update fmt

src/models/qwen2_5vl/generate.rs

@@ -222,9 +222,9 @@ impl<'a> GenerateModel for Qwen2_5VLGenerateModel<'a> {
                     "</tool_call>" => {
                         // 结束工具调用
                         let chunk = build_completion_chunk_response(
-                            decoded_token, 
+                            decoded_token,
-                            &self.model_name, 
+                            &self.model_name,
-                            tool_call_id.clone(), 
+                            tool_call_id.clone(),
                             Some(tool_call_content.clone())
                         );
                         tool_call_id = None;
@@ -245,15 +245,15 @@ impl<'a> GenerateModel for Qwen2_5VLGenerateModel<'a> {
                         } else {
                             // 正常文本输出
                             let chunk = build_completion_chunk_response(
-                                decoded_token, 
+                                decoded_token,
-                                &self.model_name, 
+                                &self.model_name,
-                                None, 
+                                None,
                                 None
                             );
                             yield Ok(chunk);
                         }
                     }
-                }          
+                }
                 // let chunk = build_completion_chunk_response(decoded_token, &self.model_name, None, None);
                 // yield Ok(chunk);
                 if next_token == self.endoftext_id || next_token == self.im_end_id {

src/models/qwen2_5vl/processor.rs

@@ -72,9 +72,9 @@ impl Qwen2_5VLProcessor {
                     if let ChatMessageContentPart::Image(img_part) = part {
                         let img_url = img_part.image_url;
                         vision_map.get_mut("image").unwrap().push(img_url.url);
-                    // } else if let ChatMessageContentPart::Video(video_part) = part {
+                        // } else if let ChatMessageContentPart::Video(video_part) = part {
-                    //     let video_url = video_part.video_url;
+                        //     let video_url = video_part.video_url;
-                    //     vision_map.get_mut("video").unwrap().push(video_url.url);
+                        //     vision_map.get_mut("video").unwrap().push(video_url.url);
                     }
                 }
             }

src/utils/mod.rs

@@ -137,24 +137,23 @@ pub fn ceil_by_factor(num: f32, factor: u32) -> u32 {
     ceil * factor
 }
 
-pub fn build_completion_response(res: String, model_name: &str, num_tokens: Option<u32>) -> ChatCompletionResponse {
+pub fn build_completion_response(
+    res: String,
+    model_name: &str,
+    num_tokens: Option<u32>,
+) -> ChatCompletionResponse {
     let id = uuid::Uuid::new_v4().to_string();
-    let usage = match num_tokens {
+    let usage = num_tokens.map(|num| Usage {
-        Some(num) => {
+        input_tokens: None,
-            Some(Usage {
+        input_tokens_details: None,
-                input_tokens: None,
+        output_tokens: None,
-                input_tokens_details: None,
+        output_tokens_details: None,
-                output_tokens: None,
+        prompt_tokens: None,
-                output_tokens_details: None,
+        completion_tokens: None,
-                prompt_tokens: None,
+        total_tokens: num,
-                completion_tokens: None,
+        prompt_tokens_details: None,
-                total_tokens: num,
+        completion_tokens_details: None,
-                prompt_tokens_details: None,
+    });
-                completion_tokens_details: None
-            })
-        },
-        None => None,
-    };
     let mut response = ChatCompletionResponse {
         id: Some(id),
         choices: vec![],
@@ -163,7 +162,7 @@ pub fn build_completion_response(res: String, model_name: &str, num_tokens: Opti
         service_tier: None,
         system_fingerprint: None,
         object: "chat.completion".to_string(),
-        usage
+        usage,
     };
     let choice = if res.contains("<tool_call>") {
         let mes: Vec<&str> = res.split("<tool_call>").collect();

tests/test_deepseek_ocr.rs

@@ -48,7 +48,7 @@ fn deepseek_ocr_generate() -> Result<()> {
         let tps = num_token as f64 / duration_secs;
         println!("Tokens per second (TPS): {:.2}", tps);
     }
-    println!("Time elapsed in generate is: {:?}", i_duration);    
+    println!("Time elapsed in generate is: {:?}", i_duration);
     Ok(())
 }
 

tests/test_hunyuan_ocr.rs

@@ -48,7 +48,7 @@ fn hunyuan_ocr_generate() -> Result<()> {
         println!("Tokens per second (TPS): {:.2}", tps);
     }
     println!("Time elapsed in generate is: {:?}", i_duration);
-    
+
     Ok(())
 }
 

tests/test_paddleocr_vl.rs

@@ -48,7 +48,7 @@ fn paddleocr_vl_generate() -> Result<()> {
         let tps = num_token as f64 / duration_secs;
         println!("Tokens per second (TPS): {:.2}", tps);
     }
-    println!("Time elapsed in generate is: {:?}", i_duration);    
+    println!("Time elapsed in generate is: {:?}", i_duration);
     Ok(())
 }
 

tests/test_qwen2_5vl.rs

@@ -45,7 +45,7 @@ fn qwen2_5vl_generate() -> Result<()> {
     println!("Time elapsed in load model is: {:?}", i_duration);
 
     let i_start = Instant::now();
-    let result = model.generate(mes)?;    
+    let result = model.generate(mes)?;
     let i_duration = i_start.elapsed();
     println!("generate: \n {:?}", result);
     if result.usage.is_some() {

tests/test_robo_brain.rs

@@ -35,7 +35,7 @@ fn robo_brain_generate() -> Result<()> {
     let i_start = Instant::now();
     let result = model.generate(mes)?;
     let i_duration = i_start.elapsed();
-    println!("generate: \n {:?}", result);    
+    println!("generate: \n {:?}", result);
     if result.usage.is_some() {
         let num_token = result.usage.as_ref().unwrap().total_tokens;
         let duration_secs = i_duration.as_secs_f64();