- Jul 24, 2017
- 2,438
- 26,902
Check any mentions of English in the script files itself and change them. Then you can try again.I can't use it to translate into Vietnamese like I edited it in the prompt
Check any mentions of English in the script files itself and change them. Then you can try again.I can't use it to translate into Vietnamese like I edited it in the prompt
I followed the steps on the page that you have passed me, now I only need these 2 things, the tiktoken and the aoihttp, I had to install visual studio build tools to install the rest, but now I don't know where I can get it or how to install it.Your pip is broken for some reason. Use the get-pip method here to fix.You must be registered to see the links
You shouldn't actually need anything other than pip and python to install those dependencies which is how I know its an error with your python or pip installation. Someone else had this error and they fixed it after reinstalling everything related to pip and python and restarting their pc, terminal, etc. can only recommend you do the same since those dependencies are needed and I'm unsure of the exact cause of that issue.I followed the steps on the page that you have passed me, now I only need these 2 things, the tiktoken and the aoihttp, I had to install visual studio build tools to install the rest, but now I don't know where I can get it or how to install it. View attachment 3092963
Your pip is broken for some reason. Use the get-pip method here to fix.You must be registered to see the links
is it in rpgmaker mv.py?Check any mentions of English in the script files itself and change them. Then you can try again.
Tool was just updated to support translating to more languagesis it in rpgmaker mv.py?
then I really failed, I can't do it
I think its only necessary if you are using the official API. If you've modified it to use local ones or others then yeah you probably don't need it.Hey dazedanon I've modified the code a bitYou must be registered to see the links(based on older version) to support self hosted or other API's but I'm debating if I should sync and create a pull request as you mentioned that organization keys are absolutely necessary inYou must be registered to see the links.
I've tried it out with a free api (took two days because of 1 thread limit) and my own self hosted one and it seems to work without an organization but I haven't really done much testing to see if anything is screwed up.
I attached the json files I tested with. The game is RJ423458
Pull request has been made, as for the openai library there's a little known variable:I think its only necessary if you are using the official API. If you've modified it to use local ones or others then yeah you probably don't need it.
I'm actually curious how you got it working without completely changing the translateGPT function since it uses the official openai API library at the moment.
As for syncing and merging, I like those QOL changes you made so I think it will be worth it.
Maybe one day when its stable but for now too busy with development and translation.Good afternoon, I have seen your MTL translation tool and I wanted to try to use it but I don't know how to do it, that's why I was wondering if you could make a video following the steps you put in github, thank you very much for your time.
I understand, but thank you for the translations and for the time you took to answer me.Maybe one day when its stable but for now too busy with development and translation.
You need to paid the API for the ChatGTPSo it's gonna cost me to MTL a game or it's totally free?
Maybe it's worth revisiting this idea now that Gemini Pro just came out to Bard. It is particularly known for being lightning fast, and from what people have said & my own testing, I would say it's maybe a step above ChatGPT, but still below GPT-4. Reportedly it is less censored than ChatGPT is as well right now, though it seems you got around that with the API anyway.Haven't tried it. I have no idea what the cost of it is and Bard is generally considered worst than GPT4 so I just haven't bothered looking into it.
So they say but I'll believe it when its out. Bard's content filter was impossible to bypass for me. Let me know how local models go, I do know they struggle with speed, but might not be so bad of it can maintain the translation quality.Maybe it's worth revisiting this idea now that Gemini Pro just came out to Bard. It is particularly known for being lightning fast, and from what people have said & my own testing, I would say it's maybe a step above ChatGPT, but still below GPT-4. Reportedly it is less censored than ChatGPT is as well right now, though it seems you got around that with the API anyway.
Personally I'd be interested in running this on localhosted models as they have gotten quite good, although I wasn't able to test any as good as ChatGPT yet. I mostly used 7B and 13B, and 7B is a miss but 13B can be tentatively passable, which makes me think the 34b - 70b models may be pretty good. I have heard some say they can approach GPT-4 in some areas already. I'm getting a 7900 XTX in a few days and will have the VRAM to try them out, so may come back and comment then.
For references sake, what sort of token/s rate do you get from ChatGPT API? And does that take long to translate say a small game?So they say but I'll believe it when its out. Bard's content filter was impossible to bypass for me. Let me know how local models go, I do know they struggle with speed, but might not be so bad of it can maintain the translation quality.
Small games can take an hour max for all the text usually. It takes long for me to format things and make edits to the code, UI, pictures, etc.For references sake, what sort of token/s rate do you get from ChatGPT API? And does that take long to translate say a small game?