tokenizer VS tiktoken

How I wish this post had appeared a few days earlier... I am writing on my own library for some agent experiments (in go, to make my life more interesting I guess), and knowing the number of tokens is important to implement a token buffer memory (as you approach the model's context window size, you prune enough messages from the beginning of the conversation that the whole thing keeps some given size, in tokens). While there's a nice native library in go for OpenAI models (https://github.com/tiktoken-go/tokenizer), the only library I found for Hugging Face models (and Claude, they published their tokenizer spec in the same JSON format) calls into HF's Rust implementation, which makes it challenging as a dependency in Go. What is more, any tokenizer needs to keep some representation of its vocabulary in memory. So, in the end I removed the true tokenizers, and ended up using an approximate version (just split it in on spaces and multiply by a factor I determined experimentally for the models I use using the real tokenizer, with a little extra for safety). If it turns out someone needs the real thing they can always provide their own token counter). I was actually rather happy with this result: I have less dependencies, and use less memory. But to get there I needed to do a deep dive too understand BPE tokenizers :)

(The library, if anyone is interested: https://github.com/ryszard/agency.)

Their tokenizer is open source: https://github.com/openai/tiktoken

Data files that contain vocabulary are listed here: https://github.com/openai/tiktoken/blob/9e79899bc248d5313c7d...

OpenAI's refference tokeniser - https://github.com/openai/tiktoken

ChatGPT presumably adds them as special tokens to the cl100k_base tokenizer, as they demo in the tiktoken documentation: https://github.com/openai/tiktoken#extending-tiktoken

I want to measure the approximate number of tokens in a piece of text to understand if I will need to modify it before passing it into the context of an OpenAI API call. Tiktoken can do this, but I'm not sure if it's overkill to use that library just for this simple task. I don't need to actually tokenize the text, I just need an approximate count (e.g. within like 1% of the text's actual token length for text that represents the visible text on a webpage).

https://platform.openai.com/tokenizer or the official python library tiktoken https://github.com/openai/tiktoken or this JS port of tiktoken https://github.com/dqbd/tiktoken

It's built on top of the tiktoken library and is basically just a lambda function in the backend.

"All" is a single token in OpenAI's tiktoken Tokenizer, unrelated to the token for capital "A". Even lowercase "all" is a distinct token from "All" or "ALL."

check tiktoken