mindstudio-probe 8.3.3__py3-none-any.whl → 26.0.0a1__py3-none-any.whl

msprobe/visualization/graph_service.py

@@ -1,36 +1,43 @@
-# Copyright (c) 2024-2025, Huawei Technologies Co., Ltd.
-# All rights reserved.
+# -------------------------------------------------------------------------
+#  This file is part of the MindStudio project.
+# Copyright (c) 2025 Huawei Technologies Co.,Ltd.
 #
-# Licensed under the Apache License, Version 2.0  (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
+# MindStudio is licensed under Mulan PSL v2.
+# You can use this software according to the terms and conditions of the Mulan PSL v2.
+# You may obtain a copy of Mulan PSL v2 at:
 #
-# http://www.apache.org/licenses/LICENSE-2.0
+#          http://license.coscl.org.cn/MulanPSL2
 #
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
+# THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND,
+# EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT,
+# MERCHANTABILITY OR FIT FOR A PARTICULAR PURPOSE.
+# See the Mulan PSL v2 for more details.
+# -------------------------------------------------------------------------
 
 import os
 import time
+import threading
 from copy import deepcopy
-from multiprocessing import cpu_count, Pool
+from dataclasses import dataclass
+from multiprocessing import cpu_count, Pool, Manager
+from typing import Callable, Optional
+
+from tqdm import tqdm
 from msprobe.core.common.file_utils import (check_file_type, create_directory, FileChecker,
                                             check_file_or_directory_path, load_json)
 from msprobe.core.common.const import FileCheckConst, Const
 from msprobe.core.common.utils import CompareException, get_dump_mode
 from msprobe.visualization.compare.graph_comparator import GraphComparator
 from msprobe.visualization.utils import GraphConst, check_directory_content, SerializableArgs, load_parallel_param, \
-    sort_rank_number_strings, check_whether_parallel_merge, validate_parallel_param, get_step_or_rank_int
+    sort_rank_number_strings, validate_parallel_param, get_step_or_rank_int, \
+    monitor_progress, ProgressInfo, calculate_list, get_log_msg_wrapper
 from msprobe.visualization.builder.graph_builder import GraphBuilder, GraphExportConfig, GraphInfo, BuildGraphTaskInfo
-from msprobe.core.common.log import logger
+from msprobe.core.common.log import logger, BaseLogger
 from msprobe.visualization.graph.node_colors import NodeColors
 from msprobe.core.compare.layer_mapping import generate_api_mapping_by_layer_mapping
 from msprobe.core.compare.utils import check_and_return_dir_contents
 from msprobe.core.common.utils import detect_framework_by_dump_json
-from msprobe.visualization.graph.distributed_analyzer import DistributedAnalyzer
+from msprobe.visualization.graph.distributed_analyzer import distributed_analyse
 from msprobe.visualization.builder.graph_merger import GraphMerger
 from msprobe.visualization.db_utils import post_process_db
 
@@ -39,12 +46,12 @@ build_output_db_name = f'build_{current_time}.vis.db'
 compare_output_db_name = f'compare_{current_time}.vis.db'
 
 
-def _compare_graph(graph_n: GraphInfo, graph_b: GraphInfo, input_param, args):
+def _compare_graph(graph_n: GraphInfo, graph_b: GraphInfo, input_param, args, pbar_info=None):
     dump_path_param = {
-        'npu_json_path': graph_n.data_path,
-        'bench_json_path': graph_b.data_path,
-        'stack_json_path': graph_n.stack_path,
-        'is_print_compare_log': input_param.get("is_print_compare_log", True)
+        'npu_path': graph_n.data_path,
+        'bench_path': graph_b.data_path,
+        'stack_path': graph_n.stack_path,
+        'is_print_compare_log': input_param.get("is_print_compare_log", False)
     }
     mapping_dict = {}
     if args.layer_mapping:
@@ -61,19 +68,19 @@ def _compare_graph(graph_n: GraphInfo, graph_b: GraphInfo, input_param, args):
         raise CompareException(CompareException.CROSS_FRAME_ERROR)
 
     graph_comparator = GraphComparator([graph_n.graph, graph_b.graph], dump_path_param, args, is_cross_framework,
-                                       mapping_dict=mapping_dict)
+                                       mapping_dict=mapping_dict, pbar_info=pbar_info)
     graph_comparator.compare()
     return graph_comparator
 
 
-def _compare_graph_result(input_param, args):
+def _compare_graph_result(input_param, args, pbar_info=None):
     logger.info('Start building model graphs...')
     # 对两个数据进行构图
-    graph_n = _build_graph_info(input_param.get('npu_path'), args)
-    graph_b = _build_graph_info(input_param.get('bench_path'), args)
+    graph_n = _build_graph_info(input_param.get('npu_path'), args, pbar_info=pbar_info)
+    graph_b = _build_graph_info(input_param.get('bench_path'), args, pbar_info=pbar_info)
     logger.info('Model graphs built successfully, start comparing graphs...')
     # 基于graph、stack和data进行比较
-    graph_comparator = _compare_graph(graph_n, graph_b, input_param, args)
+    graph_comparator = _compare_graph(graph_n, graph_b, input_param, args, pbar_info=pbar_info)
     # 增加micro step标记
     micro_steps = graph_n.graph.paging_by_micro_step(graph_b.graph)
     # 开启溢出检测
@@ -84,7 +91,7 @@ def _compare_graph_result(input_param, args):
     return CompareGraphResult(graph_n.graph, graph_b.graph, graph_comparator, micro_steps)
 
 
-def _export_compare_graph_result(args, result):
+def _export_compare_graph_result(args, result, pbar_info=None):
     graphs = [result.graph_n, result.graph_b]
     graph_comparator = result.graph_comparator
     micro_steps = result.micro_steps
@@ -97,7 +104,7 @@ def _export_compare_graph_result(args, result):
                                       args.step_list if hasattr(args, 'step_list') else [0],
                                       args.rank_list if hasattr(args, 'rank_list') else [0])
     try:
-        GraphBuilder.to_db(output_db_path, export_config)
+        GraphBuilder.to_db(output_db_path, export_config, pbar_info=pbar_info)
         logger.info(f'Exporting compare graph result successfully, the result file is saved in {output_db_path}')
         return ''
     except RuntimeError as e:
@@ -105,7 +112,7 @@ def _export_compare_graph_result(args, result):
         return compare_output_db_name
 
 
-def _build_graph_info(dump_path, args, graph=None):
+def _build_graph_info(dump_path, args, graph=None, pbar_info=None):
     construct_path = FileChecker(os.path.join(dump_path, GraphConst.CONSTRUCT_FILE), FileCheckConst.FILE,
                                  FileCheckConst.READ_ABLE).common_check()
     data_path = FileChecker(os.path.join(dump_path, GraphConst.DUMP_FILE), FileCheckConst.FILE,
@@ -113,13 +120,13 @@ def _build_graph_info(dump_path, args, graph=None):
     stack_path = FileChecker(os.path.join(dump_path, GraphConst.STACK_FILE), FileCheckConst.FILE,
                              FileCheckConst.READ_ABLE).common_check()
     if not graph:
-        graph = GraphBuilder.build(construct_path, data_path, stack_path)
+        graph = GraphBuilder.build(construct_path, data_path, stack_path, pbar_info=pbar_info)
     return GraphInfo(graph, construct_path, data_path, stack_path)
 
 
-def _build_graph_result(dump_path, args):
+def _build_graph_result(dump_path, args, pbar_info=None):
     logger.info('Start building model graphs...')
-    graph = _build_graph_info(dump_path, args).graph
+    graph = _build_graph_info(dump_path, args, pbar_info=pbar_info).graph
     # 增加micro step标记
     micro_steps = graph.paging_by_micro_step()
     # 开启溢出检测
@@ -128,30 +135,39 @@ def _build_graph_result(dump_path, args):
     return BuildGraphResult(graph, micro_steps)
 
 
-def _run_build_graph_compare(input_param, args, nr, br):
+def _run_build_graph_compare(input_param, args, nr, br, pbar_info=None):
     logger.info(f'Start building graph for {nr}...')
-    graph_n = _build_graph_info(input_param.get('npu_path'), args)
-    graph_b = _build_graph_info(input_param.get('bench_path'), args)
+    graph_n = _build_graph_info(input_param.get('npu_path'), args, pbar_info=pbar_info)
+    graph_b = _build_graph_info(input_param.get('bench_path'), args, pbar_info=pbar_info)
     logger.info(f'Building graph for {nr} finished.')
     return BuildGraphTaskInfo(graph_n, graph_b, nr, br, current_time)
 
 
-def _run_build_graph_single(dump_ranks_path, rank, step, args):
+def _run_build_graph_single(dump_ranks_path, rank, step, args, pbar_info=None):
     logger.info(f'Start building graph for {rank}...')
     dump_path = os.path.join(dump_ranks_path, rank)
-    result = _build_graph_result(dump_path, args)
+    result = _build_graph_result(dump_path, args, pbar_info=pbar_info)
     if rank != Const.RANK:
         result.rank = get_step_or_rank_int(rank, True)
     logger.info(f'Building graph for step: {step}, rank: {rank} finished.')
     return result
 
 
-def _run_graph_compare(graph_task_info, input_param, args):
+def _run_build_graph_and_export(dump_ranks_path, rank, step, args, pbar_info=None):
+    result = _run_build_graph_single(dump_ranks_path, rank, step, args, pbar_info)
+    if step is not None:
+        result.step = get_step_or_rank_int(step)
+    create_directory(args.output_path)
+
+    return _export_build_graph_result(args, result, pbar_info)
+
+
+def _run_graph_compare(graph_task_info, input_param, args, pbar_info=None):
     logger.info(f'Start comparing data for {graph_task_info.npu_rank}...')
     graph_n = graph_task_info.graph_info_n
     graph_b = graph_task_info.graph_info_b
     nr = graph_task_info.npu_rank
-    graph_comparator = _compare_graph(graph_n, graph_b, input_param, args)
+    graph_comparator = _compare_graph(graph_n, graph_b, input_param, args, pbar_info=pbar_info)
     micro_steps = graph_n.graph.paging_by_micro_step(graph_b.graph)
     # 开启溢出检测
     if args.overflow_check:
@@ -164,7 +180,7 @@ def _run_graph_compare(graph_task_info, input_param, args):
     return graph_result
 
 
-def _export_build_graph_result(args, result):
+def _export_build_graph_result(args, result, pbar_info=None):
     out_path = args.output_path
     graph = result.graph
     micro_steps = result.micro_steps
@@ -175,7 +191,7 @@ def _export_build_graph_result(args, result):
                                step=result.step, rank_list=args.rank_list if hasattr(args, 'rank_list') else [0],
                                step_list=args.step_list if hasattr(args, 'step_list') else [0])
     try:
-        GraphBuilder.to_db(output_db_path, config)
+        GraphBuilder.to_db(output_db_path, config, pbar_info=pbar_info)
         logger.info(f'Model graph exported successfully, the result file is saved in {output_db_path}')
         return None
     except RuntimeError as e:
@@ -189,57 +205,30 @@ def is_real_data_compare(input_param, npu_ranks, bench_ranks):
     has_real_data = False
     for nr, br in zip(npu_ranks, bench_ranks):
         dump_path_param = {
-            'npu_json_path': FileChecker(os.path.join(dump_rank_n, nr, GraphConst.DUMP_FILE), FileCheckConst.FILE,
-                                         FileCheckConst.READ_ABLE).common_check(),
-            'bench_json_path': FileChecker(os.path.join(dump_rank_b, br, GraphConst.DUMP_FILE), FileCheckConst.FILE,
-                                           FileCheckConst.READ_ABLE).common_check()
+            'npu_path': FileChecker(os.path.join(dump_rank_n, nr, GraphConst.DUMP_FILE), FileCheckConst.FILE,
+                                    FileCheckConst.READ_ABLE).common_check(),
+            'bench_path': FileChecker(os.path.join(dump_rank_b, br, GraphConst.DUMP_FILE), FileCheckConst.FILE,
+                                      FileCheckConst.READ_ABLE).common_check()
         }
         has_real_data |= get_dump_mode(dump_path_param) == Const.ALL
     return has_real_data
 
 
-def _mp_compare(input_param, serializable_args, nr, br):
-    graph_task_info = _run_build_graph_compare(input_param, serializable_args, nr, br)
-    return _run_graph_compare(graph_task_info, input_param, serializable_args)
-
-
-def _compare_graph_ranks(input_param, args, step=None):
-    with Pool(processes=max(int((cpu_count() + 1) // 4), 1)) as pool:
-        def err_call(err):
-            logger.error(f'Error occurred while comparing graph ranks: {err}')
-            try:
-                pool.close()
-            except OSError as e:
-                logger.error(f'Error occurred while terminating the pool: {e}')
-
-        serializable_args = SerializableArgs(args)
-        # 暂存所有rank的graph，用于匹配rank间的分布式节点
-        compare_graph_results = _get_compare_graph_results(input_param, serializable_args, step, pool, err_call)
+def _mp_compare(input_param, serializable_args, nr, br, pbar_info=None):
+    graph_task_info = _run_build_graph_compare(input_param, serializable_args, nr, br, pbar_info=pbar_info)
+    return _run_graph_compare(graph_task_info, input_param, serializable_args, pbar_info=pbar_info)
 
-        serializable_args.rank_list = [result.rank for result in compare_graph_results]
 
-        # 匹配rank间的分布式节点
-        if len(compare_graph_results) > 1:
-            DistributedAnalyzer({obj.rank: obj.graph_n for obj in compare_graph_results},
-                                args.overflow_check).distributed_match()
-            DistributedAnalyzer({obj.rank: obj.graph_b for obj in compare_graph_results},
-                                args.overflow_check).distributed_match()
+def _mp_compare_and_export(input_param, args, rank, step, pbar_info=None):
+    graph_result = _mp_compare(input_param, args, rank, rank, pbar_info=pbar_info)
+    if step is not None:
+        graph_result.step = get_step_or_rank_int(step)
+    create_directory(args.output_path)
 
-        export_res_task_list = []
-        create_directory(args.output_path)
-        for result in compare_graph_results:
-            export_res_task_list.append(pool.apply_async(_export_compare_graph_result,
-                                                         args=(serializable_args, result),
-                                                         error_callback=err_call))
-        export_res_list = [res.get() for res in export_res_task_list]
-        if any(export_res_list):
-            failed_names = list(filter(lambda x: x, export_res_list))
-            logger.error(f'Unable to export compare graph results: {", ".join(failed_names)}.')
-        else:
-            logger.info('Successfully exported compare graph results.')
+    return _export_compare_graph_result(args, graph_result, pbar_info=pbar_info)
 
 
-def _get_compare_graph_results(input_param, serializable_args, step, pool, err_call):
+def _compare_graph_ranks(input_param, args, step=None, pbar_info=None):
     dump_rank_n = input_param.get('npu_path')
     dump_rank_b = input_param.get('bench_path')
     npu_ranks = sort_rank_number_strings(check_and_return_dir_contents(dump_rank_n, Const.RANK))
@@ -251,64 +240,86 @@ def _get_compare_graph_results(input_param, serializable_args, step, pool, err_c
             raise CompareException(CompareException.INVALID_PATH_ERROR)
         npu_ranks = intersection_ranks
         bench_ranks = intersection_ranks
-    compare_graph_results = []
-    if is_real_data_compare(input_param, npu_ranks, bench_ranks):
-        mp_task_dict = {}
-        for nr, br in zip(npu_ranks, bench_ranks):
-            input_param['npu_path'] = os.path.join(dump_rank_n, nr)
-            input_param['bench_path'] = os.path.join(dump_rank_b, br)
-            build_key = f'{step}_{nr}' if step else f'{nr}'
-            input_param_copy = deepcopy(input_param)
-            mp_task_dict[build_key] = pool.apply_async(_run_build_graph_compare,
-                                                       args=(input_param_copy, serializable_args, nr, br),
-                                                       error_callback=err_call)
-
-        mp_res_dict = {k: v.get() for k, v in mp_task_dict.items()}
-        for mp_res in mp_res_dict.values():
-            compare_graph_results.append(_run_graph_compare(mp_res, input_param, serializable_args))
-    else:
-        compare_graph_tasks = []
-        for nr, br in zip(npu_ranks, bench_ranks):
-            input_param['npu_path'] = os.path.join(dump_rank_n, nr)
-            input_param['bench_path'] = os.path.join(dump_rank_b, br)
-            input_param_copy = deepcopy(input_param)
-            compare_graph_tasks.append(pool.apply_async(_mp_compare,
-                                                        args=(input_param_copy, serializable_args, nr, br),
-                                                        error_callback=err_call))
-        compare_graph_results = [task.get() for task in compare_graph_tasks]
-    if step is not None:
-        for result in compare_graph_results:
-            result.step = get_step_or_rank_int(step)
-    return compare_graph_results
+    args.rank_list = [get_step_or_rank_int(rank, True) for rank in npu_ranks]
+    serializable_args = SerializableArgs(args)
 
+    with Pool(processes=max(int((cpu_count() + 1) // 4), 1)) as pool:
+        def err_call(err):
+            logger.error(f'Error occurred while comparing graph ranks: {err}')
 
-def _compare_graph_steps(input_param, args):
-    dump_step_n = input_param.get('npu_path')
-    dump_step_b = input_param.get('bench_path')
+        if is_real_data_compare(input_param, npu_ranks, bench_ranks):
+            # 真实数据模式，考虑到tensor比对过程会使用进程池启用多进程，为了避免嵌套进程池，graph比对使用串行
+            compare_graph_results = []
+            mp_task_dict = {}
+            for nr, br in zip(npu_ranks, bench_ranks):
+                input_param['npu_path'] = os.path.join(dump_rank_n, nr)
+                input_param['bench_path'] = os.path.join(dump_rank_b, br)
+                build_key = f'{step}_{nr}' if step else f'{nr}'
+                input_param_copy = deepcopy(input_param)
+                pbar_info_copy = PbarInfo.update_task_id(pbar_info, nr)
+                mp_task_dict[build_key] = pool.apply_async(_run_build_graph_compare,
+                                                           args=(input_param_copy, serializable_args, nr, br,
+                                                                 pbar_info_copy),
+                                                           error_callback=err_call)
+            mp_res_dict = {k: v.get() for k, v in mp_task_dict.items()}
+            for build_key, mp_res in mp_res_dict.items():
+                if pbar_info:
+                    if Const.REPLACEMENT_CHARACTER in build_key:
+                        build_key = build_key.split(Const.REPLACEMENT_CHARACTER)[-1]
+                    pbar_info.task_id = build_key
+                compare_graph_results.append(_run_graph_compare(mp_res, input_param, serializable_args, pbar_info))
+            if step is not None:
+                for result in compare_graph_results:
+                    result.step = get_step_or_rank_int(step)
+
+            export_res_task_list = []
+            create_directory(args.output_path)
+            for result in compare_graph_results:
+                export_res_task_list.append(pool.apply_async(_export_compare_graph_result,
+                                                             args=(serializable_args, result),
+                                                             error_callback=err_call))
+            export_res_list = [res.get() for res in export_res_task_list]
+        else:
+            compare_graph_tasks = []
+            for nr, br in zip(npu_ranks, bench_ranks):
+                input_param['npu_path'] = os.path.join(dump_rank_n, nr)
+                input_param['bench_path'] = os.path.join(dump_rank_b, br)
+                input_param_copy = deepcopy(input_param)
+                pbar_info_copy = PbarInfo.update_task_id(pbar_info, nr)
+                compare_graph_tasks.append(pool.apply_async(_mp_compare_and_export,
+                                                            args=(input_param_copy, serializable_args, nr, step,
+                                                                  pbar_info_copy),
+                                                            error_callback=err_call))
+            export_res_list = [res.get() for res in compare_graph_tasks]
+        if any(export_res_list):
+            failed_names = list(filter(lambda x: x, export_res_list))
+            logger.error(f'Unable to export compare graph results: {", ".join(failed_names)}.')
+        else:
+            logger.info('Successfully exported compare graph results.')
 
-    npu_steps = check_and_return_dir_contents(dump_step_n, Const.STEP)
-    bench_steps = check_and_return_dir_contents(dump_step_b, Const.STEP)
 
-    if npu_steps != bench_steps:
-        intersection_steps = sort_rank_number_strings(list(set(npu_steps) & set(bench_steps)))
+def _compare_graph_steps(input_param, args, pbar_info=None):
+    dump_step_n = input_param.get('npu_path')
+    dump_step_b = input_param.get('bench_path')
 
-        if not intersection_steps:
-            logger.error('The steps in the two runs are completely different. Unable to match the steps.')
-            raise CompareException(CompareException.INVALID_PATH_ERROR)
-        npu_steps = intersection_steps
+    npu_steps = calculate_list(dump_step_n, dump_step_b, Const.STEP)
 
     args.step_list = sorted([get_step_or_rank_int(step) for step in npu_steps])
 
-    for folder_step in npu_steps:
+    for i, folder_step in enumerate(npu_steps):
         logger.info(f'Start processing data for {folder_step}...')
         input_param['npu_path'] = os.path.join(dump_step_n, folder_step)
         input_param['bench_path'] = os.path.join(dump_step_b, folder_step)
 
-        _compare_graph_ranks(input_param, args, step=folder_step) if not args.parallel_merge \
-            else _compare_graph_ranks_parallel(input_param, args, step=folder_step)
+        if pbar_info:
+            pbar_info.step = i
 
+        _compare_graph_ranks(input_param, args, step=folder_step, pbar_info=pbar_info) if not args.parallel_merge \
+            else _compare_graph_ranks_parallel(input_param, args, step=folder_step, pbar_info=pbar_info)
 
-def _build_graph_ranks(dump_ranks_path, args, step=None):
+
+def _build_graph_ranks_parallel(args, step=None, pbar_info=None):
+    dump_ranks_path = os.path.join(args.target_path, step) if step is not None else args.target_path
     ranks = sort_rank_number_strings(check_and_return_dir_contents(dump_ranks_path, Const.RANK))
     serializable_args = SerializableArgs(args)
     with Pool(processes=max(int((cpu_count() + 1) // 4), 1)) as pool:
@@ -320,10 +331,13 @@ def _build_graph_ranks(dump_ranks_path, args, step=None):
                 logger.error(f'Error occurred while terminating the pool: {e}')
 
         build_graph_tasks = []
+        if pbar_info and pbar_info.step:
+            PbarInfo.reset_progress_and_current_stage(pbar_info, ranks)
         for rank in ranks:
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, rank)
             build_graph_tasks.append(pool.apply_async(_run_build_graph_single,
-                                                      args=(dump_ranks_path, rank, step, serializable_args),
-                                                      error_callback=err_call))
+                                                      args=(dump_ranks_path, rank, step, serializable_args,
+                                                            pbar_info_copy), error_callback=err_call))
         build_graph_results = [task.get() for task in build_graph_tasks]
 
         if step is not None:
@@ -332,18 +346,19 @@ def _build_graph_ranks(dump_ranks_path, args, step=None):
 
         if args.parallel_params:
             validate_parallel_param(args.parallel_params[0], dump_ranks_path)
-            build_graph_results = GraphMerger(build_graph_results, args.parallel_params[0]).merge_graph()
-
-        if len(build_graph_results) > 1 and not args.parallel_merge:
-            DistributedAnalyzer({obj.rank: obj.graph for obj in build_graph_results},
-                                args.overflow_check).distributed_match()
+            build_graph_results = GraphMerger(build_graph_results, args.parallel_params[0],
+                                              pbar_info=pbar_info).merge_graph()
+            if pbar_info:
+                PbarInfo.del_progress_dict_item(pbar_info, ranks,
+                                                [f'{Const.RANK}{result.rank}' for result in build_graph_results])
 
         create_directory(args.output_path)
         export_build_graph_tasks = []
         serializable_args.rank_list = [result.rank for result in build_graph_results]
         for result in build_graph_results:
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, f'{Const.RANK}{result.rank}')
             export_build_graph_tasks.append(pool.apply_async(_export_build_graph_result,
-                                                             args=(serializable_args, result),
+                                                             args=(serializable_args, result, pbar_info_copy),
                                                              error_callback=err_call))
         export_build_graph_result = [task.get() for task in export_build_graph_tasks]
         if any(export_build_graph_result):
@@ -353,30 +368,55 @@ def _build_graph_ranks(dump_ranks_path, args, step=None):
             logger.info(f'Successfully exported build graph results.')
 
 
-def _build_graph_steps(dump_steps_path, args):
-    steps = sorted(check_and_return_dir_contents(dump_steps_path, Const.STEP))
+def _build_graph_ranks(args, step=None, pbar_info=None):
+    dump_ranks_path = os.path.join(args.target_path, step) if step is not None else args.target_path
+    ranks = sort_rank_number_strings(check_and_return_dir_contents(dump_ranks_path, Const.RANK))
+    args.rank_list = [get_step_or_rank_int(rank, True) for rank in ranks]
+    serializable_args = SerializableArgs(args)
+    with Pool(processes=max(int((cpu_count() + 1) // 4), 1)) as pool:
+        def err_call(err):
+            logger.error(f'Error occurred while comparing graph ranks: {err}')
+
+        tasks = []
+        for rank in ranks:
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, rank)
+            tasks.append(pool.apply_async(_run_build_graph_and_export,
+                                          args=(dump_ranks_path, rank, step, serializable_args, pbar_info_copy),
+                                          error_callback=err_call))
+        results = [task.get() for task in tasks]
+        if any(results):
+            failed_names = list(filter(lambda x: x, results))
+            logger.error(f'Unable to export build graph results: {failed_names}.')
+        else:
+            logger.info(f'Successfully exported build graph results.')
+
+
+def _build_graph_steps(args, pbar_info=None):
+    steps = sorted(check_and_return_dir_contents(args.target_path, Const.STEP))
     args.step_list = sorted([get_step_or_rank_int(step) for step in steps])
 
-    for step in steps:
+    for i, step in enumerate(steps):
         logger.info(f'Start processing data for {step}...')
-        dump_ranks_path = os.path.join(dump_steps_path, step)
-        _build_graph_ranks(dump_ranks_path, args, step)
+        if pbar_info:
+            pbar_info.step = i
+        _build_graph_ranks(args, step, pbar_info=pbar_info) if not args.parallel_merge \
+            else _build_graph_ranks_parallel(args, step, pbar_info=pbar_info)
 
 
-def _compare_and_export_graph(graph_task_info, input_param, args, step=None):
-    result = _run_graph_compare(graph_task_info, input_param, args)
+def _compare_and_export_graph(graph_task_info, input_param, args, step=None, pbar_info=None):
+    result = _run_graph_compare(graph_task_info, input_param, args, pbar_info=pbar_info)
     if step is not None:
         result.step = get_step_or_rank_int(step)
-    return _export_compare_graph_result(args, result)
+    return _export_compare_graph_result(args, result, pbar_info=pbar_info)
 
 
-def _compare_graph_ranks_parallel(input_param, args, step=None):
+def _compare_graph_ranks_parallel(input_param, args, step=None, pbar_info=None):
     args.fuzzy_match = True
     npu_path = input_param.get('npu_path')
     bench_path = input_param.get('bench_path')
     ranks_n = sort_rank_number_strings(check_and_return_dir_contents(npu_path, Const.RANK))
     ranks_b = sort_rank_number_strings(check_and_return_dir_contents(bench_path, Const.RANK))
-    parallel_params = load_parallel_param(input_param)
+    parallel_params = args.parallel_params
     if len(parallel_params) != 2:
         raise RuntimeError('Parallel params error in compare graph!')
     validate_parallel_param(parallel_params[0], npu_path)
@@ -394,24 +434,33 @@ def _compare_graph_ranks_parallel(input_param, args, step=None):
         # 1.并行构图
         build_graph_tasks_n = []
         build_graph_tasks_b = []
+        if pbar_info and pbar_info.step:
+            PbarInfo.reset_progress_and_current_stage(pbar_info, list(set(ranks_n) | set(ranks_b)))
         for rank in ranks_n:
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, rank)
             build_graph_tasks_n.append(pool.apply_async(_run_build_graph_single,
-                                                        args=(npu_path, rank, step, serializable_args),
+                                                        args=(npu_path, rank, step, serializable_args, pbar_info_copy),
                                                         error_callback=err_call))
         for rank in ranks_b:
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, rank)
             build_graph_tasks_b.append(pool.apply_async(_run_build_graph_single,
-                                                        args=(bench_path, rank, step, serializable_args),
-                                                        error_callback=err_call))
+                                                        args=(bench_path, rank, step, serializable_args,
+                                                              pbar_info_copy), error_callback=err_call))
         graph_results_n = [task.get() for task in build_graph_tasks_n]
         graph_results_b = [task.get() for task in build_graph_tasks_b]
 
         # 2.图合并
-        build_graph_results_n = GraphMerger(graph_results_n, parallel_params[0]).merge_graph()
-        build_graph_results_b = GraphMerger(graph_results_b, parallel_params[1], True).merge_graph()
+        build_graph_results_n = GraphMerger(graph_results_n, parallel_params[0], pbar_info=pbar_info).merge_graph()
+        build_graph_results_b = GraphMerger(graph_results_b, parallel_params[1], True,
+                                            pbar_info=pbar_info).merge_graph()
+
         if len(build_graph_results_n) != len(build_graph_results_b):
             raise RuntimeError(f'Parallel merge failed because the dp of npu: {len(build_graph_results_n)} '
                                f'is inconsistent with that of bench: {len(build_graph_results_b)}!')
         serializable_args.rank_list = [result.rank for result in build_graph_results_n]
+        if pbar_info:
+            PbarInfo.del_progress_dict_item(pbar_info, list(set(ranks_n) | set(ranks_b)),
+                                            [f'{Const.RANK}{result.rank}' for result in build_graph_results_n])
         # 3.并行图比对和输出
         export_res_task_list = []
         create_directory(args.output_path)
@@ -422,9 +471,10 @@ def _compare_graph_ranks_parallel(input_param, args, step=None):
                 _build_graph_info(os.path.join(npu_path, f'rank{graph_n.root.rank}'), args, graph_n),
                 _build_graph_info(os.path.join(bench_path, f'rank{graph_b.root.rank}'), args, graph_b),
                 f'rank{graph_n.root.rank}', f'rank{graph_b.root.rank}', current_time)
+            pbar_info_copy = PbarInfo.update_task_id(pbar_info, f'{Const.RANK}{result_n.rank}')
             export_res_task_list.append(pool.apply_async(_compare_and_export_graph,
-                                                         args=(graph_task_info, input_param, serializable_args, step),
-                                                         error_callback=err_call))
+                                                         args=(graph_task_info, input_param, serializable_args, step,
+                                                               pbar_info_copy), error_callback=err_call))
         export_res_list = [res.get() for res in export_res_task_list]
         if any(export_res_list):
             failed_names = list(filter(lambda x: x, export_res_list))
@@ -434,80 +484,336 @@ def _compare_graph_ranks_parallel(input_param, args, step=None):
 
 
 def _graph_service_parser(parser):
-    parser.add_argument("-i", "--input_path", dest="input_path", type=str,
-                        help="<Required> The compare input path, a dict json.", required=True)
+    # -------------------------- 基础必填参数 --------------------------
+    parser.add_argument("-tp", "--target_path", dest="target_path", type=str,
+                        help="<Required> The target path.", required=True)
     parser.add_argument("-o", "--output_path", dest="output_path", type=str,
-                        help="<Required> The compare task result out path.", required=True)
+                        help="<Required> The visualization task result out path.", required=True)
+    # -------------------------- 基础可选参数 --------------------------
+    parser.add_argument("-gp", "--golden_path", dest="golden_path", type=str,
+                        help="<Optional> The golden path.", required=False)
     parser.add_argument("-lm", "--layer_mapping", dest="layer_mapping", type=str, nargs='?', const=True,
                         help="<Optional> The layer mapping file path.", required=False)
     parser.add_argument("-oc", "--overflow_check", dest="overflow_check", action="store_true",
                         help="<Optional> whether open overflow_check for graph.", required=False)
-    parser.add_argument("-f", "--fuzzy_match", dest="fuzzy_match", action="store_true",
-                        help="<Optional> Whether to perform a fuzzy match on the api name.", required=False)
+    parser.add_argument("-fm", "--fuzzy_match", dest="fuzzy_match", action="store_true",
+                        help="<Optional> whether to perform a fuzzy match on the api name.", required=False)
+    parser.add_argument("-tensor_log", "--is_print_compare_log", dest="is_print_compare_log", action="store_true",
+                        help="<Optional> whether print tensor compare log for visualization task.", required=False)
+    parser.add_argument("-progress_log", "--is_print_progress_log", dest="is_print_progress_log", action="store_true",
+                        help="<Optional> whether print progress log for visualization task.", required=False)
+
+    # -------------------------- 不同并行切分策略合并可选参数 --------------------------
+    group_n = parser.add_argument_group("Parallel Parameters, "
+                                        "used for graph merging under different parallel partitioning strategies")
+
+    group_n.add_argument("--rank_size", type=int, nargs='+', help="<Optional> The rank size of dump path.",
+                         required=False)
+    group_n.add_argument("--tp", type=int, nargs='+',
+                         help="<Optional, but required if rank_size is not empty> The tp size of dump path.",
+                         required=False)
+    group_n.add_argument("--pp", type=int, nargs='+',
+                         help="<Optional, but required if rank_size is not empty> The pp size of dump path.",
+                         required=False)
+    group_n.add_argument("--vpp", type=int, nargs='+', default=[1], help="<Optional> The vpp size of dump path.",
+                         required=False)
+    group_n.add_argument("--order", type=str, nargs='+', default=['tp-cp-ep-dp-pp'],
+                         help="<Optional> The order of dump path.", required=False)
 
 
 def _graph_service_command(args):
-    input_param = load_json(args.input_path)
-    npu_path = input_param.get("npu_path")
-    bench_path = input_param.get("bench_path")
-    args.parallel_merge = check_whether_parallel_merge(input_param)
-    args.parallel_params = load_parallel_param(input_param) if args.parallel_merge else None
-    check_file_or_directory_path(npu_path, isdir=True)
-    if bench_path:
-        check_file_or_directory_path(bench_path, isdir=True)
-    if check_file_type(npu_path) == FileCheckConst.DIR and not bench_path:
-        content = check_directory_content(npu_path)
-        output_db_path = os.path.join(args.output_path, build_output_db_name)
-        if content == GraphConst.RANKS:
-            _build_graph_ranks(npu_path, args)
-        elif content == GraphConst.STEPS:
-            _build_graph_steps(npu_path, args)
+    try:
+        if args.is_print_progress_log:
+            # 往ProgressInfo中记录error日志，用于前端展示
+            BaseLogger.error = get_log_msg_wrapper(BaseLogger.error)
+        npu_path = args.target_path
+        bench_path = args.golden_path
+        ProgressInfo.print_progress_log = args.is_print_progress_log
+        args.parallel_merge = True if args.rank_size else False
+        args.parallel_params = load_parallel_param(args) if args.parallel_merge else None
+        check_file_or_directory_path(npu_path, isdir=True)
+        if bench_path:
+            check_file_or_directory_path(bench_path, isdir=True)
+        if check_file_type(npu_path) == FileCheckConst.DIR and not bench_path:
+            content = check_directory_content(npu_path)
+            if content == GraphConst.RANKS:
+                _build_graph_ranks_with_pbar(args)
+            elif content == GraphConst.STEPS:
+                _build_graph_steps_with_pbar(args)
+            else:
+                _build_graph_with_pbar(npu_path, args)
+        elif check_file_type(npu_path) == FileCheckConst.DIR and check_file_type(bench_path) == FileCheckConst.DIR:
+            content_n = check_directory_content(npu_path)
+            content_b = check_directory_content(bench_path)
+            if content_n != content_b:
+                raise ValueError('The directory structures of npu_path and bench_path are inconsistent.')
+            input_param = {
+                'npu_path': args.target_path,
+                'bench_path': args.golden_path,
+                'is_print_compare_log': args.is_print_compare_log
+            }
+            if content_n == GraphConst.RANKS:
+                _compare_graph_ranks_with_pbar(input_param, args)
+            elif content_n == GraphConst.STEPS:
+                _compare_graph_steps_with_pbar(input_param, args)
+            else:
+                _compare_graph_with_pbar(input_param, args)
         else:
-            result = _build_graph_result(npu_path, args)
-            create_directory(args.output_path)
-            file_name = _export_build_graph_result(args, result)
-            if file_name:
-                logger.error('Failed to export model build graph.')
-    elif check_file_type(npu_path) == FileCheckConst.DIR and check_file_type(bench_path) == FileCheckConst.DIR:
-        content_n = check_directory_content(npu_path)
-        content_b = check_directory_content(bench_path)
-        output_db_path = os.path.join(args.output_path, compare_output_db_name)
-        if content_n != content_b:
-            raise ValueError('The directory structures of npu_path and bench_path are inconsistent.')
-        if content_n == GraphConst.RANKS:
-            if args.parallel_merge:
-                _compare_graph_ranks_parallel(input_param, args)
+            logger.error("The npu_path or bench_path should be a folder.")
+            raise CompareException(CompareException.INVALID_COMPARE_MODE)
+    except KeyboardInterrupt:
+        logger.warning("Interrupted by user, terminating processes and cleaning up...")
+    except Exception as e:
+        logger.error(f"An unexpected error occurred: {e}")
+        raise e
+    finally:
+        ProgressInfo.update_process_running(False)
+
+
+@dataclass
+class ProgressConfig:
+    core_func: Callable
+    get_ranks: Callable
+    db_name: str
+    pbar_info_kwargs: dict = None
+    use_monitor_thread: bool = True
+    tqdm_total: Optional[int] = None
+
+
+def _run_with_progress(param, args, config: ProgressConfig):
+    """通用进度条处理"""
+
+    monitor_thread = None
+    pbar_info = None
+    ranks = None
+
+    try:
+        if config.use_monitor_thread:
+            manager = Manager()
+            progress_dict = manager.dict()
+            pbar_info = PbarInfo(progress_dict=progress_dict, **config.pbar_info_kwargs)
+            ranks = config.get_ranks(args)
+        else:
+            pbar_info = PbarInfo(**config.pbar_info_kwargs)
+
+        tqdm_args = {
+            "desc": GraphConst.PBAR_DESC_PREFIX,
+            "total": config.tqdm_total if config.tqdm_total is not None else pbar_info.total,
+            "bar_format": GraphConst.BAR_FORMAT
+        }
+
+        with tqdm(**tqdm_args) as pbar:
+            # 单进程场景直接更新pbar，多进程场景需要通过monitor thread从共享dict中获取进度更新pbar
+            if config.use_monitor_thread:
+                monitor_thread = threading.Thread(target=monitor_progress,
+                                                  args=(pbar_info, pbar, ranks, args.parallel_merge))
+                monitor_thread.start()
             else:
-                _compare_graph_ranks(input_param, args)
-        elif content_n == GraphConst.STEPS:
-            _compare_graph_steps(input_param, args)
+                pbar_info.pbar = pbar
+
+            if param:
+                config.core_func(param, args, pbar_info=pbar_info)
+            else:
+                config.core_func(args, pbar_info=pbar_info)
+
+            db_path = os.path.join(args.output_path, config.db_name)
+            post_process_db(db_path, pbar_info=pbar_info, is_parallel_merge=args.parallel_merge)
+
+            if not args.parallel_merge and config.use_monitor_thread:
+                distributed_analyse(db_path, args.overflow_check, pbar_info=pbar_info)
+
+            if config.use_monitor_thread and monitor_thread:
+                monitor_thread.join(timeout=5)
+
+    except KeyboardInterrupt:
+        logger.warning("Interrupted by user, terminating processes and cleaning up...")
+    except Exception as e:
+        logger.error(f"An unexpected error occurred: {e}")
+        raise e
+    finally:
+        ProgressInfo.update_process_running(False)
+        if config.use_monitor_thread and pbar_info:
+            pbar_info.stop_monitor = True
+
+
+def _build_graph_ranks_with_pbar(args):
+    def core_func(args, pbar_info):
+        if args.parallel_merge:
+            _build_graph_ranks_parallel(args, pbar_info=pbar_info)
         else:
-            result = _compare_graph_result(input_param, args)
-            create_directory(args.output_path)
-            file_name = _export_compare_graph_result(args, result)
-            if file_name:
-                logger.error('Failed to export model compare graph.')
-    else:
-        logger.error("The npu_path or bench_path should be a folder.")
-        raise CompareException(CompareException.INVALID_COMPARE_MODE)
-    # 所有数据输出db结束后，添加索引，修改权限
-    post_process_db(output_db_path)
+            _build_graph_ranks(args, pbar_info=pbar_info)
 
+    def get_ranks(args):
+        return check_and_return_dir_contents(args.target_path, Const.RANK)
 
-def _pt_graph_service_parser(parser):
-    _graph_service_parser(parser)
+    stage_total = _get_parallel_stage_total(args) if args.parallel_merge else GraphConst.BUILD_STAGES_TOTAL
 
+    _run_with_progress(
+        param=None,
+        args=args,
+        config=ProgressConfig(
+            core_func=core_func,
+            get_ranks=get_ranks,
+            pbar_info_kwargs={"stage_total": stage_total},
+            db_name=build_output_db_name,
+        )
+    )
 
-def _pt_graph_service_command(args):
-    _graph_service_command(args)
 
+def _build_graph_steps_with_pbar(args):
+    steps = check_and_return_dir_contents(args.target_path, Const.STEP)
 
-def _ms_graph_service_parser(parser):
-    _graph_service_parser(parser)
+    def get_ranks(args):
+        return check_and_return_dir_contents(os.path.join(args.target_path, steps[0]), Const.RANK)
 
+    stage_total = _get_parallel_stage_total(args, steps) if args.parallel_merge else GraphConst.BUILD_STAGES_TOTAL
 
-def _ms_graph_service_command(args):
-    _graph_service_command(args)
+    _run_with_progress(
+        param=None,
+        args=args,
+        config=ProgressConfig(
+            core_func=_build_graph_steps,
+            get_ranks=get_ranks,
+            pbar_info_kwargs={"step_total": len(steps), "stage_total": stage_total},
+            db_name=build_output_db_name,
+        )
+    )
+
+
+def _build_graph_with_pbar(npu_path, args):
+    def core_func(param, args, pbar_info):
+        result = _build_graph_result(param, args, pbar_info)
+        create_directory(args.output_path)
+        file_name = _export_build_graph_result(args, result, pbar_info)
+        if file_name:
+            logger.error('Failed to export model build graph.')
+
+    _run_with_progress(
+        param=npu_path,
+        args=args,
+        config=ProgressConfig(
+            core_func=core_func,
+            get_ranks=lambda x: None,
+            pbar_info_kwargs={},
+            db_name=build_output_db_name,
+            use_monitor_thread=False,
+            tqdm_total=GraphConst.PBAR_TOTAL
+        )
+    )
+
+
+def _compare_graph_ranks_with_pbar(input_param, args):
+    def core_func(param, args, pbar_info):
+        if args.parallel_merge:
+            _compare_graph_ranks_parallel(param, args, pbar_info=pbar_info)
+        else:
+            _compare_graph_ranks(param, args, pbar_info=pbar_info)
+
+    def get_ranks(args):
+        if args.parallel_merge:
+            return calculate_list(args.target_path, args.golden_path, mode=GraphConst.UNION)
+        return calculate_list(args.target_path, args.golden_path)
+
+    stage_total = _get_parallel_stage_total(args, is_compare=True) if args.parallel_merge \
+        else GraphConst.COMPARE_STAGES_TOTAL
+
+    _run_with_progress(
+        param=input_param,
+        args=args,
+        config=ProgressConfig(
+            core_func=core_func,
+            get_ranks=get_ranks,
+            pbar_info_kwargs={"stage_total": stage_total},
+            db_name=compare_output_db_name
+        )
+    )
+
+
+def _compare_graph_steps_with_pbar(input_param, args):
+    steps = calculate_list(args.target_path, args.golden_path, Const.STEP)
+
+    def get_ranks(args):
+        rank_path_t = os.path.join(args.target_path, steps[0])
+        rank_path_g = os.path.join(args.golden_path, steps[0])
+        if args.parallel_merge:
+            return calculate_list(rank_path_t, rank_path_g, mode=GraphConst.UNION)
+        return calculate_list(rank_path_t, rank_path_g)
+
+    stage_total = _get_parallel_stage_total(args, steps, is_compare=True) if args.parallel_merge \
+        else GraphConst.COMPARE_STAGES_TOTAL
+
+    _run_with_progress(
+        param=input_param,
+        args=args,
+        config=ProgressConfig(
+            core_func=_compare_graph_steps,
+            get_ranks=get_ranks,
+            pbar_info_kwargs={"stage_total": stage_total, "step_total": len(steps)},
+            db_name=compare_output_db_name
+        )
+    )
+
+
+def _compare_graph_with_pbar(input_param, args):
+    def core_func(param, args, pbar_info):
+        result = _compare_graph_result(param, args, pbar_info=pbar_info)
+        create_directory(args.output_path)
+        file_name = _export_compare_graph_result(args, result, pbar_info=pbar_info)
+        if file_name:
+            logger.error('Failed to export model compare graph.')
+
+    _run_with_progress(
+        param=input_param,
+        args=args,
+        config=ProgressConfig(
+            core_func=core_func,
+            get_ranks=lambda x: None,
+            pbar_info_kwargs={"pbar": None, "stage_total": GraphConst.COMPARE_STAGES_TOTAL},
+            db_name=compare_output_db_name,
+            use_monitor_thread=False,
+            tqdm_total=GraphConst.PBAR_TOTAL
+        )
+    )
+
+
+def _get_parallel_stage_total(args, steps=None, is_compare=False):
+    """
+    获取不同并行切分策略的任务阶段数
+    """
+    parallel_params = args.parallel_params
+    if not is_compare and (not parallel_params or len(parallel_params) != 1):
+        raise RuntimeError('Parallel params error in build graph!')
+    if is_compare and (not parallel_params or len(parallel_params) != 2):
+        raise RuntimeError('Parallel params error in compare graph!')
+    target_path = os.path.join(args.target_path, steps[0]) if steps else args.target_path
+    validate_parallel_param(parallel_params[0], target_path)
+
+    if is_compare:
+        golden_path = os.path.join(args.golden_path, steps[0]) if steps else args.golden_path
+        validate_parallel_param(parallel_params[1], golden_path, '[Bench]')
+
+    stage_count_map = {
+        "TPMerger": lambda param: param.rank_size // param.tp,
+        "PPMerger": lambda param: param.rank_size // param.pp,
+        "VPPMerger": lambda param: param.rank_size // param.pp,
+        "TPPPMerger": lambda param: param.rank_size // param.pp + param.rank_size // param.pp // param.tp,
+        "FullMerger": lambda param: param.rank_size // param.pp + param.rank_size // param.pp // param.tp,
+        "NoParallelMerger": 0
+    }
+
+    def _get_stage_count(parallel_param, merger_name: str) -> int:
+        rule = stage_count_map.get(merger_name, 0)
+        return rule(parallel_param) if callable(rule) else rule
+
+    merger_name_t = GraphMerger([], parallel_params[0]).strategy.__class__.__name__
+    stage_count_target = _get_stage_count(parallel_params[0], merger_name_t)
+
+    if is_compare:
+        merger_name_g = GraphMerger([], parallel_params[1]).strategy.__class__.__name__
+        stage_count_golden = _get_stage_count(parallel_params[1], merger_name_g)
+        return GraphConst.COMPARE_STAGES_TOTAL + stage_count_target + stage_count_golden
+
+    return GraphConst.BUILD_STAGES_TOTAL + stage_count_target
 
 
 class CompareGraphResult:
@@ -526,3 +832,72 @@ class BuildGraphResult:
         self.micro_steps = micro_steps
         self.rank = rank
         self.step = step
+
+
+class PbarInfo:
+    def __init__(self, pbar=None, progress_dict=None, task_id=None, step=0, step_total=1,
+                 stage_total=GraphConst.BUILD_STAGES_TOTAL):
+        self.pbar = pbar
+        self.progress_dict = progress_dict
+        self.task_id = task_id
+        self.step = step
+        self.step_total = step_total
+        self.total = GraphConst.PBAR_TOTAL * step_total
+        self.stage_total = stage_total * step_total  # 有几个阶段
+        self.current_stage_dict = Manager().dict()  # 当前阶段，进程共享
+        self.stage_progress = round(self.total / self.stage_total, 2)  # 每个阶段的最大进度
+        self.stop_monitor = False
+        self.wait_monitor = False
+        self.continue_monitor = True
+
+    def __deepcopy__(self, memo):
+        new_obj = PbarInfo()
+        new_obj.progress_dict = self.progress_dict
+        new_obj.task_id = self.task_id
+        new_obj.step = self.step
+        new_obj.step_total = self.step_total
+        new_obj.stage_total = self.stage_total
+        new_obj.current_stage_dict = self.current_stage_dict
+        new_obj.stage_progress = self.stage_progress
+        new_obj.total = self.total
+        new_obj.stop_monitor = self.stop_monitor
+        new_obj.wait_monitor = self.wait_monitor
+        new_obj.continue_monitor = self.continue_monitor
+        return new_obj
+
+    @staticmethod
+    def update_task_id(pbar_info, task_id):
+        """
+        在进程池中，实例作为入参，修改实例属性，需要深拷贝实例使修改生效
+        """
+        if pbar_info:
+            pbar_info.task_id = task_id
+            return deepcopy(pbar_info)
+        return pbar_info
+
+    @staticmethod
+    def del_progress_dict_item(pbar_info, origin_ranks, merged_ranks):
+        """
+        不同并行切分策略的图合并场景下，graph合并到一些rank中，剩余的rank作为task_id不再需要
+        """
+        diff_ranks = list(set(origin_ranks) - set(merged_ranks))
+        for rank in diff_ranks:
+            if rank in pbar_info.progress_dict:
+                del pbar_info.progress_dict[rank]
+
+    @staticmethod
+    def reset_progress_and_current_stage(pbar_info, task_ids):
+        """
+        不同并行切分策略的图合并场景下，每个step需要重置进度信息
+        """
+        for task_id in task_ids:
+            pbar_info.progress_dict[task_id] = GraphConst.PBAR_TOTAL * pbar_info.step
+            pbar_info.current_stage_dict[task_id] = pbar_info.stage_total // pbar_info.step_total * pbar_info.step
+
+    def set_continue_monitor(self, value: bool):
+        self.continue_monitor = value
+        self.wait_monitor = not value
+
+    def set_wait_monitor(self, value: bool):
+        self.wait_monitor = value
+        self.continue_monitor = not value