mirror of
https://github.com/songquanpeng/one-api.git
synced 2025-11-18 06:03:42 +08:00
🐛 fix: stream mode delay issue (#53)
This commit is contained in:
@@ -32,7 +32,7 @@ func (p *PalmProvider) CreateChatCompletion(request *types.ChatCompletionRequest
|
||||
return p.convertToChatOpenai(palmResponse, request)
|
||||
}
|
||||
|
||||
func (p *PalmProvider) CreateChatCompletionStream(request *types.ChatCompletionRequest) (requester.StreamReaderInterface[types.ChatCompletionStreamResponse], *types.OpenAIErrorWithStatusCode) {
|
||||
func (p *PalmProvider) CreateChatCompletionStream(request *types.ChatCompletionRequest) (requester.StreamReaderInterface[string], *types.OpenAIErrorWithStatusCode) {
|
||||
req, errWithCode := p.getChatRequest(request)
|
||||
if errWithCode != nil {
|
||||
return nil, errWithCode
|
||||
@@ -50,7 +50,7 @@ func (p *PalmProvider) CreateChatCompletionStream(request *types.ChatCompletionR
|
||||
Request: request,
|
||||
}
|
||||
|
||||
return requester.RequestStream[types.ChatCompletionStreamResponse](p.Requester, resp, chatHandler.handlerStream)
|
||||
return requester.RequestStream[string](p.Requester, resp, chatHandler.handlerStream)
|
||||
}
|
||||
|
||||
func (p *PalmProvider) getChatRequest(request *types.ChatCompletionRequest) (*http.Request, *types.OpenAIErrorWithStatusCode) {
|
||||
@@ -142,11 +142,11 @@ func convertFromChatOpenai(request *types.ChatCompletionRequest) *PaLMChatReques
|
||||
}
|
||||
|
||||
// 转换为OpenAI聊天流式请求体
|
||||
func (h *palmStreamHandler) handlerStream(rawLine *[]byte, isFinished *bool, response *[]types.ChatCompletionStreamResponse) error {
|
||||
func (h *palmStreamHandler) handlerStream(rawLine *[]byte, dataChan chan string, errChan chan error) {
|
||||
// 如果rawLine 前缀不为data:,则直接返回
|
||||
if !strings.HasPrefix(string(*rawLine), "data: ") {
|
||||
*rawLine = nil
|
||||
return nil
|
||||
return
|
||||
}
|
||||
|
||||
// 去除前缀
|
||||
@@ -155,19 +155,21 @@ func (h *palmStreamHandler) handlerStream(rawLine *[]byte, isFinished *bool, res
|
||||
var palmChatResponse PaLMChatResponse
|
||||
err := json.Unmarshal(*rawLine, &palmChatResponse)
|
||||
if err != nil {
|
||||
return common.ErrorToOpenAIError(err)
|
||||
errChan <- common.ErrorToOpenAIError(err)
|
||||
return
|
||||
}
|
||||
|
||||
error := errorHandle(&palmChatResponse.PaLMErrorResponse)
|
||||
if error != nil {
|
||||
return error
|
||||
errChan <- error
|
||||
return
|
||||
}
|
||||
|
||||
return h.convertToOpenaiStream(&palmChatResponse, response)
|
||||
h.convertToOpenaiStream(&palmChatResponse, dataChan, errChan)
|
||||
|
||||
}
|
||||
|
||||
func (h *palmStreamHandler) convertToOpenaiStream(palmChatResponse *PaLMChatResponse, response *[]types.ChatCompletionStreamResponse) error {
|
||||
func (h *palmStreamHandler) convertToOpenaiStream(palmChatResponse *PaLMChatResponse, dataChan chan string, errChan chan error) {
|
||||
var choice types.ChatCompletionStreamChoice
|
||||
if len(palmChatResponse.Candidates) > 0 {
|
||||
choice.Delta.Content = palmChatResponse.Candidates[0].Content
|
||||
@@ -182,10 +184,9 @@ func (h *palmStreamHandler) convertToOpenaiStream(palmChatResponse *PaLMChatResp
|
||||
Created: common.GetTimestamp(),
|
||||
}
|
||||
|
||||
*response = append(*response, streamResponse)
|
||||
responseBody, _ := json.Marshal(streamResponse)
|
||||
dataChan <- string(responseBody)
|
||||
|
||||
h.Usage.CompletionTokens += common.CountTokenText(palmChatResponse.Candidates[0].Content, h.Request.Model)
|
||||
h.Usage.TotalTokens = h.Usage.PromptTokens + h.Usage.CompletionTokens
|
||||
|
||||
return nil
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user