gptcli

gptcli is a command-line interface tool that leverages the GPT (Generative Pre-trained Transformer) model from OpenAI to generate text based on user-provided input. This tool simplifies text generation using the GPT model directly from the command line.

Installation

To use the gptcli tool, ensure you have Python installed on your system. You can install the tool via pip:

pip install git+https://github.com/ivandustin/gptcli.git

Usage

The basic syntax for utilizing the gptcli tool is as follows:

gpt instructions.txt input.txt > output.txt

You can also use - for any file parameters to read from stdin. For example:

cat input.txt | gpt instructions.txt - > README.md

Practical Example

Suppose you need to write a technical document.

1. Write a draft of your technical document in draft.txt.
2. Create an instructions file named tech-doc.txt with the following content:

Create a technical document.

Generate the document with the following command:

gpt tech-doc.txt draft.txt > README.md

OpenAI Model

Specify the OpenAI model to be used by setting the OPENAI_MODEL environment variable.

You can also use the --model argument to specify the OpenAI model directly in the command:

gpt instructions.txt input.txt --model gpt-4o > output.txt

OpenAI API Key

To authenticate with the OpenAI API, you need to specify your API key via the OPENAI_API_KEY environment variable.

Structured Output

For structured output, provide a JSON Schema in the --schema option:

gpt instructions.txt input.txt --schema schema.json > output.json

You can use Pydantic to generate a JSON Schema document.

GPT Function

You can import the gpt function in your project that allows you to generate text based on a model, temperature, instructions, and input.

Example

from gptcli import gpt

model = "gpt-4o"
temperature = 0.7
instructions = "Generate a creative story based on the following input."
input_text = "Once upon a time in a land far, far away..."

output = gpt(model, temperature, instructions, input_text)
print(output)

GPT Function (Structured Output)

Example

from gptcli.json import gpt
import json

model = "gpt-4o"
temperature = 0
instructions = "Get the age of Anne."
schema = {
    "properties": {
        "age": {
            "title": "Age",
            "type": "integer"
        }
    },
    "required": [
        "age"
    ],
    "title": "Schema",
    "type": "object"
}
input_text = "The age of Prince is 26. The age of Anne is 21. The age of Anna is 19."

output = gpt(model, temperature, instructions, schema, input_text)
print(json.dumps(output, indent=4))

Output

{
    "age": 21
}

Encoding Issues

If you encounter any character encoding issue, especially in Windows, you may need to set the PYTHONUTF8 environment variable to 1.