docs: integration paddleyolo (#57)

.vitepress/config.mts

@@ -162,6 +162,7 @@ function sidebarGuideCloud(): DefaultTheme.SidebarItem[] {
       { text: 'MMSegmentation', link: 'integration/integration-mmsegmentation' },
       { text: 'OpenAI', link: 'integration/integration-openai' },
       { text: 'Omegaconf', link: 'integration/integration-omegaconf' },
+      { text: 'PaddleYOLO', link: 'integration/integration-paddleyolo' },
       { text: 'PyTorch', link: 'integration/integration-pytorch' },
       { text: 'PyTorch Lightning', link: 'integration/integration-pytorch-lightning' },
       { text: 'PyTorch torchtune', link: 'integration/integration-pytorch-torchtune' },

zh/guide_cloud/integration/integration-paddleyolo.md

@@ -0,0 +1,150 @@
+# PaddleYolo
+
+[PaddleYolo](https://github.com/PaddlePaddle/PaddleYOLO) 是飞桨（PaddlePaddle）框架下的一个目标检测库，主要用于图像和视频中的物体检测。PaddleYOLO包含YOLO系列模型的相关代码，支持YOLOv3、PP-YOLO、PP-YOLOv2、PP-YOLOE、PP-YOLOE+、RT-DETR、YOLOX、YOLOv5、YOLOv6、YOLOv7、YOLOv8、YOLOv5u、YOLOv7u、YOLOv6Lite、RTMDet等模型
+
+你可以使用PaddleYolo快速进行目标检测模型训练，同时使用SwanLab进行实验跟踪与可视化。
+
+[Demo](https://swanlab.cn/@ZeyiLin/PaddleYOLO/runs/10zy8zickn2062kubch34/chart)
+
+## 1. 引入SwanLabCallback
+
+首先在你clone的PaddleYolo项目中，找到`ppdet/engine/callbacks.py`文件，在代码的底部添加如下代码：
+
+```python
+class SwanLabCallback(Callback):
+    def __init__(self, model):
+        super(SwanLabCallback, self).__init__(model)
+
+        try:
+            import swanlab
+            self.swanlab = swanlab
+        except Exception as e:
+            logger.error('swanlab not found, please install swanlab. '
+                         'Use: `pip install swanlab`.')
+            raise e
+
+        self.swanlab_params = {k.lstrip("swanlab_"): v for k, v in model.cfg.items() if k.startswith("swanlab_")}
+
+        self._run = None
+        if dist.get_world_size() < 2 or dist.get_rank() == 0:
+            _ = self.run
+            self.run.config.update(self.model.cfg)
+
+        self.best_ap = -1000.
+        self.fps = []
+
+    @property
+    def run(self):
+        if self._run is None:
+            self._run = self.swanlab.get_run() or self.swanlab.init(**self.swanlab_params)
+        return self._run
+
+    def on_step_end(self, status):
+        if dist.get_world_size() < 2 or dist.get_rank() == 0 and status['mode'] == 'train':
+            training_status = status['training_staus'].get()
+            batch_time = status['batch_time']
+            data_time = status['data_time']
+            batch_size = self.model.cfg['{}Reader'.format(status['mode'].capitalize())]['batch_size']
+
+            ips = float(batch_size) / float(batch_time.avg)
+            metrics = {
+                "train/" + k: float(v) for k, v in training_status.items()
+            }
+            metrics.update({
+                "train/ips": ips,
+                "train/data_cost": float(data_time.avg),
+                "train/batch_cost": float(batch_time.avg)
+            })
+
+            self.fps.append(ips)
+            self.run.log(metrics)
+
+    def on_epoch_end(self, status):
+        if dist.get_world_size() < 2 or dist.get_rank() == 0:
+            mode = status['mode']
+            epoch_id = status['epoch_id']
+
+            if mode == 'train':
+                fps = sum(self.fps) / len(self.fps)
+                self.fps = []
+
+                end_epoch = self.model.cfg.epoch
+                if (epoch_id + 1) % self.model.cfg.snapshot_epoch == 0 or epoch_id == end_epoch - 1:
+                    save_name = str(epoch_id) if epoch_id != end_epoch - 1 else "model_final"
+                    tags = ["latest", f"epoch_{epoch_id}"]
+
+            elif mode == 'eval':
+                fps = status['sample_num'] / status['cost_time']
+
+                merged_dict = {
+                    f"eval/{key}-mAP": map_value[0]
+                    for metric in self.model._metrics
+                    for key, map_value in metric.get_results().items()
+                }
+                merged_dict.update({
+                    "epoch": status["epoch_id"],
+                    "eval/fps": fps
+                })
+
+                self.run.log(merged_dict)
+
+                if status.get('save_best_model'):
+                    for metric in self.model._metrics:
+                        map_res = metric.get_results()
+                        key = next((k for k in ['bbox', 'keypoint', 'mask'] if k in map_res), None)
+
+                        if not key:
+                            logger.warning("Evaluation results empty, this may be due to "
+                                           "training iterations being too few or not "
+                                           "loading the correct weights.")
+                            return
+
+                        if map_res[key][0] >= self.best_ap:
+                            self.best_ap = map_res[key][0]
+                            save_name = 'best_model'
+                            tags = ["best", f"epoch_{epoch_id}"]
+
+    def on_train_end(self, status):
+        self.run.finish()
+```
+
+## 2. 修改trainer代码
+
+在`ppdet/engine/trainer.py`文件中，在`from .callbacks import`那一行添加SwanLabCallback：
+
+```python
+from .callbacks import Callback, ComposeCallback, LogPrinter, Checkpointer, VisualDLWriter, WandbCallback, SwanLabCallback
+```
+
+接着，我们找到`Trainer`类的`__init_callbacks`方法，在`if self.mode == 'train':`下添加如下代码：
+
+```python
+if self.cfg.get('use_swanlab', False) or 'swanlab' in self.cfg:
+    self._callbacks.append(SwanLabCallback(self))
+```
+
+至此，你已经完成了SwanLab与PaddleYolo的集成！接下来，只需要在训练的配置文件中添加`use_swanlab: True`，即可开始可视化跟踪训练。
+
+## 3. 修改配置文件
+
+我们以`yolov3_mobilenet_v1_roadsign`为例。
+
+在`configs/yolov3/yolov3_mobilenet_v1_roadsign.yml`文件中，在下面添加如下代码：
+
+```yaml
+use_swanlab: true
+swanlab_project: PaddleYOLO # 可选
+swanlab_experiment_name: yolov3_mobilenet_v1_roadsign # 可选
+swanlab_description: 对PaddleYOLO的一次训练测试 # 可选
+# swanlab_workspace: swanhub # 组织名，可选
+```
+
+## 4. 开始训练
+
+```bash
+python -u tools/train.py -c configs/yolov3/yolov3_mobilenet_v1_roadsign.yml --eval
+```
+
+在训练过程中，即可看到整个训练过程的日志，以及训练结束后自动生成的可视化图表。
+
+![paddleyolo-image](/assets/ig-paddleyolo.png)