Fix by using the actual real value as outputted by the validation

workers.

router/src/infer.rs

@@ -90,6 +90,7 @@ impl Infer {
     ) -> Result<
         (
             OwnedSemaphorePermit,
+            u32,
             UnboundedReceiverStream<Result<InferStreamResponse, InferError>>,
         ),
         InferError,
@@ -114,6 +115,7 @@ impl Infer {
 
         // MPSC channel to communicate with the background batching task
         let (response_tx, response_rx) = mpsc::unbounded_channel();
+        let input_length = valid_request.input_length;
 
         // Append the request to the queue
         self.queue.append(Entry {
@@ -130,7 +132,11 @@ impl Infer {
         self.shared.batching_task.notify_one();
 
         // Return stream
-        Ok((permit, UnboundedReceiverStream::new(response_rx)))
+        Ok((
+            permit,
+            input_length,
+            UnboundedReceiverStream::new(response_rx),
+        ))
     }
 
     /// Add a new request to the queue and return a InferResponse
@@ -142,7 +148,7 @@ impl Infer {
         let use_top_tokens = request.parameters.top_n_tokens.is_some_and(|x| x > 0);
 
         // Create stream and keep semaphore permit as long as generate lives
-        let (_permit, mut stream) = self.generate_stream(request).await?;
+        let (_permit, input_length, mut stream) = self.generate_stream(request).await?;
 
         // Return values
         let mut result_prefill = Vec::new();
@@ -196,6 +202,7 @@ impl Infer {
         {
             Ok(InferResponse {
                 prefill: result_prefill,
+                input_length,
                 tokens: result_tokens,
                 generated_text,
                 queued,
@@ -636,6 +643,7 @@ pub(crate) enum InferStreamResponse {
 
 #[derive(Debug)]
 pub(crate) struct InferResponse {
+    pub(crate) input_length: u32,
     pub(crate) prefill: Vec<PrefillToken>,
     pub(crate) tokens: Vec<Token>,
     pub(crate) generated_text: GeneratedText,

router/src/server.rs

@@ -170,7 +170,7 @@ async fn generate(
     };
 
     // Token details
-    let prompt_tokens = response.prefill.len();
+    let input_length = response.input_length;
     let details = match details {
         true => {
             // convert best_of_responses
@@ -258,7 +258,7 @@ async fn generate(
         "x-time-per-token",
         time_per_token.as_millis().to_string().parse().unwrap(),
     );
-    headers.insert("x-prompt-tokens", prompt_tokens.into());
+    headers.insert("x-prompt-tokens", input_length.into());
     headers.insert(
         "x-generated-tokens",
         response.generated_text.generated_tokens.into(),
@@ -384,7 +384,7 @@ async fn generate_stream(
         } else {
             match infer.generate_stream(req).instrument(info_span!(parent: &span, "async_stream")).await {
                 // Keep permit as long as generate_stream lives
-                Ok((_permit, mut response_stream)) => {
+                Ok((_permit, _input_length, mut response_stream)) => {
                     // Server-Sent Event stream
                     while let Some(response) = response_stream.next().await {
                         match response {