Comparing the model completion

Each model completion text box under the prompt text box in Prompt Catalog, provides the completion of the sent prompt to the Large Language Model (LLM).

Additionally, these metrics are displayed as part of the text box:

  • A list of supported models that include:
    • Description of model
    • Token context size
  • Gear icon for parameters to influence your response:
    • Temperature: Number between 0 and 1, where lower temperature results in concise and factual responses, while higher temperature produces creative and diverse results.
    • Max Response Length: Maximum number of tokens to generate a response.
    • Top P: Control the randomness from the LLM by setting a greater than or equal to probability of output.
    • Frequency Penalty: Number between -2.0 and 2.0, where positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model’s likelihood to repeat the same line verbatim.
    • Presence Penalty: Number between -2.0 and 2.0, where positive values penalize new tokens based on whether they appear in the text so far, increasing the model’s likelihood to talk about new topics.
  • Response calculations for generating your response:
    • Response Time
    • Total Tokens