Implement lazy loading to avoid unnecessary imports

swanlab/converter/__init__.py

@@ -7,5 +7,20 @@
 @Description:
     转换部分，兼容其他可视化工具，转换为swanlab格式
 """
-from .tfb import TFBConverter
-from .wb import WandbConverter
+
+
+# 使用延迟导入的方式
+def __getattr__(name):
+    if name == "TFBConverter":
+        from .tfb import TFBConverter
+
+        return TFBConverter
+    if name == "WandbConverter":
+        from .wb import WandbConverter
+
+        return WandbConverter
+
+    raise AttributeError(f"module 'convert' has no attribute '{name}'")
+
+
+__all__ = ["TFBConverter", "WandbConverter"]

swanlab/converter/tfb/_utils.py

@@ -1,17 +1,15 @@
 import os
-from PIL import Image
 import io
-import numpy as np
 
-try:
-    import tensorflow as tf
-except ImportError as e:
-    raise TypeError(
-        "Tensorboard Converter requires tensorflow when process tfevents file. Install with 'pip install tensorflow'."
-    )
 
+def get_tf_events_tags_type(tf_event_path: str):  
+    try:
+        import tensorflow as tf
+    except ImportError as e:
+        raise TypeError(
+            "Tensorboard Converter requires tensorflow when process tfevents file. Install with 'pip install tensorflow'."
+        )
 
-def get_tf_events_tags_type(tf_event_path: str):
     """获取TFEvent文件中所有tag的类型，并返回一个字典
     比如{"tag1": "scalar", "tag2": "image", "tag3": "audio", "tag4": "text"}
 
@@ -44,6 +42,15 @@ def get_tf_events_tags_type(tf_event_path: str):
 
 
 def get_tf_events_tags_data(tf_event_path: str, tags: dict):
+    try:
+        import tensorflow as tf
+    except ImportError as e:
+        raise TypeError(
+            "Tensorboard Converter requires tensorflow when process tensorboard tfevents file. Install with 'pip install tensorflow'."
+        )
+    import numpy as np
+    from PIL import Image
+
     """获取TFEvent文件中所有tag的数据，并返回一个字典
     比如{"tag1": [(step1, value1), (step2, value2)], "tag2": [(step1, value1), (step2, value2)]}
 

swanlab/converter/wb/wb_converter.py

@@ -70,7 +70,12 @@ def parse_wandb_logs(self, wb_project: str, wb_entity: str, wb_run_id: str = Non
             swanlab_run.config.update(wb_config)
             swanlab_run.config.update(wb_run.config)
 
-            keys = [key for key in wb_run.history(stream="default").keys() if not key.startswith("_")]
+            # Get the first history record to extract available keys
+            history = wb_run.history(stream="default")
+            if len(history) > 0:
+                keys = [key for key in history[0].keys() if not key.startswith("_")]
+            else:
+                keys = []
 
             # 记录标量指标
             for record in wb_run.scan_history():