370
edits
Tokens: Difference between revisions
→Tokenization Tools
No edit summary |
|||
| (6 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
==Introduction== | ==Introduction== | ||
Tokens are fragments of words, which may include trailing spaces or sub-words. | [[Tokens]] are fragments of words, which may include trailing spaces or sub-words. Before [[natural language processing]] (NLP) systems, like [[large language models]] (LLMs), process a [[prompt]], the prompt is transformed into tokens. The way words are broken down into tokens is language-dependent, which can affect the implementation cost of the API for languages other than English. | ||
==Understanding Token Lengths== | ==Understanding Token Lengths== | ||
To comprehend the idea of tokens, consider these general approximations concerning token lengths: | To comprehend the idea of tokens, consider these general approximations concerning token lengths: | ||
1 token ≈ 4 characters in English | *1 token ≈ 4 characters in English | ||
1 token ≈ ¾ words | *1 token ≈ ¾ words | ||
100 tokens ≈ 75 words | *100 tokens ≈ 75 words | ||
Regarding sentences and paragraphs: | Regarding sentences and paragraphs: | ||
1-2 sentences ≈ 30 tokens | *1-2 sentences ≈ 30 tokens | ||
1 paragraph ≈ 100 tokens | *1 paragraph ≈ 100 tokens | ||
1,500 words ≈ 2048 tokens | *1,500 words ≈ 2048 tokens | ||
For more context, consider these examples: | For more context, consider these examples: | ||
Wayne Gretzky’s quote "You miss 100% of the shots you don't take" contains 11 tokens. | *Wayne Gretzky’s quote "You miss 100% of the shots you don't take" contains 11 tokens. | ||
OpenAI’s charter contains 476 tokens. | *OpenAI’s charter contains 476 tokens. | ||
The transcript of the US Declaration of Independence contains 1,695 tokens. | *The transcript of the US Declaration of Independence contains 1,695 tokens. | ||
==Tokenization Tools== | ==Tokenization Tools== | ||
To delve deeper into tokenization, the following tools and libraries are available: | To delve deeper into tokenization, the following tools and libraries are available: | ||
*[ | *[https://platform.openai.com/tokenizer OpenAI's interactive Tokenizer tool] | ||
*[ | *[https://github.com/openai/tiktoken Tiktoken], a fast BPE tokenizer specifically for OpenAI models | ||
*Transformers package for Python | *[[Transformers]] package for Python | ||
*gpt-3-encoder package for node.js | *[https://www.npmjs.com/package/gpt-3-encoder gpt-3-encoder package for node.js] | ||
==Token Limits== | |||
The fragment limit for requests is contingent on the [[model]] employed. | |||
===OpenAI API Token Limit=== | |||
OpenAI API has a maximum of 4097 tokens shared between the prompt and its completion. If a prompt consists of 4000 tokens, the completion can have a maximum of 97 tokens. This limitation is a technical constraint, but there are strategies to work within it, such as shortening prompts or dividing text into smaller sections. | |||
==Token Pricing== | ==Token Pricing== | ||
| Line 40: | Line 45: | ||
*Uppercase at the beginning of a sentence: "Red" (token: "7738") | *Uppercase at the beginning of a sentence: "Red" (token: "7738") | ||
*The more likely or common a token is, the lower the token number assigned to it. For example, the token for the period ("13") remains consistent in all three sentences because its usage is similar throughout the corpus data. | *The more likely or common a token is, the lower the token number assigned to it. For example, the token for the period ("13") remains consistent in all three sentences because its usage is similar throughout the corpus data. | ||
<gallery mode=packed> | |||
File:tokens3.png | |||
File:tokens2.png | |||
File:tokens1.png | |||
</gallery> | |||
==Prompt Design and Token Knowledge== | ==Prompt Design and Token Knowledge== | ||
| Line 48: | Line 59: | ||
===Using the logit_bias Parameter=== | ===Using the logit_bias Parameter=== | ||
The logit_bias parameter enables setting biases for specific tokens, altering the probability of those tokens appearing in the completion. For example, if designing an AI Baking Assistant that takes users' egg allergies into account, the logit_bias parameter can be used to deter the model from generating responses that include any form of the word "egg". | The [[logit_bias]] [[parameter]] enables setting biases for specific tokens, altering the probability of those tokens appearing in the completion. For example, if designing an AI Baking Assistant that takes users' egg allergies into account, the logit_bias parameter can be used to deter the model from generating responses that include any form of the word "egg". | ||
First, use a tokenizer tool to identify the tokens for which biases should be set: | First, use a tokenizer tool to identify the tokens for which biases should be set: | ||
| Line 55: | Line 66: | ||
Plural with trailing space: " eggs" (token: "9653") | Plural with trailing space: " eggs" (token: "9653") | ||
Subword token generated for "Egg" or "Eggs": "gg" (token: "1130") | Subword token generated for "Egg" or "Eggs": "gg" (token: "1130") | ||
The logit_bias parameter accepts bias values ranging from -100 to +100. Extreme values result in either prohibiting (-100) or exclusively selecting (100) the related token. | The logit_bias parameter accepts bias values ranging from -100 to +100. Extreme values result in either prohibiting (-100) or exclusively selecting (100) the related token. | ||
By incorporating the logit biases into the prompt, the likelihood of the word "egg" (and its variations) appearing in the response for a banana bread recipe is diminished. Consequently, the AI Baking Assistant generates a response that excludes eggs, meeting its requirement of being mindful of the user's egg allergies. | By incorporating the logit biases into the prompt, the likelihood of the word "egg" (and its variations) appearing in the response for a banana bread recipe is diminished. Consequently, the AI Baking Assistant generates a response that excludes eggs, meeting its requirement of being mindful of the user's egg allergies. | ||
[[Category:Terms]] [[Category:Artificial intelligence terms]] | |||