Call Ollama for Text Generation — call_ollama • TextAnalysisR

Sends a prompt to Ollama and returns the generated text.

Usage

call_ollama(
  prompt,
  model = "phi3:mini",
  temperature = 0.3,
  max_tokens = 512,
  timeout = 60,
  verbose = FALSE
)

Arguments

prompt: Character string containing the prompt.
model: Character string specifying the Ollama model (default: "phi3:mini").
temperature: Numeric value controlling randomness (default: 0.3).
max_tokens: Maximum number of tokens to generate (default: 512).
timeout: Timeout in seconds for the request (default: 60).
verbose: Logical, if TRUE, prints progress messages.

Value

Character string with the generated text, or NULL if failed.

Examples

if (FALSE) { # \dontrun{
response <- call_ollama(
  prompt = "Summarize these keywords: machine learning, neural networks, AI",
  model = "phi3:mini"
)
print(response)
} # }