potential bug / error using cgpt4 & default 1000 or 2000 token length

Viewing 2 replies - 1 through 2 (of 2 total)
  • Plugin Support Val Meow

    (@valwa)

    Hello @xenek ,

    Can you replicate this issue multiple times ? and with using a different model besides GPT-4?

    I’m curious about the reason behind the last AI response being empty prior to sending a new message. It’s possible that this could be an error resulting from multiple requests being sent simultaneously.

    Thread Starter xenek

    (@xenek)

    Using Version 1.8.5.

    I copied my chatbot

    I reset the copy to defaults

    (

    Turbo

    1024 tokens

    Max 15 input 512)

    tested.

    Seems to work ok.

    So either the bug has been repaired, was local or related to carriage service, or was only on the first install/first chatbot.

    update:

    I set the type to ChatGPT4 and left the other setting (including token) the same, and it continues to work. This is to the 2nd ‘newchatbot’ not the ‘default’ 1st chatbot, that I originally customised

    — hope this helps ?? Thanks for functional software! I’m grateful to use it.

Viewing 2 replies - 1 through 2 (of 2 total)
  • The topic ‘potential bug / error using cgpt4 & default 1000 or 2000 token length’ is closed to new replies.