diff --git a/docs/Dockerfile.docs b/docs/Dockerfile.docs
index cd9e712dcd..b3f078baaa 100644
--- a/docs/Dockerfile.docs
+++ b/docs/Dockerfile.docs
@@ -24,7 +24,7 @@
 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
 # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
-FROM ubuntu:22.04
+FROM ubuntu:24.04
 
 # various documentation dependencies
 RUN apt-get update -q=2 \
@@ -53,6 +53,7 @@ RUN wget https://github.com/pseudomuto/protoc-gen-doc/releases/download/v1.3.2/p
     && rm /tmp/protoc-gen-doc.tar.gz
 
 # install sphinx et al
+ENV PIP_BREAK_SYSTEM_PACKAGES=1
 RUN pip3 install \
       ablog \
       attrs  \
@@ -75,8 +76,13 @@ RUN pip3 install \
       sphinxcontrib-bibtex
 
 RUN pip3 install \
-      --index-url https://urm.nvidia.com/artifactory/api/pypi/ct-omniverse-pypi/simple/ \
-      nvidia-sphinx-theme
+      --extra-index-url https://pypi.nvidia.com \
+      nvidia-sphinx-theme \
+      sphinx==7.4.7
+
+RUN curl -fL https://install-cli.jfrog.io | sh
+
+RUN git config --global --add safe.directory "*"
 
 # Set visitor script to be included on every HTML page
 ENV VISITS_COUNTING_SCRIPT="//assets.adobedtm.com/b92787824f2e0e9b68dc2e993f9bd995339fe417/satelliteLib-7ba51e58dc61bcb0e9311aadd02a0108ab24cc6c.js"
diff --git a/docs/conf.py b/docs/conf.py
index 0b44f7c8b2..387e0f3c37 100755
--- a/docs/conf.py
+++ b/docs/conf.py
@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 
-# Copyright 2023-2024, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# Copyright 2023-2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -35,9 +35,12 @@
 # -- Path setup --------------------------------------------------------------
 
 import json
+import logging
 import os
 import re
+import subprocess
 from datetime import date
+from logging.handlers import RotatingFileHandler
 
 # If extensions (or modules to document with autodoc) are in another directory,
 # add these directories to sys.path here. If the directory is relative to the
@@ -61,6 +64,45 @@
 # at the end of the file.
 # current_dir = os.getcwd()
 # os.chdir("docs")
+# -- Setup logger ------------------------------------------------------------
+
+
+def setup_logger(name, log_file, level=logging.INFO, max_bytes=1048576, backup_count=5):
+    logger = logging.getLogger(name)
+    logger.setLevel(level)
+
+    # Prevent adding multiple handlers if the function is called multiple times
+    if not logger.handlers:
+        # Create handlers
+        file_handler = RotatingFileHandler(
+            log_file, maxBytes=max_bytes, backupCount=backup_count
+        )
+        console_handler = logging.StreamHandler()
+
+        # Set the logging level for handlers
+        file_handler.setLevel(level)
+        console_handler.setLevel(level)
+
+        # Create a logging format
+        BLUE = "\033[94m"
+        RESET = "\033[0m"
+        formatter = logging.Formatter(
+            f"{BLUE}%(asctime)s - %(name)s - %(levelname)s - {RESET}%(message)s"
+        )
+        file_handler.setFormatter(formatter)
+        console_handler.setFormatter(formatter)
+
+        # Add handlers to the logger
+        logger.addHandler(file_handler)
+        logger.addHandler(console_handler)
+    return logger
+
+
+logger = setup_logger(
+    os.path.basename(__file__),
+    os.environ.get("TRITON_SERVER_DOCS_LOG_FILE", "/tmp/docs.log"),
+)
+logger.info(f"Defined logger for {os.path.basename(__file__)}")
 
 # -- Project information -----------------------------------------------------
 
@@ -70,14 +112,19 @@
 
 # Get the version of Triton this is building.
 version_long = "0.0.0"
+logger.info(f"Getting version from ../TRITON_VERSION")
 with open("../TRITON_VERSION") as f:
     version_long = f.readline()
     version_long = version_long.strip()
+    logger.info(f"Version: {version_long}")
+
 
 version_short = re.match(r"^[\d]+\.[\d]+\.[\d]+", version_long).group(0)
+logger.info(f"Version short: {version_short}")
 version_short_split = version_short.split(".")
+logger.info(f"Version short split: {version_short_split}")
 one_before = f"{version_short_split[0]}.{int(version_short_split[1]) - 1}.{version_short_split[2]}"
-
+logger.info(f"One before: {one_before}")
 
 # maintain left-side bar toctrees in `contents` file
 # so it doesn't show up needlessly in the index page
@@ -180,7 +227,7 @@
         "json_url": "https://docs.nvidia.com/deeplearning/triton-inference-server/user-guide/docs/_static/switcher.json",
         "version_match": one_before if "dev" in version_long else version_short,
     },
-    "navbar_start": ["navbar-logo", "version-switcher"],
+    # "navbar_start": ["navbar-logo", "version-switcher"],
     "primary_sidebar_end": [],
 }
 
@@ -194,6 +241,8 @@
     }
 )
 
+logger.info(f"html_theme_options: {html_theme_options}")
+
 deploy_ngc_org = "nvidia"
 deploy_ngc_team = "triton"
 myst_substitutions = {
@@ -203,6 +252,8 @@
     else deploy_ngc_org,
 }
 
+logger.info(f"myst_substitutions: {myst_substitutions}")
+
 
 def ultimateReplace(app, docname, source):
     result = source[0]
@@ -219,6 +270,7 @@ def ultimateReplace(app, docname, source):
     if deploy_ngc_team
     else deploy_ngc_org,
 }
+logger.info(f"ultimate_replacements: {ultimate_replacements}")
 
 # bibtex_bibfiles = ["references.bib"]
 # To test that style looks good with common bibtex config
@@ -233,25 +285,24 @@ def ultimateReplace(app, docname, source):
 # SETUP SWITCHER
 ###############################
 switcher_path = os.path.join(html_static_path[0], "switcher.json")
+logger.info(f"switcher_path: {switcher_path}")
 versions = []
-# Triton 2 releases
-correction = -1 if "dev" in version_long else 0
-upper_bound = version_short.split(".")[1]
-for i in range(2, int(version_short.split(".")[1]) + correction):
-    versions.append((f"2.{i}.0", f"triton-inference-server-2{i}0"))
 
-# Triton 1 releases
-for i in range(0, 15):
-    versions.append((f"1.{i}.0", f"tensorrt_inference_server_1{i}0"))
-
-# Triton Beta Releases
-for i in range(1, 11):
-    versions.append((f"0.{i}.0_beta", f"inference_server_0{i}0_beta"))
+# Obtain Triton Server Release Tags.
+tags = subprocess.run(["git", "tag", "--list", "v*"], capture_output=True, text=True)
+tags_list = sorted(tags.stdout.strip().splitlines(), key=Version, reverse=True)
+logger.info(f"Found source tags: {tags_list}")
+
+for v in tags_list:
+    versions.append(
+        (
+            v.replace("v", ""),
+            f"triton-inference-server-{v.replace('v', '').replace('.', '')}",
+        )
+    )
 
-# Patch releases
-# Add here.
+logger.info(f"Defined dictionary of versions: {versions}")
 
-versions = sorted(versions, key=lambda v: Version(v[0]), reverse=True)
 
 # Build switcher data
 json_data = []
@@ -263,6 +314,7 @@ def ultimateReplace(app, docname, source):
             "url": f"https://docs.nvidia.com/deeplearning/triton-inference-server/archives/{v[1]}/user-guide/docs",
         }
     )
+
 if "dev" in version_long:
     json_data.insert(
         0,
@@ -284,6 +336,7 @@ def ultimateReplace(app, docname, source):
 
 # Trim to last N releases.
 json_data = json_data[0:12]
+logger.info(f"Trimmed to last 12 release...")
 
 json_data.append(
     {
@@ -293,19 +346,22 @@ def ultimateReplace(app, docname, source):
     }
 )
 
-# validate the links
 for i, d in enumerate(json_data):
+    logger.info(f"Validating link: {d['url']}")
     h = httplib2.Http()
     resp = h.request(d["url"], "HEAD")
     if int(resp[0]["status"]) >= 400:
         print(d["url"], "NOK", resp[0]["status"])
-        exit(1)
+        # exit(1)
 
-# Write switcher data to file
+logger.info(f"Writing switcher data to file: {switcher_path}")
 with open(switcher_path, "w") as f:
     json.dump(json_data, f, ensure_ascii=False, indent=4)
 
 
+logger.info("Configuration completed...")
+
+
 def setup(app):
     app.add_config_value("ultimate_replacements", {}, True)
     app.connect("source-read", ultimateReplace)
diff --git a/docs/generate_docs.py b/docs/generate_docs.py
index b7abe51ed0..a38b06166b 100755
--- a/docs/generate_docs.py
+++ b/docs/generate_docs.py
@@ -31,8 +31,8 @@
 import os
 import re
 import subprocess
-from collections import defaultdict
 from functools import partial
+from logging.handlers import RotatingFileHandler
 
 # Global constants
 server_abspath = os.environ.get("SERVER_ABSPATH", os.getcwd())
@@ -59,208 +59,200 @@
 triton_github_url_reg = re.compile(
     rf"{triton_repo_patn}/([^/#]+)(?:{tag_patn})?/*([^#]*)\s*(?=#|$)"
 )
-# relpath_patn = r"]\s*\(\s*([^)]+)\)"
 # Hyperlink in a .md file, excluding embedded images.
 hyperlink_reg = re.compile(r"((?<!\!)\[[^\]]+\]\s*\(\s*)([^)]+?)(\s*\))")
 
-exclusions = None
-with open(f"{server_docs_abspath}/exclusions.txt", "r") as f:
-    exclusions = f.read()
-    f.close()
-exclude_patterns = exclusions.strip().split("\n")
+# Load exclusion patterns
+with open(f"{server_docs_abspath}/exclusions.txt") as f:
+    exclude_patterns = f.read().strip().split("\n")
 
-# Parser
-parser = argparse.ArgumentParser(description="Process some arguments.")
-parser.add_argument("--repo-tag", type=str, help="Repository tags in format value")
+
+# Setup logger once
+def setup_logger(name, log_file, level=logging.INFO, max_bytes=1048576, backup_count=5):
+    logger = logging.getLogger(name)
+    logger.setLevel(level)
+
+    # Prevent adding multiple handlers if the function is called multiple times
+    if not logger.handlers:
+        # Create handlers
+        file_handler = RotatingFileHandler(
+            log_file, maxBytes=max_bytes, backupCount=backup_count
+        )
+        console_handler = logging.StreamHandler()
+
+        # Set the logging level for handlers
+        file_handler.setLevel(level)
+        console_handler.setLevel(level)
+
+        # Create a logging format
+        BLUE = "\033[94m"
+        RESET = "\033[0m"
+        formatter = logging.Formatter(
+            f"{BLUE}%(asctime)s - %(name)s - %(levelname)s - {RESET}%(message)s"
+        )
+        file_handler.setFormatter(formatter)
+        console_handler.setFormatter(formatter)
+
+        # Add handlers to the logger
+        logger.addHandler(file_handler)
+        logger.addHandler(console_handler)
+    return logger
+
+
+parser = argparse.ArgumentParser(description="Setup Triton Server Docs")
+parser.add_argument(
+    "--repo-tag",
+    type=str,
+    default=os.environ.get("TRITON_SERVER_REPO_TAG", "main"),
+    help="Repository tags in format value",
+)
+parser.add_argument(
+    "--log-file",
+    type=str,
+    default=os.environ.get("TRITON_SERVER_DOCS_LOG_FILE", "/tmp/docs.log"),
+    help="The path to the log file",
+)
 parser.add_argument(
     "--repo-file",
     default="repositories.txt",
     help="File which lists the repositories to add. File should be"
     " one repository name per line, newline separated.",
 )
-parser.add_argument("--github-organization", help="GitHub organization name")
-
-
-def setup_logger():
-    """
-    This function is to setup logging
-    """
-    # Create a custom logger
-    logger = logging.getLogger(__name__)
-    # Set the log level
-    logger.setLevel(logging.INFO)
-    # Create handlers
-    file_handler = logging.FileHandler("/tmp/docs.log")
-    # Create formatters and add it to the handlers
-    formatter = logging.Formatter(
-        "%(asctime)s - %(name)s - %(levelname)s - %(message)s"
-    )
-    file_handler.setFormatter(formatter)
-    # Add handlers to the logger
-    logger.addHandler(file_handler)
-    return logger
+parser.add_argument(
+    "--github-organization",
+    type=str,
+    default=os.environ.get(
+        "TRITON_SERVER_REPO_ORG", "https://github.com/triton-inference-server"
+    ),
+    help="GitHub organization name",
+)
+args = parser.parse_args()
 
 
-def log_message(message):
-    """
-    This function is for logging to /tmp
-    - message: Message to log
-    """
-    # Setup the logger
-    logger = setup_logger()
-    # Log the message
-    logger.info(message)
+logger = setup_logger(os.path.basename(__file__), args.log_file)
+logger.info(f"Defined arguments: {args}")
 
 
 def run_command(command):
-    """
-    This function runs any command using subprocess and logs failures
-    - command: Command to execute
-    """
-    log_message(f"Running command: {command}")
-    try:
-        subprocess.run(
-            command,
-            shell=True,
-            check=True,
-            text=True,
-            stdout=subprocess.PIPE,
-            stderr=subprocess.PIPE,
-        )
-    except subprocess.CalledProcessError as e:
-        raise (e)
+    """Run command using subprocess and log execution."""
+    logger.info(f"Running command: {command}")
+    subprocess.run(
+        command,
+        shell=True,
+        check=True,
+        text=True,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+    )
 
 
 def clone_from_github(repo, tag, org):
-    """
-    This function clones from github, in-sync with build.py
-    - repo: Repo Name
-    - tag: Tag Name
-    - org: Org Name
-    """
-    # Construct the full GitHub repository URL
-    repo_url = f"https://github.com/{org}/{repo}.git"
-    # Construct the git clone command
+    """Clone repository from GitHub (in-sync with build.py)."""
+    logger.info(f"Cloning... {org}/{repo}.git@{tag}")
+    repo_url = f"{org}/{repo}.git"
+
     if tag:
         if re.match("model_navigator", repo):
             tag = "main"
 
-        if re.match("tensorrtllm_backend", repo):
-            tag = os.getenv("TENSORRTLLM_BACKEND_REPO_TAG", "main")
-            token = os.getenv("CI_JOB_TOKEN")
-            host_fqdn = os.getenv("CI_SERVER_FQDN")
-            repo_url = (
-                f"https://gitlab-ci-token:{token}@{host_fqdn}/dl/triton/{repo}.git"
-            )
-
-        clone_command = [
-            "git",
-            "clone",
-            "--branch",
-            tag,
-            "--single-branch",
-            repo_url,
-        ]
+        clone_command = ["git", "clone", "--branch", tag, "--single-branch", repo_url]
     else:
         clone_command = ["git", "clone", repo_url]
-    # Execute the git clone command
-    try:
-        subprocess.run(clone_command, check=True)
-        log_message(f"Successfully cloned {repo}")
-    except subprocess.CalledProcessError as e:
-        raise (e)
+
+    subprocess.run(clone_command, check=True)
+    logger.info(f"Successfully cloned... {org}/{repo}.git@{tag}")
 
 
 def is_excluded(file_path):
-    for exclude_pattern in exclude_patterns:
-        file_abspath = os.path.abspath(file_path)
-        exclude_pattern = os.path.abspath(exclude_pattern)
-        if os.path.commonpath([file_abspath, exclude_pattern]) == exclude_pattern:
+    """Check if file path matches any exclusion pattern."""
+    file_abspath = os.path.abspath(file_path)
+    for pattern in exclude_patterns:
+        exclude_abspath = os.path.abspath(pattern)
+        if os.path.commonpath([file_abspath, exclude_abspath]) == exclude_abspath:
             return True
     return False
 
 
-# Return the Git repo name of given file path
 def get_git_repo_name(file_path):
-    # Execute git command to get remote URL
-    try:
-        # Get the directory containing the file
-        directory = os.path.dirname(file_path)
-        # Execute git command with the file's directory as the cwd
-        remote_url = (
-            subprocess.check_output(
-                ["git", "-C", directory, "remote", "get-url", "origin"]
-            )
-            .decode()
-            .strip()
-        )
-    except subprocess.CalledProcessError as e:
-        raise (e)
+    """Return the Git repo name of given file path."""
+    directory = os.path.dirname(file_path)
+    remote_url = (
+        subprocess.check_output(["git", "-C", directory, "remote", "get-url", "origin"])
+        .decode()
+        .strip()
+    )
 
-    # Extract repository name from the remote URL.
+    # Extract repository name from the remote URL
     if remote_url.endswith(".git"):
-        # Remove '.git' extension.
         remote_url = remote_url[:-4]
-    repo_name = os.path.basename(remote_url)
-    return repo_name
+    return os.path.basename(remote_url)
 
 
 def replace_url_with_relpath(url, src_doc_path):
     """
-    This function replaces Triton Inference Server GitHub URLs with relative paths in following cases.
-    1. URL is a doc file, e.g. ".md" file.
-    2. URL is a directory which contains README.md and URL ends with "#<section>".
+    Replace Triton Inference Server GitHub URLs with relative paths for:
+    1. URL is a doc file (e.g., ".md" file).
+    2. URL is a directory with README.md and ends with "#<section>".
 
     Examples:
         https://github.com/triton-inference-server/server/blob/main/docs/protocol#restricted-protocols
         https://github.com/triton-inference-server/server/blob/main/docs/protocol/extension_shared_memory.md
         https://github.com/triton-inference-server/server/blob/main/docs/user_guide/model_configuration.md#dynamic-batcher
-
-    Keep URL in the following cases:
-        https://github.com/triton-inference-server/server/tree/r24.02
-        https://github.com/triton-inference-server/server/blob/main/build.py
-        https://github.com/triton-inference-server/server/blob/main/qa
-        https://github.com/triton-inference-server/server/blob/main/CONTRIBUTING.md
     """
     m = triton_github_url_reg.match(url)
-    # Do not replace URL if it is not a Triton GitHub file.
     if not m:
         return url
 
     target_repo_name = m.group(1)
+    logger.info(f"Found target repository: {target_repo_name}")
     target_relpath_from_target_repo = os.path.normpath(m.groups("")[1])
+    logger.info(
+        f"Found target relative path from target repository: {target_relpath_from_target_repo}"
+    )
     section = url[len(m.group(0)) :]
+    logger.info(f"Found section: {section}")
     valid_hashtag = section not in ["", "#"] and section.startswith("#")
 
-    if target_repo_name == "server":
-        target_path = os.path.join(server_abspath, target_relpath_from_target_repo)
-    else:
-        target_path = os.path.join(
+    target_path = (
+        os.path.join(server_abspath, target_relpath_from_target_repo)
+        if target_repo_name == "server"
+        else os.path.join(
             server_docs_abspath, target_repo_name, target_relpath_from_target_repo
         )
-
-    # Return URL if it points to a path outside server/docs.
+    )
+    logger.info(f"Found target path: {target_path}")
+    # Return URL if it points to a path outside server/docs
     if os.path.commonpath([server_docs_abspath, target_path]) != server_docs_abspath:
         return url
-
-    if (
+    logger.info(
+        f"Target path is under server/docs directory: {os.path.commonpath([server_docs_abspath, target_path]) == server_docs_abspath}"
+    )
+    # Check if target is valid for conversion
+    is_md_file = (
         os.path.isfile(target_path)
         and os.path.splitext(target_path)[1] == ".md"
         and not is_excluded(target_path)
-    ):
-        pass
-    elif (
+    )
+    logger.info(f"Target path is a valid .md file: {is_md_file}")
+    is_dir_with_readme = (
         os.path.isdir(target_path)
         and os.path.isfile(os.path.join(target_path, "README.md"))
         and valid_hashtag
         and not is_excluded(os.path.join(target_path, "README.md"))
-    ):
+    )
+    logger.info(f"Target path is a directory with README.md: {is_dir_with_readme}")
+    if is_md_file:
+        pass
+    elif is_dir_with_readme:
         target_path = os.path.join(target_path, "README.md")
     else:
         return url
+    logger.info(
+        f"Target path is a valid .md file or a directory with README.md: {is_md_file or is_dir_with_readme}"
+    )
 
-    # The "target_path" must be a file at this line.
     relpath = os.path.relpath(target_path, start=os.path.dirname(src_doc_path))
+    logger.info(f"Found relative path: {relpath}")
     return re.sub(triton_github_url_reg, relpath, url, 1)
 
 
@@ -274,32 +266,26 @@ def replace_relpath_with_url(relpath, src_doc_path):
     Examples:
         ../examples/model_repository
         ../examples/model_repository/inception_graphdef/config.pbtxt
-
-    Keep relpath in the following cases:
-        build.md
-        build.md#building-with-docker
-        #building-with-docker
-        ../getting_started/quickstart.md
-        ../protocol#restricted-protocols
     """
-    target_path = relpath.rsplit("#")[0]
+    target_path = relpath.rsplit("#", 1)[0]
     section = relpath[len(target_path) :]
     valid_hashtag = section not in ["", "#"]
+
     if relpath.startswith("#"):
         target_path = os.path.basename(src_doc_path)
-    target_path = os.path.join(os.path.dirname(src_doc_path), target_path)
-    target_path = os.path.normpath(target_path)
+
+    target_path = os.path.normpath(
+        os.path.join(os.path.dirname(src_doc_path), target_path)
+    )
     src_git_repo_name = get_git_repo_name(src_doc_path)
 
-    url = f"https://github.com/triton-inference-server/{src_git_repo_name}/blob/main/"
-    if src_git_repo_name == "server":
-        src_repo_abspath = server_abspath
-        # TODO: Assert the relative path not pointing to cloned repo, e.g. client.
-        # This requires more information which may be stored in a global variable.
-    else:
-        src_repo_abspath = os.path.join(server_docs_abspath, src_git_repo_name)
+    src_repo_abspath = (
+        server_abspath
+        if src_git_repo_name == "server"
+        else os.path.join(server_docs_abspath, src_git_repo_name)
+    )
 
-    # Assert target path is under the current repo directory.
+    # Assert target path is under the current repo directory
     assert os.path.commonpath([src_repo_abspath, target_path]) == src_repo_abspath
 
     target_path_from_src_repo = os.path.relpath(target_path, start=src_repo_abspath)
@@ -310,9 +296,10 @@ def replace_relpath_with_url(relpath, src_doc_path):
         and valid_hashtag
         and os.path.isfile(os.path.join(target_path, "README.md"))
     ):
-        relpath = os.path.join(relpath.rsplit("#")[0], "README.md") + section
+        relpath = os.path.join(relpath.rsplit("#", 1)[0], "README.md") + section
         target_path = os.path.join(target_path, "README.md")
 
+    # Keep relpath if it's a valid .md file in docs
     if (
         os.path.isfile(target_path)
         and os.path.splitext(target_path)[1] == ".md"
@@ -321,53 +308,37 @@ def replace_relpath_with_url(relpath, src_doc_path):
         and not is_excluded(target_path)
     ):
         return relpath
-    else:
-        return url + target_path_from_src_repo + section
+
+    return f"https://github.com/triton-inference-server/{src_git_repo_name}/blob/main/{target_path_from_src_repo}{section}"
 
 
 def replace_hyperlink(m, src_doc_path):
     """
-    TODO: Support of HTML tags for future docs.
-    Markdown allows <link>, e.g. <a href=[^>]+>. Whether we want to
-    find and replace the link depends on if they link to internal .md files
-    or allows relative paths. I haven't seen one such case in our doc so
-    should be safe for now.
+    Replace hyperlinks in markdown files.
+    TODO: Support HTML tags for future docs (e.g., <a href=...>).
     """
-
     hyperlink_str = m.group(2)
-    match = http_reg.match(hyperlink_str)
-
-    if match:
-        # Hyperlink is a URL.
-        res = replace_url_with_relpath(hyperlink_str, src_doc_path)
-    else:
-        # Hyperlink is a relative path.
-        res = replace_relpath_with_url(hyperlink_str, src_doc_path)
-
+    res = (
+        replace_url_with_relpath(hyperlink_str, src_doc_path)
+        if http_reg.match(hyperlink_str)
+        else replace_relpath_with_url(hyperlink_str, src_doc_path)
+    )
     return m.group(1) + res + m.group(3)
 
 
-def preprocess_docs(exclude_paths=[]):
-    # Find all ".md" files inside the current repo.
-    if exclude_paths:
-        cmd = (
-            ["find", server_docs_abspath, "-type", "d", "\\("]
-            + " -o ".join([f"-path './{dir}'" for dir in exclude_paths]).split(" ")
-            + ["\\)", "-prune", "-o", "-type", "f", "-name", "'*.md'", "-print"]
-        )
-    else:
-        cmd = ["find", server_docs_abspath, "-name", "'*.md'"]
-    cmd = " ".join(cmd)
+def preprocess_docs(exclude_paths=None):
+    """Find all .md files and preprocess their hyperlinks."""
+    # Find all ".md" files
+    cmd = f"find {server_docs_abspath} -name '*.md'"
     result = subprocess.run(cmd, check=True, capture_output=True, text=True, shell=True)
-    docs_list = list(filter(None, result.stdout.split("\n")))
+    docs_list = [path for path in result.stdout.split("\n") if path]
 
-    # Read, preprocess and write back to each document file.
+    # Read, preprocess and write back to each document file
     for doc_abspath in docs_list:
         if is_excluded(doc_abspath):
             continue
 
-        content = None
-        with open(doc_abspath, "r") as f:
+        with open(doc_abspath) as f:
             content = f.read()
 
         content = hyperlink_reg.sub(
@@ -380,35 +351,21 @@ def preprocess_docs(exclude_paths=[]):
 
 
 def main():
-    args = parser.parse_args()
-    repo_tag = args.repo_tag
-    repository_filename = args.repo_file
-    github_org = args.github_organization
-
-    # Change working directory to server/docs.
+    """Main function to clone repositories, preprocess docs, and build HTML."""
+    logger.info("Starting setup Triton Server documentation for Sphinx build...")
+    logger.info(f"Collecting repositories from {args.repo_file}...")
     os.chdir(server_docs_abspath)
-    run_command("make clean")
 
-    repositories = ""
-    with open(repository_filename, "r") as f:
-        repositories = f.read()
-        f.close()
+    with open(args.repo_file) as f:
+        repository_list = f.read().strip().split("\n")
 
-    repository_list = repositories.strip().split("\n")
+    # Clone repositories
     for repository in repository_list:
         run_command(f"rm -rf {repository}")
-        clone_from_github(repository, repo_tag, github_org)
+        clone_from_github(repository, args.repo_tag, args.github_organization)
 
-    # Preprocess documents in server_docs_abspath after all repos are cloned.
+    # Preprocess documents after all repos are cloned
     preprocess_docs()
-    run_command("make html")
-
-    # Clean up working directory.
-    for repository in repository_list:
-        run_command(f"rm -rf {repository}")
-
-    # Return to previous working directory server/.
-    os.chdir(server_abspath)
 
 
 if __name__ == "__main__":