xinyin025/one-api
1
0
Fork 0
mirror of https://github.com/songquanpeng/one-api.git synced 2025-11-12 03:13:41 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'origin/upstream/main'

Browse Source
This commit is contained in:
Laisky.Cai
2024-03-15 09:49:49 +00:00
parent d953da1ff2 b204f6d82b
commit 41afad713e
49 changed files with 623 additions and 205 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
relay/controller/helper.go
View File

@@ -107,18 +107,18 @@ func getPromptTokens(textRequest *relaymodel.GeneralOpenAIRequest, relayMode int
return 0
}
func getPreConsumedQuota(textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64) int {
func getPreConsumedQuota(textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64) int64 {
preConsumedTokens := config.PreConsumedQuota
if textRequest.MaxTokens != 0 {
preConsumedTokens = promptTokens + textRequest.MaxTokens
preConsumedTokens = int64(promptTokens) + int64(textRequest.MaxTokens)
}
return int(float64(preConsumedTokens) * ratio)
return int64(float64(preConsumedTokens) * ratio)
}
func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64, meta *util.RelayMeta) (int, *relaymodel.ErrorWithStatusCode) {
func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIRequest, promptTokens int, ratio float64, meta *util.RelayMeta) (int64, *relaymodel.ErrorWithStatusCode) {
preConsumedQuota := getPreConsumedQuota(textRequest, promptTokens, ratio)
userQuota, err := model.CacheGetUserQuota(meta.UserId)
userQuota, err := model.CacheGetUserQuota(ctx, meta.UserId)
if err != nil {
return preConsumedQuota, openai.ErrorWrapper(err, "get_user_quota_failed", http.StatusInternalServerError)
}
@@ -144,16 +144,16 @@ func preConsumeQuota(ctx context.Context, textRequest *relaymodel.GeneralOpenAIR
return preConsumedQuota, nil
}
func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.RelayMeta, textRequest *relaymodel.GeneralOpenAIRequest, ratio float64, preConsumedQuota int, modelRatio float64, groupRatio float64) {
func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.RelayMeta, textRequest *relaymodel.GeneralOpenAIRequest, ratio float64, preConsumedQuota int64, modelRatio float64, groupRatio float64) {
if usage == nil {
logger.Error(ctx, "usage is nil, which is unexpected")
return
}
quota := 0
var quota int64
completionRatio := common.GetCompletionRatio(textRequest.Model)
promptTokens := usage.PromptTokens
completionTokens := usage.CompletionTokens
quota = int(math.Ceil((float64(promptTokens) + float64(completionTokens)*completionRatio) * ratio))
quota = int64(math.Ceil((float64(promptTokens) + float64(completionTokens)*completionRatio) * ratio))
if ratio != 0 && quota <= 0 {
quota = 1
}
@@ -168,7 +168,7 @@ func postConsumeQuota(ctx context.Context, usage *relaymodel.Usage, meta *util.R
if err != nil {
logger.Error(ctx, "error consuming token remain quota: "+err.Error())
}
err = model.CacheUpdateUserQuota(meta.UserId)
err = model.CacheUpdateUserQuota(ctx, meta.UserId)
if err != nil {
logger.Error(ctx, "error update user quota cache: "+err.Error())
}