소스 검색

feat: 添加日志子目录和全局日志配置,优化处理器日志管理

zhch158_admin 1 주 전
부모
커밋
a9a8e8cf3b
1개의 변경된 파일19개의 추가작업 그리고 8개의 파일을 삭제
  1. 19 8
      batch_ocr/processor_configs.yaml

+ 19 - 8
batch_ocr/processor_configs.yaml

@@ -6,7 +6,6 @@
 processors:
   # -------------------------------------------------------------------------
   # PaddleOCR-VL 处理器
-  # 用于视觉语言模型的 OCR 处理
   # -------------------------------------------------------------------------
   paddleocr_vl_single_process:
     script: "/Users/zhch158/workspace/repository.git/PaddleX/zhch/paddleocr_vl_single_process.py"
@@ -14,14 +13,15 @@ processors:
     output_arg: "--output_dir"
     extra_args:
       - "--pipeline=/Users/zhch158/workspace/repository.git/PaddleX/zhch/my_config/PaddleOCR-VL-Client-RT-DETR-H_layout_17cls.yaml"
-      - "--no-adapter"
+      - "--device=cpu"
+      # - "--no-adapter"
     output_subdir: "paddleocr_vl_results"
+    log_subdir: "logs/paddleocr_vl"  # 🎯 新增:日志子目录
     venv: "source /Users/zhch158/workspace/repository.git/PaddleX/paddle_env/bin/activate"
     description: "PaddleOCR-VL 处理器 - 视觉语言模型OCR"
 
   # -------------------------------------------------------------------------
   # PP-StructureV3 本地处理器
-  # 用于文档结构化分析(本地GPU/CPU处理)
   # -------------------------------------------------------------------------
   ppstructurev3_single_process:
     script: "/home/ubuntu/zhch/PaddleX/zhch/ppstructurev3_single_process.py"
@@ -29,22 +29,25 @@ processors:
     output_arg: "--output_dir"
     extra_args:
       - "--pipeline=/home/ubuntu/zhch/PaddleX/zhch/my_config/PP-StructureV3.yaml"
+      - "--device=cpu"
     output_subdir: "ppstructurev3_results"
+    log_subdir: "logs/ppstructurev3"
     venv: "conda activate paddle"
     description: "PP-StructureV3 处理器 - 本地处理"
 
-  # -------------------------------------------------------------------------
-  # PP-StructureV3 GPU 处理器
-  # 明确使用 GPU 加速
-  # -------------------------------------------------------------------------
   ppstructurev3_gpu:
     script: "/home/ubuntu/zhch/PaddleX/zhch/ppstructurev3_single_process.py"
     input_arg: "--input_file"
     output_arg: "--output_dir"
     extra_args:
       - "--pipeline=/home/ubuntu/zhch/PaddleX/zhch/my_config/PP-StructureV3.yaml"
+    input_arg: "--input_file"
+    output_arg: "--output_dir"
+    extra_args:
+      - "--pipeline=/home/ubuntu/zhch/PaddleX/zhch/my_config/PP-StructureV3.yaml"
       - "--device=gpu"
     output_subdir: "ppstructurev3_gpu_results"
+    log_subdir: "logs/ppstructurev3_gpu"
     venv: "conda activate paddle"
     description: "PP-StructureV3 处理器 - GPU加速"
 
@@ -60,6 +63,7 @@ processors:
       - "--pipeline=/Users/zhch158/workspace/repository.git/PaddleX/zhch/my_config/PP-StructureV3-zhch.yaml"
       - "--device=cpu"
     output_subdir: "ppstructurev3_cpu_results"
+    log_subdir: "logs/ppstructurev3_cpu"
     venv: "source /Users/zhch158/workspace/repository.git/PaddleX/paddle_env/bin/activate"
     description: "PP-StructureV3 处理器 - CPU处理"
 
@@ -75,6 +79,7 @@ processors:
       - "--api_url=http://10.192.72.11:8111/layout-parsing"
       - "--timeout=300"
     output_subdir: "ppstructurev3_client_results"
+    log_subdir: "logs/ppstructurev3_client"
     venv: "source /Users/zhch158/workspace/repository.git/PaddleX/paddle_env/bin/activate"
     description: "PP-StructureV3 HTTP API 客户端 - 远程服务"
 
@@ -91,6 +96,7 @@ processors:
       - "--timeout=300"
       - "--batch_size=1"
     output_subdir: "mineru_vllm_results"
+    log_subdir: "logs/mineru_vllm"
     venv: "conda activate mineru2"
     description: "MinerU vLLM 处理器 - 支持PDF和图片"
 
@@ -111,6 +117,7 @@ processors:
       - "--max_workers=1"
       - "--dpi=200"
     output_subdir: "dotsocr_vllm_results"
+    log_subdir: "logs/dotsocr_vllm"
     venv: "conda activate py312"
     description: "DotsOCR vLLM 处理器 - 支持PDF和图片"
 
@@ -123,4 +130,8 @@ global:
   
   # 默认输出子目录名称(如果处理器未指定)
   output_subdir: "results"
-  
+  
+  # 🎯 新增:全局日志配置
+  log_dir: "logs"  # 全局日志目录(相对于 base_dir)
+  log_retention_days: 30  # 日志保留天数
+  log_level: "INFO"  # 日志级别: DEBUG, INFO, WARNING, ERROR