INNER CODE UNIT · Python

num_tokens_from_string

eugeneyan/obsidian-copilot · src/app.py:99

def num_tokens_from_string(string: str, model_name: str) -> int:
    """
    Returns the number of tokens in a string based on tiktoken encoding.

    Args:
        string: String to count tokens for
        model_name: Tokenizer model type

    Returns:
        Number of tokens in the string
    """
    encoding = tiktoken.encoding_for_model(model_name)
    num_tokens = len(encoding.encode(string))
    return num_tokens


def get_chunks_from_hits(
    hits: List[dict], model_name: str = "gpt-3.5-turbo", max_tokens: int = 3200

View source record →

📰 Research Paper
Loading…
⏳ Fetching content…