I have some encouraging results in using the command “correct into simple English” for a sentence alone with text-davinci-003 and using the API and text completion. However, my issue is that each promt takes time to process, so I am trying to speed up the process and send all the sentences (approx 50) at once. The problem here is, of course, that everything seems to get tokenized alltogether and the sentences returned are not the same as when processing each sentence alone. I assume it processes it as a single coherent paragraph. Is there a way to send a batch of sentences in one go that gets processed as if each sentence is being sent by themselves?
Maybe tell the model to treat each sentence separately?
If you send them all together then they become part of your prompt.
You need to create a simple script that loops through your sentences and sends multiple parallel API requests to Open AI.
That will do the job, and fast!
You can also use a Google sheets addon to do the same thing.
Yes, there is a way to send a batch of sentences at once and process them as if each sentence was sent independently. Python and NLTK are great tools for processing individual sentences and helping to simplify and translate complex sentences into simple English. We hope this can help you speed up the process of translating and simplifying your sentences.
This answer was created with the help of text-davinci-003 and the text completion API using the playground on the date and time mentioned in the post.