Skip to content

Latest commit

 

History

History
59 lines (48 loc) · 1.56 KB

playground-settings.org

File metadata and controls

59 lines (48 loc) · 1.56 KB
Raw

OpenAI Playground Settings

These are the official playground settings of the OpenAI Advanced Playground for GPT-3.

Also, here are some glossaries
http://github.com/semiosis/pen.el/blob/master/glossary.txt https://github.com/semiosis/glossaries-gh/blob/master/openai-api.txt

Engine

The different GPT-3 models the API is able to access:

  • Ada,
  • Babbage,
  • Curie, and
  • Davinci.

Frequency Penalty

How much to penalize new tokens based on their existing frequency in the text so far. Decreases the model’s likelihood to repeat the same line.

Presence Penalty

How muoh to penalize new tokens based on whether they.r appear in the text so far. Increases the model’s likelihood to talk about new topics.

Best of (Value from 1-20, Default: 1)

Generates multiple completions server-side, and displays only the best. Streaming only works when set to 1.

Since it acts as a multiplier on the number of completions; this parameters can eat into your token quota very quickly - use caution!

Stop Sequences

Up to four sequences where the API will step generating further tokens. The returned text will not contain the step sequence.

Inject Start Text

Text to append after the user’s input to format the model for a response.

Inject Restart Text

Text to append after the model’s generation to continue the patterned structure.

Show Probabilities

Toggle token highlighting which indicates how likely a token was to be generated. Helps to debug a given generation, or see alternative options for a token.

  • States
    • Off
    • Most likely
    • Least likely
    • Full spectrum