Request Parameters

This guide provides an overview of the Completion endpoint in Konko's API, designed for generating text based on a single input prompt.

Model Parameter

Selecting the appropriate model is crucial for the quality and relevance of the output. Each model specializes in different tasks. For model options, visit: List of Available Models.

Prompt Parameter

The prompt to generate completions for, encoded as a string.

Advanced Parameters

  • max_tokens: Sets the maximum length for the generated text.
  • temperature: Adjusts randomness in the response (0 to 1). Higher values yield more varied outputs.
  • top_p: Controls the breadth of possible responses. Higher values lead to more diverse results.
  • n: Determines the number of completions to generate per prompt.
  • stop: Sequences signaling the model to stop generating text.
  • presence_penalty/frequency_penalty: Modifies token generation based on their frequency or presence.
  • logit_bias: Influences the likelihood of specific tokens appearing in the completion.
  • seed: For deterministic outputs, ensuring repeat requests yield the same result.
  • stream: Enables real-time streaming of the generated text.

Example SDK Usage

import konko

response = konko.completions.create(
    model="meta-llama/llama-2-13b",
    prompt="Summarize the Foundation by Isaac Asimov",
  	max_tokens=500
    
)
print(response)

Output

Shows the structure of the generated text based on the provided prompt and parameters.

Completion(id='82c4fe31-4d48-487c-a409-1aedb2738ac3', 
           choices=[
  									CompletionChoice
                      (
                        finish_reason=None,
                        index=0, 
                        logprobs=None, 
                        text='\nThe Foundation by Isaac Asimov is a science fiction novel that tells the story of a group of scientists who are trying to save humanity from extinction. The novel is set in the future, and it follows the story of Hari Seldon, a scientist who has developed a new field of study called psychohistory. Psychohistory is a way of predicting the future by studying the behavior of large groups of people.\nSeldon uses psychohistory to predict that the Galactic Empire will fall, and he creates a plan to save humanity from the chaos that will follow. The Foundation is a group of scientists who are working to preserve human knowledge and culture in the face of the Empire’s collapse.\nThe Foundation is a classic science fiction novel that is still relevant today. The novel explores themes of power, corruption, and the importance of knowledge. The Foundation is a must-read for any fan of science fiction.\nThe Foundation by Isaac Asimov is a science fiction novel that tells the story of a group of scientists who are trying to save humanity from extinction. The novel is set in the future, and it follows the story of Hari Seldon, a scientist who has developed a new field of study called psychohistory. Psychohistory is a way of predicting the future by studying the behavior of large groups of people.\nThe Foundation is a classic science fiction novel that is still relevant today. The novel explores themes of power, corruption, and the importance of knowledge. The Foundation is a must-read for any fan of science fiction. The Foundation by Isaac Asimov is a science fiction novel that tells the story of a group of scientists who are trying to save humanity from extinction. The novel is set in the future, and it follows the story of Hari Seldon, a scientist who has developed a new field of study called psychohistory. Psychohistory is a way of predicting the future by studying the behavior of large groups of people.\nThe Foundation is a classic science fiction novel that is still relevant today. The novel explores themes of power, corruption, and the importance of knowledge. The Foundation is a must-read for any fan of science fiction. The Foundation by Isaac Asimov is a science fiction novel that tells the story of a group of scientists who are trying'
                      )
                   ], 
           created=1704895437, 
           model='meta-llama/llama-2-13b', 
           object='text_completion', 
           system_fingerprint=None,
           usage=None
          )