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

[Shisaye]

Links for the models mentioned above.

static/unweighted

You must be registered to see the links

You must be registered to see the links

imatrix/weighted

You must be registered to see the links

You must be registered to see the links

I would guess that most people will get them through the LM Studio UI.
Also I mean this 8B model:

You must be registered to see the links

Not entirely sure what the difference to the v1 is, but I've only tested the v2.

 

[Entai2965]

I fixed the links. There are quite a few different 8b models. v2 is different than v2.0 apparently.

 

[Shisaye]

I fixed the links. There are quite a few different 8b models. v2 is different than v2.0 apparently.

No clue... They are exactly the same size and were built from the same

You must be registered to see the links

.

Maybe he published it as v2.0 accidentally and then just made a new post with just v2?

 

[Shisaye]

I've implemented a very simple fallback model option in v2.027 now.
Basically it's 3 new options to set another api url, model name, and optional api key.
And whenever a LLM fails so badly that it would normally abort the batch translation, it will now try the current request batch again with the chosen fallback model using the same prompts, settings, etc.

 

[fantasmic]

And as if that wasn't enough, yet, It's a VL model, it has vision. You can use it for SEP picture translations, too. (Although I wouldn't use the 8B variant for that, that probably wouldn't be a meaningful step up from just using paddle.)

I've been using QWEN2.5 7B for SEP and feel that it's a meaningful step up from just using paddle. Maybe not "you should buy a new gpu for this", but definitely "best use it if you have it." I'd expect QWEN3 8B is probably in the same boat.

 

[Shisaye]

I've been using QWEN2.5 7B for SEP and feel that it's a meaningful step up from just using paddle. Maybe not "you should buy a new gpu for this", but definitely "best use it if you have it." I'd expect QWEN3 8B is probably in the same boat.

Since OCR stuff takes "relatively" little tokens and you usually only have a few pictures that need translation I don't really see why someone shouldn't use the 32B version running on CPU though. All that's needed for the IQ3_XS version for example would be 16GB of ram. That's like $30 if you buy the Chinese stuff.