When I click (or double-click) the icon nothing happens. I have uninstalled and re-installed it, with the same result. (When I first tried to use it, some weeks ago, it worked.)
When I click (or double-click) the icon nothing happens. I have uninstalled and re-installed it, with the same result. (When I first tried to use it, some weeks ago, it worked.)
Hi Mats Linder ,
I was trying to get the same to reproduce, but it works for me when I double click to open. However, might I suggest you first go to View tab and do a Reset window layout, maybe this will force to restore the closed windows.
Also, please remove the plugin from all the below paths:
c:\Users\[USERNAME]\AppData\Roaming\Trados\Trados Studio\17\Plugins\Packages\
c:\Users\[USERNAME]\AppData\Roaming\Trados\Trados Studio\17\Plugins\Packages\
c:\Users\[USERNAME]\AppData\Roaming\Trados\Trados Studio\17\Plugins\Unpacked\
c:\Users\[USERNAME]\AppData\Local\Trados\Trados Studio\17\Plugins\Packages\
c:\Users\[USERNAME]\AppData\Local\Trados\Trados Studio\17\Plugins\Unpacked\
c:\ProgramData\Trados\Trados Studio\17\Plugins\Packages\
C:\ProgramData\Trados\Trados Studio\17\Plugins\Unpacked\
Then install it again from the official page: https://appstore.rws.com/plugin/200/
Hope it helps?
Regards
Thanks! Resetting did the trick. I'd really like to know why. (Why does not closing and restarting Studio have the same effect?)
Now I have another (related) question: I can't get OpenAI to show any translations. No matter what I feed into it, this is what I get:
Hi Mats Linder ,
I know the OpenAI server is really slow sometimes especially if there are multiple calls, so there may be no results offered. Have you tried selecting the default prompt as well?
In addition, and to also backup your observations, we are aware of several blocking bits for this, such as regional restriction or your subscription service type .
This other thread may give you some overall of the above, can you please take a look and see if it helps?
Regards,
Hi Oana,
Thanks for your suggestions. However, regardless of what settings I try, I get no further than the message "Searching", which goes on forever without any results. The thread you point to is of no use to me, since my problem is not any of those mentioned there. So maybe the problem is one of the blocking bits that you mention. (I don't even know what the subscription type you mention is.)
Regards,
Mats
PS. However, I do receive an error message: Too many requests (when I click Search from source, which I believe is the way to start the search). Is that a clue?
Hi Oana,
Thanks for your suggestions. However, regardless of what settings I try, I get no further than the message "Searching", which goes on forever without any results. The thread you point to is of no use to me, since my problem is not any of those mentioned there. So maybe the problem is one of the blocking bits that you mention. (I don't even know what the subscription type you mention is.)
Regards,
Mats
PS. However, I do receive an error message: Too many requests (when I click Search from source, which I believe is the way to start the search). Is that a clue?
I don't even know what the subscription type you mention is.
Have you created your OpenAI account?
https://platform.openai.com/signup
You need a subscription to OpenAI before you'll be able to use it. It's not free!
I'm on a free trial (have not set up a paid account). My impression was that (a) you cannot even start OpenAI without an API key, and (b) you cannot get an API key without an account. So yes, I do have an account, but I still don't get any results.
Still struggning, also with the updated version...
More information: When I click Search from source I get this error messge: Too many requests. This is diffucult to understand since I have only made one.
The rate limits for the OpenAI API are determined by the type of account you have and the specific endpoint you're using. They are measured in two ways: requests per minute (RPM) and tokens per minute (TPM). The rate limit can be hit by either option (RPM or TPM) depending on what occurs first.
This page from OpenAI might be helpful: https://help.openai.com/en/articles/7416438-rate-limits
A simple summary of this might be to think about how the plugin works. It works by sending your text to OpenAI's servers, which then return the translated text based on whatever prompt you created or used.
However, there's a limit to how much text the plugin can send within a certain period of time. This limit is based not only on the number of requests but also on the total amount of text (or "tokens") being sent. If the plugin sends too much text too quickly, OpenAI's servers will say, "Hold on, you're asking me to do too much too quickly. I need a break." This is when you see the "too many requests" error message.
Now, if you're asking the plugin to do something complex, like creating 30 different translations of a single sentence and making sure they're all no more than 10% longer than the original, this could make the plugin hit the limit faster. That's because this task requires more processing power and time from OpenAI's servers.
So, if you're seeing the "too many requests" error, it could be because the plugin is trying to translate too much text or perform complex tasks too quickly. You might need to wait a bit before trying to translate more text or simplify the tasks you're asking the plugin to do.
If waiting or simplifying tasks isn't an option, we could ask OpenAI to increase the limit of how much text the plugin can send within a certain period. But OpenAI will only agree to this if they believe the plugin really needs it and our users won't abuse it.... and more importantly if doing this fits in with their own work. Note this statement on the link I provided "Please note that during the limited beta phase of GPT-4 we will be unable to accommodate requests for rate limit increases. In its current state, the model is intended for experimentation and prototyping".
Tagging Patrick Andrew Hartnett for info.
Hi Mats Linder , the model text-davinci-003 is available for free for the first 3 months. You need a paid account to access both gpt-3.5-turbo and gpt-4
You can always check the usage from your account on OpenAI, here: platform.openai.com/.../usage
Thank you Paul and Andrew!
It might be that the problem was that my trial account was no longer valid (the Usage was shown to have expired), although the first API key was created on 16 April. (I cannot find a date for the creation of the trial account.) Anyway, when I signed for a paid account, the app worked. Strangely, though, I always have to Reset Window Layout in order to be able to open it. Not a big issue but mysterious.
So your careful explanations, Paul, were unnecessary in this particular case (sorry to have taken your time once again when there was no need). However, they certainly give more understanding of the mechanisms of this very unusual and interesting feature.