Been using chatting to generate long and complex vba scripts.
The issues is the script is too long to display on a single display page
The bot has attempted to “ chunk it” and send in concesuctive pages automatically but it’s not constant and also hangs up before all chunks have been received.
Are there any ways of receiving the full script ,eg………hyperlink , drop box, emails etc
MikeH
I have also experienced similar issues with Condice Lungo when using the 3.5 model and asking it to continue where it left off. After a few requests, it starts going into a loop.
I haven’t encountered this error with gpt4.
You could try asking for code division by tokens.
For example,
PROMPT ASKING FOR CODE GENERATION:
Give me the code as if you’re providing it in separate calls, knowing that for each response, you should give me a maximum of 5000 tokens and wait for me to write “Continue” to give me the next block.
Thanks for that .
Using the text script you suggested about token display seems to have worked
Is this in the api? I’ve often wondered about this. In the web interface I can continue, but it has some kind of built in chat context.
Not sure how to duplicate this behavior in the API?