Token Limit Warning Message

When working with QPilot, you may encounter messages related to session token limits. These messages indicate whether the token limit issue is due to the latest prompt or the session history. Here’s how to troubleshoot these situations:

Note: Limitations are set out by the GPT model you have chosen when configuring your organisation QPilot. Click here for more on configuring your hub.

If the Session Token Limit is Reached Due to the Latest Prompt:

Message Displayed: “Prompt failed. Prompt text and references exceed the maximum word count. Reduce the number of words in the prompt and attachments and try again.”
Cause: This occurs when your latest prompt, including any references, exceeds 60% of the maximum session token limit. The token limit includes both the prompt and the response, so if the latest prompt uses up a significant portion of the available tokens, you may encounter this issue.
Resolution: To address this, consider simplifying or shortening your prompt. Remove unnecessary references or split the prompt into smaller parts. You can also try adjusting the prompt to use fewer tokens, ensuring it remains within the allowable limit.
If the Session Token Limit is Reached Due to the Session History:

Message Displayed: “Prompt failed. Your current chat conversation reached the maximum word count. Start a new chat and include any necessary context.”
Cause: This occurs when the cumulative tokens used across the entire session, including previous prompts and responses, approach the maximum session token limit. Even if the latest prompt itself does not exceed 60% of the token limit, the accumulated tokens from prior interactions can contribute to reaching the limit.
Resolution: To resolve this issue, consider clearing the session history or starting a new session. This can be done by ending the current session and beginning a fresh one. Clearing the history will help free up tokens and prevent hitting the limit too quickly.
General Tips:

Monitor Token Usage: Keep an eye on how many tokens are being used in both prompts and responses. This can help you better manage and avoid reaching the limit.
Break Down Prompts: If you have a complex query, break it down into smaller, more manageable parts. This approach helps reduce token usage and makes it easier to manage session limits.
Review Prompt Efficiency: Ensure that your prompts are clear and concise. Removing extraneous details can help minimize token consumption.
By understanding these messages and their causes, you can effectively manage your token usage and maintain smooth interactions with QPilot.