Ollama connection actions and objects

• •The Chat Completion action generates a chat interaction with a provided model and without streaming output.
To use this action, specify the following input fields:
• - ChatCompletionsRequestBody: The request body consists of a series of messages. The model generates a response to the last message, using earlier messages as context.
• - model: The name of your deployed model. You must deploy a model before making calls.
The following snippet is a sample of the request body:
<ChatCompletionsRequestBody xmlns:m="urn:informatica:ae:xquery:json2xml:meta-data">
<model>llama3.2:Latest</model>
<messages m:isArray="true">
<role>system</role>
<content>you are a helpfull assistant</content>
</messages>
<messages m:isArray="true">
<role>user</role>
<content>why is the sky blue</content>
</messages>
<stream>false</stream>
</ChatCompletionsRequestBody>
• •The Embeddings action generates embeddings from a model.
To use this action, specify the following input fields:
• - embeddingsRequest: The request body to get the embeddings.
• - model: The name of your deployed model. You must deploy a model before making calls.
The following snippet is a sample of the request body:
<EmbeddingsRequestBody xmlns:m="urn:informatica:ae:xquery:json2xml:meta-data">
<model>llama3.2:Latest</model>
<prompt>Here is an article </prompt>
<temperature>0.7</temperature>
<top_k>50</top_k>
<top_p>0.7</top_p>
</EmbeddingsRequestBody>
• •The Generate action generates a chat interaction with streaming output for a given prompt with a provided model. This is a streaming endpoint that contains a series of responses. The final response object will include statistics and additional data from the request.
To use this action, specify the following input fields:
• - GenerateRequestBody: The request body consists of a series of messages. The model generates a response to the last message.
• - model: The name of your deployed model. You must deploy a model before making calls.
The following snippet is a sample of the request body:
<GenerateRequestBody xmlns:m="urn:informatica:ae:xquery:json2xml:meta-data">
<model>llama3.2:Latest</model>
<prompt>Why is the sky blue?</prompt>
</GenerateRequestBody>
• •The List Local Models action lists all the models you have downloaded locally.
The following snippet is a sample response:
<root>
<models>
<size>2019393189</size>
<name>llama3.2:Latest</name>
<digest>a80c4f17acd55265feec403c7aef86be0c25983ab279d83f3bcd3abbcb5b8b72</digest>
<model>llama3.2:Latest</model>
<details>
<format>gguf</format>
<parameter_size>3.2B</parameter_size>
<parent_model/>
<families>llama</families>
<family>llama</family>
<quantization_level>Q4_K_M</quantization_level>
</details>
<modified_at>2025-01-20T00:39:23.4959669-08:00</modified_at>
</models>
</root>
• •The Show Model action shows information about a model including model details, model file, template, parameters, license, and a system prompt.
To use this action, specify the following input fields:
• - ShowModelRequestBody: The request body to get the model details.
• - model: The name of your deployed model. You must deploy a model before making calls.
The following snippet is a sample of the request body:
<ShowModelRequestBody>
<model>llama3.2:Latest</model>
</ShowModelRequestBody>