fix name of usage.prompt_tokens_details (#559)

Signed-off-by: Maya Barnea <mayab@il.ibm.com>

Author: mayabar

docs/http-enpdpoints.md

@@ -75,7 +75,7 @@ Structure of requests/responses
           - prompt_tokens
           - completion_tokens
           - total_tokens
-          - prompt_tokens_detail
+          - prompt_tokens_details
             - cached_tokens
         - object
         - kv_transfer_params

docs/kv-cache.md

@@ -216,7 +216,7 @@ When KV cache is enabled, the response `usage` object includes cached token info
     "prompt_tokens": 100,
     "completion_tokens": 50,
     "total_tokens": 150,
-    "prompt_tokens_detail": {
+    "prompt_tokens_details": {
       "cached_tokens": 64
     }
   }

pkg/api/response.go

@@ -79,12 +79,12 @@ type Usage struct {
 	CompletionTokens int `json:"completion_tokens"`
 	// TotalTokens is the total number of tokens processed for the request (the sum of the two values above)
 	TotalTokens int `json:"total_tokens"`
-	// PromptTokensDetail contains detailed token usage statistics for the prompt
-	PromptTokensDetail *PromptTokensDetail `json:"prompt_tokens_detail"`
+	// PromptTokensDetails contains detailed token usage statistics for the prompt
+	PromptTokensDetails *PromptTokensDetails `json:"prompt_tokens_details"`
 }
 
-// PromptTokensDetail contains detailed token usage statistics for the prompt
-type PromptTokensDetail struct {
+// PromptTokensDetails contains detailed token usage statistics for the prompt
+type PromptTokensDetails struct {
 	// CacheTokens is the number of tokens in the prompt that are in the local KV Cache
 	CachedTokens int `json:"cached_tokens"`
 }

pkg/communication/response_builder.go

@@ -115,8 +115,8 @@ func aggregateUsage(respCtxPerChoice []vllmsim.ResponseContext) *api.Usage {
 		if !seenIDs[rc.RequestID()] {
 			seenIDs[rc.RequestID()] = true
 			agg.PromptTokens += u.PromptTokens
-			if u.PromptTokensDetail != nil && agg.PromptTokensDetail == nil {
-				agg.PromptTokensDetail = u.PromptTokensDetail
+			if u.PromptTokensDetails != nil && agg.PromptTokensDetails == nil {
+				agg.PromptTokensDetails = u.PromptTokensDetails
 			}
 		}
 	}

pkg/llm-d-inference-sim/request.go

@@ -238,7 +238,7 @@ func (reqCtx *baseRequestContext) handleRequest() (ResponseContext, *api.Error)
 		PromptTokens:     numOfInputTokens,
 		CompletionTokens: completionTokens,
 		TotalTokens:      numOfInputTokens + completionTokens,
-		PromptTokensDetail: &api.PromptTokensDetail{
+		PromptTokensDetails: &api.PromptTokensDetails{
 			CachedTokens: prefixCacheStats.CachedTokens,
 		},
 	}