Feat: add OCR's muti-gpus and parallel processing support (#5972)

### What problem does this PR solve?

Add OCR's muti-gpus and parallel processing support

### Type of change

- [x] New Feature (non-breaking change which adds functionality)

@yuzhichang I've tried to resolve the comments in #5697. OCR jobs can
now be done on both CPU and GPU. ( By the way, I've encountered a
“Generate embedding error” issue #5954 that might be due to my outdated
GPUs? idk. ) Please review it and give me suggestions.

GPU:

![gpu_ocr](https://github.com/user-attachments/assets/0ee2ecfb-a665-4e50-8bc7-15941b9cd80e)

![smi](https://github.com/user-attachments/assets/a2312f8c-cf24-443d-bf89-bec50503546d)

CPU:

![cpu_ocr](https://github.com/user-attachments/assets/1ba6bb0b-94df-41ea-be79-790096da4bf1)

rag/app/naive.py

@@ -128,6 +128,9 @@ class Docx(DocxParser):
 
 
 class Pdf(PdfParser):
+    def __init__(self, parallel_devices = None):
+        super().__init__(parallel_devices)
+
     def __call__(self, filename, binary=None, from_page=0,
                  to_page=100000, zoomin=3, callback=None):
         start = timer()
@@ -194,7 +197,7 @@ class Markdown(MarkdownParser):
 
 
 def chunk(filename, binary=None, from_page=0, to_page=100000,
-          lang="Chinese", callback=None, **kwargs):
+          lang="Chinese", parallel_devices=None, callback=None, **kwargs):
     """
         Supported file formats are docx, pdf, excel, txt.
         This method apply the naive ways to chunk files.
@@ -234,7 +237,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
         return res
 
     elif re.search(r"\.pdf$", filename, re.IGNORECASE):
-        pdf_parser = Pdf()
+        pdf_parser = Pdf(parallel_devices)
         if parser_config.get("layout_recognize", "DeepDOC") == "Plain Text":
             pdf_parser = PlainParser()
         sections, tables = pdf_parser(filename if not binary else binary, from_page=from_page, to_page=to_page,