwen3 models that should be applicable to people from a variety of hardware levels. They should all be useful for folks on constrained hardware. They only were utilized in batchsize 500 or less.

You must be registered to see the links


You must be registered to see the links


You must be registered to see the links


You will need to manually assign a jinja chat template to all of these when using LM Studio or VLLM. I never put them into the tokenizer_config.json.

You must be registered to see the links
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
RenderedFatality said:
Theres some fancier things you can do with that particular model to speed it up, ik_llama.cpp and a specific argument

--override-tensor "([0-2]).ffn_.*_exps.=CUDA0" --override-tensor "([3-9]|[1-9][0-9]+).ffn_.*_exps.=CPU" or something like that, + tuning --ngl argument. + --fmoe on ik_llama.cpp

I was taking a better portion of an hour earlier capturing responses and creating a tiny dataset to finetune the 4b and 14b earlier, and the 4b is fairly promising, accidently baked it for 8 epochs on 405 examples so is it overfitted? absolutely. but it definitely wasnt bad.
Click to expand...
RenderedFatality said:
Put together a few fine tunes on some of the Qwen3 models that should be applicable to people from a variety of hardware levels. They should all be useful for folks on constrained hardware. They only were utilized in batchsize 500 or less.

You must be registered to see the links


You must be registered to see the links


You must be registered to see the links


You will need to manually assign a jinja chat template to all of these when using LM Studio or VLLM. I never put them into the tokenizer_config.json.

You must be registered to see the links
Click to expand...
Sounds good, and feel free to use this thread and the anime sharing thread to share models give advice and stuff, but I think I'm done for now.
I was just trying to find an example that shows that DSLR works if you use it with a good model, so that I no longer get bug reports claiming that it doesn't and me not really being able to prove to them that it does, because they won't pay DeepSeek for it.

My focus is now back to improving parsing, plugin compatibility, etc., since I'm not actually planning to use a local LLM.
My hardware is so terrible I definitely spend more money on power and AC using them, than I would paying for DeepSeek during their off hour discount.
The only reason I was able to test so many models is because most already failed at the second batch.
The actual full test with the qwen3 model now took me 9.5h and that was only a single game. :HideThePain:
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
I've released v1.141
Fixed an issue with the DSLR wrapper that sometimes destroyed \N[] commands at the start of cells.
Fixed also some other stuff concerning \n[] commands that made translation quality worse.
There was no issue with SLR, so if you don't use DSLR you can ignore this update.
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
Something I didn't think about much when designing SLR Translator, because it wasn't really necessary when everyone was just using SLR is user friendly debugging.
For example that \N[] error just now was completely invisible to the user because the error already happened in the wrapping stage which means it wasn't actually fully sent to the DSLR engine and as a result the error detection also did not reject those translations since they technically weren't wrong, and the average user looking at the result would have thought it's just the LLM screwing up.
Because since no error was detected by DSLR it also did not log the original and translation texts in the F12 console.

The way I found out was directly placing alerts in the code for those tests, that then showed me that the LLM wasn't actually sent half of the \N[].


To get to the point of writing this, should I make more "built in" debugging stuff? In the sense that it either logs even more stuff in the console by default, even if no error was detected? (That could get really messy and potentially eat a lot of ram)
Should I in general log more stuff directly inside the batch-translation, pre-processing, and post-processing windows?

I also don't want to give information overload that might make it seem like something went wrong when it didn't, or make someone miss an actual error because it's buried inside hundreds of logs.
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
wibuntr said:
View attachment 5018056
i'm always using the first SLR but what is other SLR and DSLR?
Click to expand...
The heck? :confused:
There should only be one SLR entry.
Red Google and Red Piggyback shouldn't be in that list either.
Maybe you need to delete the profile folder in Appdata, because there's some wrong old information in there.

DSLR is my universal LLM engine. It's designed to be used with DeepSeekv3, but can also be used with free locally hosted LLMs like qwen3-30b-a3b-abliterated-erotic-i1-IQ3_XSS as well.
It's mostly for people trying to make an AI assisted proper translation, it's way too slow for a quick MTL.
 
  • Like
Reactions: wibuntr
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
I've released v1.143.
There was a bug in DSLR, that could cause a failure loop when a \n[] placeholder was at the very end of the input.
Seems like a small thing, but can crash the entire batch translation.
 
  • Like
Reactions: fantasmic
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
LinkR34 said:
Hi Shisaye, can you help me with this game?
https://f95zone.to/threads/translat...uilding-the-agency-rj01286203-imoniapo.238311
DLsite code:
You must be registered to see the links


I tried translating it with SLR, but it only translated half of the game (used vesion 1.03). The other half doesn't show up for some reason, and it doesn't even give any errors. Could you take a look? Thank you
Click to expand...
That game uses the ExternMessage plugin which lets the dev definde a random file with text in it to be able to write the dialogue without having to use the RPGM editor.
You can find the missing text in the "ExternMessage.csv" located in "data".

To load that file into SLR Translator you can use a custom parser.
I've included a fitting one in v1.144.
Start a new project with the custom parser option.
For the parser option choose the ExternMessageMZCSV.tpm in the custom parser folder of SLR Translator.
Then choose the folder the ExternMessage.csv is in for the project and start it.

It will grab all the text from it and you can do the normal SLR translation process on it.
 
  • Like
Reactions: LinkR34
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
testing101hi said:
Is there a way to use for older RPG Maker?
Click to expand...
SLR Translator currently only works for VX, VX Ace, MV, or MZ.

Old outdated RPG Maker engines like 95, 98, 2000, 2003, XP, or Wolf have very different structure and you would need completely new parsers for them.
They are really not nice to work with.
 
  • Sad
Reactions: derakino999
LinkR34

LinkR34

Member
Dec 9, 2017
295
261
Shisaye said:
That game uses the ExternMessage plugin which lets the dev definde a random file with text in it to be able to write the dialogue without having to use the RPGM editor.
You can find the missing text in the "ExternMessage.csv" located in "data".

To load that file into SLR Translator you can use a custom parser.
I've included a fitting one in v1.144.
Start a new project with the custom parser option.
For the parser option choose the ExternMessageMZCSV.tpm in the custom parser folder of SLR Translator.
Then choose the folder the ExternMessage.csv is in for the project and start it.

It will grab all the text from it and you can do the normal SLR translation process on it.
Click to expand...
Thank you, it worked. But now the following error popped up:

CnP_22072025_203851.png
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
LinkR34 said:
Thank you, it worked. But now the following error popped up:

View attachment 5068791
Click to expand...
This is a bit of a weird one.
I've attached a more strict custom parser to this post.

Use it on the ExternMessage.csv.
Click on the Prepare Project for Batch Translation button.
Mark all cells by clicking on the first cell then scrolling down to the last used cell and clicking it while holding shift. (Do not select the last empty cell.)
Rightclick one of the selected cells and click on Tags > Green.
So that every used cell in your ExternMessage.csv project is green.
Then do the batch translation.
Press the "Fix Cells and Check for Errors" button.
Now you can export and it should work.

The issue was that it doesn't treat the cells as scripts, it treats them as normal dialogue, which disables a lot of the script fixing functionality of SLR.
But if you tag the cells green it treats them as "approved scripts" instead.
 
  • Heart
Reactions: LinkR34
LinkR34

LinkR34

Member
Dec 9, 2017
295
261
Shisaye said:
This is a bit of a weird one.
I've attached a more strict custom parser to this post.

Use it on the ExternMessage.csv.
Click on the Prepare Project for Batch Translation button.
Mark all cells by clicking on the first cell then scrolling down to the last used cell and clicking it while holding shift. (Do not select the last empty cell.)
Rightclick one of the selected cells and click on Tags > Green.
So that every used cell in your ExternMessage.csv project is green.
Then do the batch translation.
Press the "Fix Cells and Check for Errors" button.
Now you can export and it should work.

The issue was that it doesn't treat the cells as scripts, it treats them as normal dialogue, which disables a lot of the script fixing functionality of SLR.
But if you tag the cells green it treats them as "approved scripts" instead.
Click to expand...
Thanks, it's working for now. But the characters have their names in japanese. I tried looking up each character's name in the translator, but they were already edited (I also checked ExternMessage.csv). Do you think they could be images?

Screenshot2.png
 
Shisaye

Shisaye

Engaged Member
Modder
Dec 29, 2017
3,392
6,061
I'm working on a very large new module for SLR Translator and I'd like to have some testers/opinions.
If you're willing to do it please tell me, but for you to use the full potential of the system you would need at least 24gb of ram and around 50gb of free disc space. (If you only have 16gb of ram, but you have a cuda capable GPU with at least 8gb of vram it would work, too.)
 
N

natsugajeel2000

Active Member
Jan 4, 2018
751
372
Hi Shisaye,

I tried to put this game onto SLR, its a RPGM MZ, it was orignally just the game.exe but I unpacked it, but still it doesnt recognize it:

Game:
You must be registered to see the links


The game won't come on dlsite due to the AI blocking.
1755098584253.png
1755098603405.png

1755098369922.png
 
You must log in or register to reply here.
Top Bottom