WOLFRAM

Wolfram Language & System Documentation Center

"PaLM" (Service Connection)

Obsolete. Use "GoogleGemini" instead.
Use the PaLM API with the Wolfram Language.

Connecting & Authenticating

ServiceConnect["PaLM"] creates a connection to the PaLM API. If a previously saved connection can be found, it will be used; otherwise, a new authentication request will be launched.
Use of this connection requires internet access and a Google account.

Requests

Request:

"Completion" — create text completion for a given prompt

Required parameters:
  • "Prompt"the prompt for which to generate completions
    • Optional parameters:
  • "MaxTokens"Automaticmaximum number of tokens to generate
    "Model"Automaticname of the model to use
    "N"Automaticnumber of completions to return (1 to 8)
    "StopTokens"Noneup to four strings where the API will stop generating further tokens
    "Temperature"Automaticsampling temperature (between 0 and 1)
    "TopProbabilities"Automaticsample only among the k highest-probability classes
    "TotalProbabilityCutoff"Nonesample among the most probable classes with an accumulated probability of at least p (nucleus sampling)
    • Request:

    "Chat" — create a response for the given chat conversation

    Required parameters:
  • "Messages"a list of messages in the conversation
    • Optional parameters:
  • "Model"Automaticname of the model to use
    "N"Automaticnumber of completions to return (1 to 8)
    "Temperature"Automaticsampling temperature (between 0 and 1)
    "TopProbabilities"Automaticsample only among the k highest-probability classes
    "TotalProbabilityCutoff"Nonesample among the most probable classes with an accumulated probability of at least p (nucleus sampling)
    • Request:

    "TextEmbedding" — create an embedding vector representing the input text

    Required parameters:
  • "Text"text that the model will turn into an embedding
    • Optional parameters:
  • "Model"Automaticname of the model to use
    • Request:

    "TokenCount" — run a model's tokenizer on a prompt and return the token count

    Required parameters:
  • "Input"text or messages to tokenize
    • Optional parameters:
  • "Model"Automaticname of the model to use

    • Examples

    open all close all

    Basic Examples  (2)

    Create a new connection:

    Wolfram Language code: palm = ServiceConnect["PaLM"]

    Complete a piece of text:

    Wolfram Language code: ServiceExecute[palm, "Completion", {"Prompt" -> "Roses are red Violets are blue, "}]

    Generate a response from a chat:

    Wolfram Language code: ServiceExecute["PaLM", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "hello there"|>}}]

    Scope  (11)

    Completion  (3)

    Change the sampling temperature:

    Wolfram Language code: ServiceExecute["PaLM", "Completion", {"Prompt" -> "please complete this text.", "Temperature" -> 1}]

    Increase the number of characters returned:

    Wolfram Language code: ServiceExecute["PaLM", "Completion", {"Prompt" -> "please complete this text.", "MaxTokens" -> 3}]

    Return multiple completions:

    Wolfram Language code: ServiceExecute["PaLM", "Completion", {"Prompt" -> "please complete this text.", "N" -> 2}]

    Chat  (3)

    Respond to a chat containing multiple messages:

    Wolfram Language code: ServiceExecute["PaLM", "Chat", {"Messages" -> { <|"Role" -> "User", "Content" -> "tell me three colors"|>}}]

    Change the sampling temperature:

    Wolfram Language code: ServiceExecute["PaLM", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "tell me three colors"|>}, "Temperature" -> 0}]

    Return multiple completions:

    Wolfram Language code: ServiceExecute["PaLM", "Chat", {"Messages" -> {<|"Role" -> "User", "Content" -> "What colors are bears?"|>}, "N" -> 2}]

    TextEmbedding  (2)

    Compute the vector embedding of some text:

    Wolfram Language code: ServiceExecute["PaLM", "TextEmbedding", {"Text" -> "ciao"}]//Short

    Compute the distance between vector embeddings to find semantic similarities:

    Wolfram Language code: embeddings = Table[ ServiceExecute["PaLM", "TextEmbedding", {"Text" -> text}], {text, {"Paris", "London", "I love rock'n'roll"}} ];
    Wolfram Language code: DistanceMatrix[embeddings]//MatrixPlot

    TokenCount  (3)

    Get a token count for a string prompt:

    Wolfram Language code: ServiceExecute["PaLM", "TokenCount", {"Input" -> "hello", "Model" -> "text-bison-001"}]

    Get a token count for a messages prompt:

    Wolfram Language code: ServiceExecute["PaLM", "TokenCount", {"Input" -> {<|"Role" -> "User", "Content" -> "hello"|>}, "Model" -> "chat-bison-001"}]

    When not specified, the model is chosen automatically:

    Wolfram Language code: ServiceExecute["PaLM", "TokenCount", {"Input" -> "hello"}]
    Wolfram Language code: ServiceExecute["PaLM", "TokenCount", {"Input" -> {<|"Role" -> "User", "Content" -> "hello"|>}}]