one-api/common/token.go

package common

import (
	"errors"
	"fmt"
	"math"
	"strings"

	"one-api/common/image"
	"one-api/types"

	"github.com/MartialBE/tiktoken-go"
	"github.com/spf13/viper"
)

var tokenEncoderMap = map[string]*tiktoken.Tiktoken{}
var gpt35TokenEncoder *tiktoken.Tiktoken
var gpt4TokenEncoder *tiktoken.Tiktoken
var gpt4oTokenEncoder *tiktoken.Tiktoken

func InitTokenEncoders() {
	if viper.GetBool("disable_token_encoders") {
		DISABLE_TOKEN_ENCODERS = true
		SysLog("token encoders disabled")
		return
	}
	SysLog("initializing token encoders")
	var err error
	gpt35TokenEncoder, err = tiktoken.EncodingForModel("gpt-3.5-turbo")
	if err != nil {
		FatalLog(fmt.Sprintf("failed to get gpt-3.5-turbo token encoder: %s", err.Error()))
	}

	gpt4TokenEncoder, err = tiktoken.EncodingForModel("gpt-4")
	if err != nil {
		FatalLog(fmt.Sprintf("failed to get gpt-4 token encoder: %s", err.Error()))
	}

	gpt4oTokenEncoder, err = tiktoken.EncodingForModel("gpt-4o")
	if err != nil {
		FatalLog(fmt.Sprintf("failed to get gpt-4o token encoder: %s", err.Error()))
	}

	SysLog("token encoders initialized")
}

func getTokenEncoder(model string) *tiktoken.Tiktoken {
	if DISABLE_TOKEN_ENCODERS {
		return nil
	}

	tokenEncoder, ok := tokenEncoderMap[model]
	if ok {
		return tokenEncoder
	}

	if strings.HasPrefix(model, "gpt-3.5") {
		tokenEncoder = gpt35TokenEncoder
	} else if strings.HasPrefix(model, "gpt-4o") {
		tokenEncoder = gpt4oTokenEncoder
	} else if strings.HasPrefix(model, "gpt-4") {
		tokenEncoder = gpt4TokenEncoder
	} else {
		var err error
		tokenEncoder, err = tiktoken.EncodingForModel(model)
		if err != nil {
			SysError(fmt.Sprintf("failed to get token encoder for model %s: %s, using encoder for gpt-3.5-turbo", model, err.Error()))
			tokenEncoder = gpt35TokenEncoder
		}
	}

	tokenEncoderMap[model] = tokenEncoder
	return tokenEncoder
}

func getTokenNum(tokenEncoder *tiktoken.Tiktoken, text string) int {
	if DISABLE_TOKEN_ENCODERS || ApproximateTokenEnabled {
		return int(float64(len(text)) * 0.38)
	}
	return len(tokenEncoder.Encode(text, nil, nil))
}

func CountTokenMessages(messages []types.ChatCompletionMessage, model string) int {
	tokenEncoder := getTokenEncoder(model)
	// Reference:
	// https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb
	// https://github.com/pkoukk/tiktoken-go/issues/6
	//
	// Every message follows <|start|>{role/name}\n{content}<|end|>\n
	var tokensPerMessage int
	var tokensPerName int
	if model == "gpt-3.5-turbo-0301" {
		tokensPerMessage = 4
		tokensPerName = -1 // If there's a name, the role is omitted
	} else {
		tokensPerMessage = 3
		tokensPerName = 1
	}
	tokenNum := 0
	for _, message := range messages {
		tokenNum += tokensPerMessage
		switch v := message.Content.(type) {
		case string:
			tokenNum += getTokenNum(tokenEncoder, v)
		case []any:
			for _, it := range v {
				m := it.(map[string]any)
				switch m["type"] {
				case "text":
					tokenNum += getTokenNum(tokenEncoder, m["text"].(string))
				case "image_url":
					imageUrl, ok := m["image_url"].(map[string]any)
					if ok {
						url := imageUrl["url"].(string)
						detail := ""
						if imageUrl["detail"] != nil {
							detail = imageUrl["detail"].(string)
						}
						imageTokens, err := countImageTokens(url, detail)
						if err != nil {
							//Due to the excessive length of the error information, only extract and record the most critical part.
							SysError("error counting image tokens: " + err.Error())
						} else {
							tokenNum += imageTokens
						}
					}
				}
			}
		}
		tokenNum += getTokenNum(tokenEncoder, message.Role)
		if message.Name != nil {
			tokenNum += tokensPerName
			tokenNum += getTokenNum(tokenEncoder, *message.Name)
		}
	}
	tokenNum += 3 // Every reply is primed with <|start|>assistant<|message|>
	return tokenNum
}

const (
	lowDetailCost         = 85
	highDetailCostPerTile = 170
	additionalCost        = 85
)

// https://platform.openai.com/docs/guides/vision/calculating-costs
// https://github.com/openai/openai-cookbook/blob/05e3f9be4c7a2ae7ecf029a7c32065b024730ebe/examples/How_to_count_tokens_with_tiktoken.ipynb
func countImageTokens(url string, detail string) (_ int, err error) {
	var fetchSize = true
	var width, height int
	// Reference: https://platform.openai.com/docs/guides/vision/low-or-high-fidelity-image-understanding
	// detail == "auto" is undocumented on how it works, it just said the model will use the auto setting which will look at the image input size and decide if it should use the low or high setting.
	// According to the official guide, "low" disable the high-res model,
	// and only receive low-res 512px x 512px version of the image, indicating
	// that image is treated as low-res when size is smaller than 512px x 512px,
	// then we can assume that image size larger than 512px x 512px is treated
	// as high-res. Then we have the following logic:
	// if detail == "" || detail == "auto" {
	// 	width, height, err = image.GetImageSize(url)
	// 	if err != nil {
	// 		return 0, err
	// 	}
	// 	fetchSize = false
	// 	// not sure if this is correct
	// 	if width > 512 || height > 512 {
	// 		detail = "high"
	// 	} else {
	// 		detail = "low"
	// 	}
	// }

	// However, in my test, it seems to be always the same as "high".
	// The following image, which is 125x50, is still treated as high-res, taken
	// 255 tokens in the response of non-stream chat completion api.
	// https://upload.wikimedia.org/wikipedia/commons/1/10/18_Infantry_Division_Messina.jpg
	if detail == "" || detail == "auto" {
		// assume by test, not sure if this is correct
		detail = "high"
	}
	switch detail {
	case "low":
		return lowDetailCost, nil
	case "high":
		if fetchSize {
			width, height, err = image.GetImageSize(url)
			if err != nil {
				return 0, err
			}
		}
		if width > 2048 || height > 2048 { // max(width, height) > 2048
			ratio := float64(2048) / math.Max(float64(width), float64(height))
			width = int(float64(width) * ratio)
			height = int(float64(height) * ratio)
		}
		if width > 768 && height > 768 { // min(width, height) > 768
			ratio := float64(768) / math.Min(float64(width), float64(height))
			width = int(float64(width) * ratio)
			height = int(float64(height) * ratio)
		}
		numSquares := int(math.Ceil(float64(width)/512) * math.Ceil(float64(height)/512))
		result := numSquares*highDetailCostPerTile + additionalCost
		return result, nil
	default:
		return 0, errors.New("invalid detail option")
	}
}

func CountTokenInput(input any, model string) int {
	switch v := input.(type) {
	case string:
		return CountTokenText(v, model)
	case []string:
		text := ""
		for _, s := range v {
			text += s
		}
		return CountTokenText(text, model)
	}
	return 0
}

func CountTokenText(text string, model string) int {
	tokenEncoder := getTokenEncoder(model)
	return getTokenNum(tokenEncoder, text)
}

func CountTokenImage(input interface{}) (int, error) {
	switch v := input.(type) {
	case types.ImageRequest:
		// 处理 ImageRequest
		return calculateToken(v.Model, v.Size, v.N, v.Quality)
	case types.ImageEditRequest:
		// 处理 ImageEditsRequest
		return calculateToken(v.Model, v.Size, v.N, "")
	default:
		return 0, errors.New("unsupported type")
	}
}

func calculateToken(model string, size string, n int, quality string) (int, error) {
	imageCostRatio, hasValidSize := DalleSizeRatios[model][size]

	if hasValidSize {
		if quality == "hd" && model == "dall-e-3" {
			if size == "1024x1024" {
				imageCostRatio *= 2
			} else {
				imageCostRatio *= 1.5
			}
		}
	} else {
		imageCostRatio = 1
		// return 0, errors.New("size not supported for this image model")
	}

	return int(imageCostRatio*1000) * n, nil
}