Model to use
Optional
apiAPI key to use.
Optional
endpointCustom inference endpoint URL to use
Optional
frequencyPenalizes repeated tokens according to frequency
Optional
includeCredentials to use for the request. If this is a string, it will be passed straight on. If it's a boolean, true will be "include" and false will not send credentials at all.
Optional
maxMaximum number of tokens to generate in the completion.
Optional
stopThe model will stop generating text when one of the strings in the list is generated.
Optional
temperatureSampling temperature to use
Optional
topKInteger to define the top tokens considered within the sample operation to create new text.
Optional
topPTotal probability mass of tokens to consider at each step
Generated using TypeDoc
Interface defining the parameters for configuring the Hugging Face model for text generation.