feat: support Ali stable-diffusion-xl and wanx-v1 model (#1240)

* Fix ali ConvertRequest function to use baidu keyword * Support Ali stable-diffusion-xl and wanx-v1 model * Support Ali stable-diffusion-xl and wanx-v1 model * Support Ali stable-diffusion-xl and wanx-v1 model * chore: update ali constants and model ratio --------- Co-authored-by: JustSong <[email protected]> Co-authored-by: JustSong <[email protected]>
songquanpeng · Apr 5, 2024 · d4347e7 · d4347e7
1 parent b50b43e
commit d4347e7
Show file tree

Hide file tree

Showing 22 changed files with 489 additions and 87 deletions.
diff --git a/common/model-ratio.go b/common/model-ratio.go
@@ -104,11 +104,15 @@ var ModelRatio = map[string]float64{
 	"chatglm_pro":               0.7143, // ￥0.01 / 1k tokens
 	"chatglm_std":               0.3572, // ￥0.005 / 1k tokens
 	"chatglm_lite":              0.1429, // ￥0.002 / 1k tokens
-	"qwen-turbo":                0.5715, // ￥0.008 / 1k tokens  // https://help.aliyun.com/zh/dashscope/developer-reference/tongyi-thousand-questions-metering-and-billing
+	// https://help.aliyun.com/zh/dashscope/developer-reference/tongyi-thousand-questions-metering-and-billing
+	"qwen-turbo":                0.5715, // ￥0.008 / 1k tokens
 	"qwen-plus":                 1.4286, // ￥0.02 / 1k tokens
 	"qwen-max":                  1.4286, // ￥0.02 / 1k tokens
 	"qwen-max-longcontext":      1.4286, // ￥0.02 / 1k tokens
 	"text-embedding-v1":         0.05,   // ￥0.0007 / 1k tokens
+	"ali-stable-diffusion-xl":   8,
+	"ali-stable-diffusion-v1.5": 8,
+	"wanx-v1":                   8,
 	"SparkDesk":                 1.2858, // ￥0.018 / 1k tokens
 	"SparkDesk-v1.1":            1.2858, // ￥0.018 / 1k tokens
 	"SparkDesk-v2.1":            1.2858, // ￥0.018 / 1k tokens

diff --git a/relay/channel/aiproxy/adaptor.go b/relay/channel/aiproxy/adaptor.go
@@ -38,6 +38,13 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	return aiProxyLibraryRequest, nil
 }
 
+func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+	return request, nil
+}
+
 func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
 	return channel.DoRequestHelper(a, c, meta, requestBody)
 }

diff --git a/relay/channel/ali/adaptor.go b/relay/channel/ali/adaptor.go
@@ -23,21 +23,29 @@ func (a *Adaptor) Init(meta *util.RelayMeta) {
 }
 
 func (a *Adaptor) GetRequestURL(meta *util.RelayMeta) (string, error) {
-	fullRequestURL := fmt.Sprintf("%s/api/v1/services/aigc/text-generation/generation", meta.BaseURL)
-	if meta.Mode == constant.RelayModeEmbeddings {
+	fullRequestURL := ""
+	switch meta.Mode {
+	case constant.RelayModeEmbeddings:
 		fullRequestURL = fmt.Sprintf("%s/api/v1/services/embeddings/text-embedding/text-embedding", meta.BaseURL)
+	case constant.RelayModeImagesGenerations:
+		fullRequestURL = fmt.Sprintf("%s/api/v1/services/aigc/text2image/image-synthesis", meta.BaseURL)
+	default:
+		fullRequestURL = fmt.Sprintf("%s/api/v1/services/aigc/text-generation/generation", meta.BaseURL)
 	}
+
 	return fullRequestURL, nil
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Request, meta *util.RelayMeta) error {
 	channel.SetupCommonRequestHeader(c, req, meta)
 	if meta.IsStream {
 		req.Header.Set("Accept", "text/event-stream")
+		req.Header.Set("X-DashScope-SSE", "enable")
 	}
 	req.Header.Set("Authorization", "Bearer "+meta.APIKey)
-	if meta.IsStream {
-		req.Header.Set("X-DashScope-SSE", "enable")
+
+	if meta.Mode == constant.RelayModeImagesGenerations {
+		req.Header.Set("X-DashScope-Async", "enable")
 	}
 	if c.GetString(common.ConfigKeyPlugin) != "" {
 		req.Header.Set("X-DashScope-Plugin", c.GetString(common.ConfigKeyPlugin))
@@ -51,14 +59,23 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, relayMode int, request *model.G
 	}
 	switch relayMode {
 	case constant.RelayModeEmbeddings:
-		baiduEmbeddingRequest := ConvertEmbeddingRequest(*request)
-		return baiduEmbeddingRequest, nil
+		aliEmbeddingRequest := ConvertEmbeddingRequest(*request)
+		return aliEmbeddingRequest, nil
 	default:
-		baiduRequest := ConvertRequest(*request)
-		return baiduRequest, nil
+		aliRequest := ConvertRequest(*request)
+		return aliRequest, nil
 	}
 }
 
+func (a *Adaptor) ConvertImageRequest(request *model.ImageRequest) (any, error) {
+	if request == nil {
+		return nil, errors.New("request is nil")
+	}
+
+	aliRequest := ConvertImageRequest(*request)
+	return aliRequest, nil
+}
+
 func (a *Adaptor) DoRequest(c *gin.Context, meta *util.RelayMeta, requestBody io.Reader) (*http.Response, error) {
 	return channel.DoRequestHelper(a, c, meta, requestBody)
 }
@@ -70,6 +87,8 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, meta *util.Rel
 		switch meta.Mode {
 		case constant.RelayModeEmbeddings:
 			err, usage = EmbeddingHandler(c, resp)
+		case constant.RelayModeImagesGenerations:
+			err, usage = ImageHandler(c, resp)
 		default:
 			err, usage = Handler(c, resp)
 		}

diff --git a/relay/channel/ali/constants.go b/relay/channel/ali/constants.go
@@ -3,4 +3,5 @@ package ali
 var ModelList = []string{
 	"qwen-turbo", "qwen-plus", "qwen-max", "qwen-max-longcontext",
 	"text-embedding-v1",
+	"ali-stable-diffusion-xl", "ali-stable-diffusion-v1.5", "wanx-v1",
 }
diff --git a/relay/channel/ali/main.go b/relay/channel/ali/main.go
@@ -2,7 +2,10 @@ package ali
 
 import (
 	"bufio"
+	"encoding/base64"
 	"encoding/json"
+	"errors"
+	"fmt"
 	"github.com/gin-gonic/gin"
 	"github.com/songquanpeng/one-api/common"
 	"github.com/songquanpeng/one-api/common/helper"
@@ -12,6 +15,7 @@ import (
 	"io"
 	"net/http"
 	"strings"
+	"time"
 )
 
 // https://help.aliyun.com/document_detail/613695.html?spm=a2c4g.2399480.0.0.1adb778fAdzP9w#341800c0f8w0r
@@ -66,6 +70,17 @@ func ConvertEmbeddingRequest(request model.GeneralOpenAIRequest) *EmbeddingReque
 	}
 }
 
+func ConvertImageRequest(request model.ImageRequest) *ImageRequest {
+	var imageRequest ImageRequest
+	imageRequest.Input.Prompt = request.Prompt
+	imageRequest.Model = request.Model
+	imageRequest.Parameters.Size = strings.Replace(request.Size, "x", "*", -1)
+	imageRequest.Parameters.N = request.N
+	imageRequest.ResponseFormat = request.ResponseFormat
+
+	return &imageRequest
+}
+
 func EmbeddingHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
 	var aliResponse EmbeddingResponse
 	err := json.NewDecoder(resp.Body).Decode(&aliResponse)
@@ -265,3 +280,173 @@ func Handler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *
 	_, err = c.Writer.Write(jsonResponse)
 	return nil, &fullTextResponse.Usage
 }
+
+func ImageHandler(c *gin.Context, resp *http.Response) (*model.ErrorWithStatusCode, *model.Usage) {
+	apiKey := c.Request.Header.Get("Authorization")
+	apiKey = strings.TrimPrefix(apiKey, "Bearer ")
+	responseFormat := c.GetString("response_format")
+
+	var aliTaskResponse TaskResponse
+	responseBody, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return openai.ErrorWrapper(err, "read_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = resp.Body.Close()
+	if err != nil {
+		return openai.ErrorWrapper(err, "close_response_body_failed", http.StatusInternalServerError), nil
+	}
+	err = json.Unmarshal(responseBody, &aliTaskResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "unmarshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+
+	if aliTaskResponse.Message != "" {
+		logger.SysError("aliAsyncTask err: " + string(responseBody))
+		return openai.ErrorWrapper(errors.New(aliTaskResponse.Message), "ali_async_task_failed", http.StatusInternalServerError), nil
+	}
+
+	aliResponse, err, _ := asyncTaskWait(aliTaskResponse.Output.TaskId, apiKey)
+	if err != nil {
+		return openai.ErrorWrapper(err, "ali_async_task_wait_failed", http.StatusInternalServerError), nil
+	}
+
+	if aliResponse.Output.TaskStatus != "SUCCEEDED" {
+		return &model.ErrorWithStatusCode{
+			Error: model.Error{
+				Message: aliResponse.Output.Message,
+				Type:    "ali_error",
+				Param:   "",
+				Code:    aliResponse.Output.Code,
+			},
+			StatusCode: resp.StatusCode,
+		}, nil
+	}
+
+	fullTextResponse := responseAli2OpenAIImage(aliResponse, responseFormat)
+	jsonResponse, err := json.Marshal(fullTextResponse)
+	if err != nil {
+		return openai.ErrorWrapper(err, "marshal_response_body_failed", http.StatusInternalServerError), nil
+	}
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, err = c.Writer.Write(jsonResponse)
+	return nil, &fullTextResponse.Usage
+}
+
+func asyncTask(taskID string, key string) (*TaskResponse, error, []byte) {
+	url := fmt.Sprintf("https://dashscope.aliyuncs.com/api/v1/tasks/%s", taskID)
+
+	var aliResponse TaskResponse
+
+	req, err := http.NewRequest("GET", url, nil)
+	if err != nil {
+		return &aliResponse, err, nil
+	}
+
+	req.Header.Set("Authorization", "Bearer "+key)
+
+	client := &http.Client{}
+	resp, err := client.Do(req)
+	if err != nil {
+		logger.SysError("aliAsyncTask client.Do err: " + err.Error())
+		return &aliResponse, err, nil
+	}
+	defer resp.Body.Close()
+
+	responseBody, err := io.ReadAll(resp.Body)
+
+	var response TaskResponse
+	err = json.Unmarshal(responseBody, &response)
+	if err != nil {
+		logger.SysError("aliAsyncTask NewDecoder err: " + err.Error())
+		return &aliResponse, err, nil
+	}
+
+	return &response, nil, responseBody
+}
+
+func asyncTaskWait(taskID string, key string) (*TaskResponse, error, []byte) {
+	waitSeconds := 2
+	step := 0
+
+	var taskResponse TaskResponse
+	var responseBody []byte
+
+	for {
+		step++
+		rsp, err, body := asyncTask(taskID, key)
+		responseBody = body
+		if err != nil {
+			return &taskResponse, err, responseBody
+		}
+
+		if rsp.Output.TaskStatus == "" {
+			return &taskResponse, nil, responseBody
+		}
+
+		switch rsp.Output.TaskStatus {
+		case "FAILED":
+			fallthrough
+		case "CANCELED":
+			fallthrough
+		case "SUCCEEDED":
+			fallthrough
+		case "UNKNOWN":
+			return rsp, nil, responseBody
+		}
+
+		time.Sleep(time.Duration(waitSeconds) * time.Second)
+	}
+
+	return &taskResponse, nil, responseBody
+}
+
+func responseAli2OpenAIImage(response *TaskResponse, responseFormat string) *openai.ImageResponse {
+	imageRespones := openai.ImageResponse{}
+
+	for _, data := range response.Output.Results {
+		var b64Json string
+		if responseFormat == "b64_json" {
+			// 读取 data.Url 的图片数据并转存到 b64Json
+			imageData, err := getImageData(data.Url)
+			if err != nil {
+				// 处理获取图片数据失败的情况
+				logger.SysError("getImageData Error getting image data: " + err.Error())
+				continue
+			}
+
+			// 将图片数据转为 Base64 编码的字符串
+			b64Json = Base64Encode(imageData)
+		} else {
+			// 如果 responseFormat 不是 "b64_json"，则直接使用 data.B64Image
+			b64Json = data.B64Image
+		}
+
+		imageRespones.Data = append(imageRespones.Data, openai.ImageData{
+			Url:           data.Url,
+			B64Json:       b64Json,
+			RevisedPrompt: "",
+		})
+	}
+	return &imageRespones
+}
+
+func getImageData(url string) ([]byte, error) {
+	response, err := http.Get(url)
+	if err != nil {
+		return nil, err
+	}
+	defer response.Body.Close()
+
+	imageData, err := io.ReadAll(response.Body)
+	if err != nil {
+		return nil, err
+	}
+
+	return imageData, nil
+}
+
+func Base64Encode(data []byte) string {
+	b64Json := base64.StdEncoding.EncodeToString(data)
+	return b64Json
+}
diff --git a/relay/channel/ali/model.go b/relay/channel/ali/model.go
@@ -33,6 +33,79 @@ type ChatRequest struct {
 	Parameters Parameters `json:"parameters,omitempty"`
 }
 
+type ImageRequest struct {
+	Model string `json:"model"`
+	Input struct {
+		Prompt         string `json:"prompt"`
+		NegativePrompt string `json:"negative_prompt,omitempty"`
+	} `json:"input"`
+	Parameters struct {
+		Size  string `json:"size,omitempty"`
+		N     int    `json:"n,omitempty"`
+		Steps string `json:"steps,omitempty"`
+		Scale string `json:"scale,omitempty"`
+	} `json:"parameters,omitempty"`
+	ResponseFormat string `json:"response_format,omitempty"`
+}
+
+type TaskResponse struct {
+	StatusCode int    `json:"status_code,omitempty"`
+	RequestId  string `json:"request_id,omitempty"`
+	Code       string `json:"code,omitempty"`
+	Message    string `json:"message,omitempty"`
+	Output     struct {
+		TaskId     string `json:"task_id,omitempty"`
+		TaskStatus string `json:"task_status,omitempty"`
+		Code       string `json:"code,omitempty"`
+		Message    string `json:"message,omitempty"`
+		Results    []struct {
+			B64Image string `json:"b64_image,omitempty"`
+			Url      string `json:"url,omitempty"`
+			Code     string `json:"code,omitempty"`
+			Message  string `json:"message,omitempty"`
+		} `json:"results,omitempty"`
+		TaskMetrics struct {
+			Total     int `json:"TOTAL,omitempty"`
+			Succeeded int `json:"SUCCEEDED,omitempty"`
+			Failed    int `json:"FAILED,omitempty"`
+		} `json:"task_metrics,omitempty"`
+	} `json:"output,omitempty"`
+	Usage Usage `json:"usage"`
+}
+
+type Header struct {
+	Action       string `json:"action,omitempty"`
+	Streaming    string `json:"streaming,omitempty"`
+	TaskID       string `json:"task_id,omitempty"`
+	Event        string `json:"event,omitempty"`
+	ErrorCode    string `json:"error_code,omitempty"`
+	ErrorMessage string `json:"error_message,omitempty"`
+	Attributes   any    `json:"attributes,omitempty"`
+}
+
+type Payload struct {
+	Model      string `json:"model,omitempty"`
+	Task       string `json:"task,omitempty"`
+	TaskGroup  string `json:"task_group,omitempty"`
+	Function   string `json:"function,omitempty"`
+	Parameters struct {
+		SampleRate int     `json:"sample_rate,omitempty"`
+		Rate       float64 `json:"rate,omitempty"`
+		Format     string  `json:"format,omitempty"`
+	} `json:"parameters,omitempty"`
+	Input struct {
+		Text string `json:"text,omitempty"`
+	} `json:"input,omitempty"`
+	Usage struct {
+		Characters int `json:"characters,omitempty"`
+	} `json:"usage,omitempty"`
+}
+
+type WSSMessage struct {
+	Header  Header  `json:"header,omitempty"`
+	Payload Payload `json:"payload,omitempty"`
+}
+
 type EmbeddingRequest struct {
 	Model string `json:"model"`
 	Input struct {