generate remove unnecessary code

src/models/deepseek_ocr/generate.rs

@@ -80,22 +80,32 @@ impl GenerateModel for DeepseekOCRGenerateModel {
         let (mut input_ids, images_ori, image_crop, images_seq_mask, images_spatial_crop_t) = self
             .processor
             .process_info(&mes, &self.tokenizer, base_size, image_size, crop_mode)?;
-        let mut images_ori = Some(&images_ori);
-        let mut image_crop = Some(&image_crop);
-        let mut images_seq_mask = Some(&images_seq_mask);
-        let mut images_spatial_crop_t = Some(&images_spatial_crop_t);
         let mut seqlen_offset = 0;
         let mut seq_len = input_ids.dim(1)?;
         let prompt_tokens = seq_len as u32;
         let mut generate = Vec::new();
+        let logits = self.deepseekocr_model.forward(
+            &input_ids,
+            Some(&images_ori),
+            Some(&image_crop),
+            Some(&images_seq_mask),
+            Some(&images_spatial_crop_t),
+            seqlen_offset,
+        )?;
+        let logits = logits.squeeze(0)?.squeeze(0)?.to_dtype(DType::F32)?;
+        let next_token = logit_processor.sample(&logits)?;
+        generate.push(next_token);
+        input_ids = Tensor::from_vec(vec![next_token], (1, 1), &self.device)?;
+        seqlen_offset += seq_len;
+        seq_len = 1;
         let sample_len = mes.max_tokens.unwrap_or(1024);
-        for _ in 0..sample_len {
+        for _ in 1..sample_len {
             let logits = self.deepseekocr_model.forward(
                 &input_ids,
-                images_ori,
-                image_crop,
-                images_seq_mask,
-                images_spatial_crop_t,
+                None,
+                None,
+                None,
+                None,
                 seqlen_offset,
             )?;
             let logits = logits.squeeze(0)?.squeeze(0)?.to_dtype(DType::F32)?;
@@ -107,10 +117,6 @@ impl GenerateModel for DeepseekOCRGenerateModel {
             seqlen_offset += seq_len;
             seq_len = 1;
             input_ids = Tensor::from_vec(vec![next_token], (1, 1), &self.device)?;
-            images_ori = None;
-            image_crop = None;
-            images_seq_mask = None;
-            images_spatial_crop_t = None;
         }
         let num_token = generate.len() as u32;
         let res = self.tokenizer.token_decode(generate)?;

src/models/qwen2_5vl/generate.rs

@@ -144,12 +144,6 @@ impl<'a> GenerateModel for Qwen2_5VLGenerateModel<'a> {
             .text_encode(input.replace_text.clone(), &self.device)?;
         let mut seq_len = input_ids.dim(1)?;
         let mut seqlen_offset = 0;
-        let pixel_values = input.pixel_values.clone();
-        let image_grid_thw = input.image_grid_thw.clone();
-        let pixel_values_video = input.pixel_values_video.clone();
-        let video_grid_thw = input.video_grid_thw.clone();
-        let second_per_grid_ts = input.second_per_grid_ts.clone();
-
         let mut mask = Tensor::ones_like(&input_ids)?;
         let mut cache_position = Tensor::ones_like(&input_ids.i(0)?)?
             .to_dtype(candle_core::DType::F64)?
@@ -160,10 +154,10 @@ impl<'a> GenerateModel for Qwen2_5VLGenerateModel<'a> {
         let sample_len = mes.max_tokens.unwrap_or(512);
         let stream = stream! {
             let mut error_tokens = Vec::new();
-            let mut pixel_values = pixel_values.as_ref();
-            let image_grid_thw = image_grid_thw.as_ref();
-            let mut pixel_values_video = pixel_values_video.as_ref();
-            let video_grid_thw = video_grid_thw.as_ref();
+            let mut pixel_values = input.pixel_values.as_ref();
+            let image_grid_thw = input.image_grid_thw.as_ref();
+            let mut pixel_values_video = input.pixel_values_video.as_ref();
+            let video_grid_thw = input.video_grid_thw.as_ref();
             let mut tool_call_id = None;
             let mut tool_call_content = String::new();
             for _ in 0..sample_len {
@@ -176,7 +170,7 @@ impl<'a> GenerateModel for Qwen2_5VLGenerateModel<'a> {
                     &mask,
                     Some(&cache_position),
                     seqlen_offset,
-                    second_per_grid_ts.clone(),
+                    input.second_per_grid_ts.clone(),
                 )?;
                 let logits = logits.squeeze(0)?.squeeze(0)?.to_dtype(DType::F32)?;
                 let next_token = logit_processor.sample(&logits)?;

src/models/qwen3/generate.rs

@@ -116,7 +116,6 @@ impl<'a> GenerateModel for Qwen3GenerateModel<'a> {
         let mut logit_processor =
             get_logit_processor(Some(temperature), Some(top_p), Some(top_k), seed);
         let enable_thinking = extract_metadata_value::<bool>(&mes.metadata, "enable_thinking");
-        // let mes_render = self.chat_template.apply_chat_template(&mes)?;
         let mes_render = self
             .chat_template
             .apply_chat_temp_think(&mes, enable_thinking)?;

src/models/qwen3vl/generate.rs

@@ -153,18 +153,14 @@ impl<'a> GenerateModel for Qwen3VLGenerateModel<'a> {
             .text_encode(input.replace_text.clone(), &self.device)?;
         let mut seq_len = input_ids.dim(1)?;
         let mut seqlen_offset = 0;
-        let pixel_values = input.pixel_values.clone();
-        let image_grid_thw = input.image_grid_thw.clone();
-        let pixel_values_video = input.pixel_values_video.clone();
-        let video_grid_thw = input.video_grid_thw.clone();
         let mut cache_position = Tensor::arange(0u32, seq_len as u32, &self.device)?;
         let sample_len = mes.max_tokens.unwrap_or(1024);
         let stream = stream! {
             let mut error_tokens = Vec::new();
-            let mut pixel_values = pixel_values.as_ref();
-            let image_grid_thw = image_grid_thw.as_ref();
-            let mut pixel_values_video = pixel_values_video.as_ref();
-            let video_grid_thw = video_grid_thw.as_ref();
+            let mut pixel_values = input.pixel_values.as_ref();
+            let image_grid_thw = input.image_grid_thw.as_ref();
+            let mut pixel_values_video = input.pixel_values_video.as_ref();
+            let video_grid_thw = input.video_grid_thw.as_ref();
             let mut tool_call_id = None;
             let mut tool_call_content = String::new();
             for _ in 0..sample_len {

tests/test_qwen2_5vl.rs

@@ -63,7 +63,7 @@ fn qwen2_5vl_generate() -> Result<()> {
 
 #[tokio::test]
 async fn qwen2_5vl_stream() -> Result<()> {
-    // test with cuda+flash-attn: RUST_BACKTRACE=1 cargo test -F cuda,flash-attn qwen2_5vl_stream -r -- --nocapture
+    // test with cuda: RUST_BACKTRACE=1 cargo test -F cuda qwen2_5vl_stream -r -- --nocapture
     // let device = Device::cuda_if_available(0)?;
     // let dtype = DType::BF16;