Is There a Case for Conversation Optimized Tokenizers in Large Language Models?

TL;DR

This paper investigates whether customizing tokenizers for chatbot conversations can improve efficiency, finding that conversation-optimized tokenizers reduce token counts in dialogues by 5-10%, potentially saving energy without harming training corpus performance.

Contribution

It introduces conversation-specific tokenizer optimization for chatbots and demonstrates its benefits in reducing token counts and energy consumption.

Findings

Conversation-optimized tokenizers reduce dialogue tokens by 5-10%.

Energy savings are achievable through specialized tokenization.

Minimal or positive impact on training corpus tokenization efficiency.

Abstract

The computational and energy costs of Large Language Models (LLMs) have increased exponentially driven by the growing model sizes and the massive adoption of LLMs by hundreds of millions of users. The unit cost of an LLM is the computation of a token. Therefore, the tokenizer plays an important role in the efficiency of a model, and they are carefully optimized to minimize the number of tokens for the text in their training corpus. One of the most popular applications of LLMs are chatbots that interact with users. A key observation is that, for those chatbots, what is important is the performance of the tokenizer in the user text input and the chatbot responses. Those are most likely different from the text in the training corpus. So, a question that immediately arises is whether there is a potential benefit in optimizing tokenizers for chatbot conversations. In this paper, this idea is explored for different tokenizers by using a publicly available corpus of chatbot conversations to redesign their vocabularies and evaluate their performance in this domain. The results show that conversation-optimized tokenizers consistently reduce the number of tokens in chatbot dialogues, which can lead to meaningful energy savings, in the range of 5% to 10% while having minimal or even slightly positive impact on tokenization efficiency for the original training corpus.