Merge pull request #719 from yukavio/develop

add slim/prune

deploy/slim/prune/README.md

+> 运行示例前请先安装develop版本PaddleSlim
+
+# 模型裁剪压缩教程
+
+## 概述
+
+该示例使用PaddleSlim提供的[裁剪压缩API](https://paddlepaddle.github.io/PaddleSlim/api/prune_api/)对OCR模型进行压缩。
+在阅读该示例前，建议您先了解以下内容：
+
+- [OCR模型的常规训练方法](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/doc/doc_ch/detection.md)
+- [PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)
+
+## 安装PaddleSlim
+可按照[PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)中的步骤安装PaddleSlim。
+
+
+
+## 敏感度分析训练
+
+进入PaddleOCR根目录，通过以下命令对模型进行敏感度分析：
+
+```bash
+python deploy/slim/prune/sensitivity_anal.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./deploy/slim/prune/pretrain_models/det_mv3_db/best_accuracy Global.test_batch_size_per_card=1
+```
+
+## 裁剪模型与fine-tune
+
+```bash
+python deploy/slim/prune/pruning_and_finetune.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./deploy/slim/prune/pretrain_models/det_mv3_db/best_accuracy Global.test_batch_size_per_card=1
+```
+
+
+
+## 评估并导出
+
+在得到裁剪训练保存的模型后，我们可以将其导出为inference_model，用于预测部署：
+
+```bash
+python deploy/slim/prune/export_prune_model.py -c configs/det/det_mv3_db.yml -o Global.pretrain_weights=./output/det_db/best_accuracy Global.test_batch_size_per_card=1 Global.save_inference_dir=inference_model
+```

deploy/slim/prune/export_prune_model.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# 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.
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import os
+import sys
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..', '..', '..'))
+sys.path.append(os.path.join(__dir__, '..', '..', '..', 'tools'))
+
+import program
+from paddle import fluid
+from ppocr.utils.utility import initial_logger
+logger = initial_logger()
+from ppocr.utils.save_load import init_model
+from paddleslim.prune import load_model
+
+
+def main():
+    startup_prog, eval_program, place, config, _ = program.preprocess()
+
+    feeded_var_names, target_vars, fetches_var_name = program.build_export(
+        config, eval_program, startup_prog)
+    eval_program = eval_program.clone(for_test=True)
+    exe = fluid.Executor(place)
+    exe.run(startup_prog)
+
+    if config['Global']['checkpoints'] is not None:
+        path = config['Global']['checkpoints']
+    else:
+        path = config['Global']['pretrain_weights']
+
+    load_model(exe, eval_program, path)
+
+    save_inference_dir = config['Global']['save_inference_dir']
+    if not os.path.exists(save_inference_dir):
+        os.makedirs(save_inference_dir)
+    fluid.io.save_inference_model(
+        dirname=save_inference_dir,
+        feeded_var_names=feeded_var_names,
+        main_program=eval_program,
+        target_vars=target_vars,
+        executor=exe,
+        model_filename='model',
+        params_filename='params')
+    print("inference model saved in {}/model and {}/params".format(
+        save_inference_dir, save_inference_dir))
+    print("save success, output_name_list:", fetches_var_name)
+
+
+if __name__ == '__main__':
+    main()

deploy/slim/prune/pruning_and_finetune.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# 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.
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import os
+import sys
+import numpy as np
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..', '..', '..'))
+sys.path.append(os.path.join(__dir__, '..', '..', '..', 'tools'))
+
+import tools.program as program
+from paddle import fluid
+from ppocr.utils.utility import initial_logger
+logger = initial_logger()
+from ppocr.data.reader_main import reader_main
+from ppocr.utils.save_load import init_model
+from ppocr.utils.character import CharacterOps
+from ppocr.utils.utility import initial_logger
+from paddleslim.prune import Pruner, save_model
+from paddleslim.analysis import flops
+from paddleslim.core.graph_wrapper import *
+from paddleslim.prune import load_sensitivities, get_ratios_by_loss, merge_sensitive
+logger = initial_logger()
+
+skip_list = [
+    'conv10_linear_weights', 'conv11_linear_weights', 'conv12_expand_weights',
+    'conv12_linear_weights', 'conv12_se_2_weights', 'conv13_linear_weights',
+    'conv2_linear_weights', 'conv4_linear_weights', 'conv5_expand_weights',
+    'conv5_linear_weights', 'conv5_se_2_weights', 'conv6_linear_weights',
+    'conv7_linear_weights', 'conv8_expand_weights', 'conv8_linear_weights',
+    'conv9_expand_weights', 'conv9_linear_weights'
+]
+
+
+def main():
+    config = program.load_config(FLAGS.config)
+    program.merge_config(FLAGS.opt)
+    logger.info(config)
+
+    # check if set use_gpu=True in paddlepaddle cpu version
+    use_gpu = config['Global']['use_gpu']
+    program.check_gpu(use_gpu)
+
+    alg = config['Global']['algorithm']
+    assert alg in ['EAST', 'DB', 'Rosetta', 'CRNN', 'STARNet', 'RARE']
+    if alg in ['Rosetta', 'CRNN', 'STARNet', 'RARE']:
+        config['Global']['char_ops'] = CharacterOps(config['Global'])
+
+    place = fluid.CUDAPlace(0) if use_gpu else fluid.CPUPlace()
+    startup_program = fluid.Program()
+    train_program = fluid.Program()
+    train_build_outputs = program.build(
+        config, train_program, startup_program, mode='train')
+    train_loader = train_build_outputs[0]
+    train_fetch_name_list = train_build_outputs[1]
+    train_fetch_varname_list = train_build_outputs[2]
+    train_opt_loss_name = train_build_outputs[3]
+
+    eval_program = fluid.Program()
+    eval_build_outputs = program.build(
+        config, eval_program, startup_program, mode='eval')
+    eval_fetch_name_list = eval_build_outputs[1]
+    eval_fetch_varname_list = eval_build_outputs[2]
+    eval_program = eval_program.clone(for_test=True)
+
+    train_reader = reader_main(config=config, mode="train")
+    train_loader.set_sample_list_generator(train_reader, places=place)
+
+    eval_reader = reader_main(config=config, mode="eval")
+
+    exe = fluid.Executor(place)
+    exe.run(startup_program)
+
+    # compile program for multi-devices
+    init_model(config, train_program, exe)
+
+    sen = load_sensitivities("sensitivities_0.data")
+    for i in skip_list:
+        sen.pop(i)
+    back_bone_list = ['conv' + str(x) for x in range(1, 5)]
+    for i in back_bone_list:
+        for key in list(sen.keys()):
+            if i + '_' in key:
+                sen.pop(key)
+    ratios = get_ratios_by_loss(sen, 0.03)
+    logger.info("FLOPs before pruning: {}".format(flops(eval_program)))
+    pruner = Pruner(criterion='geometry_median')
+    print("ratios: {}".format(ratios))
+    pruned_val_program, _, _ = pruner.prune(
+        eval_program,
+        fluid.global_scope(),
+        params=ratios.keys(),
+        ratios=ratios.values(),
+        place=place,
+        only_graph=True)
+
+    pruned_program, _, _ = pruner.prune(
+        train_program,
+        fluid.global_scope(),
+        params=ratios.keys(),
+        ratios=ratios.values(),
+        place=place)
+    logger.info("FLOPs after pruning: {}".format(flops(pruned_val_program)))
+    train_compile_program = program.create_multi_devices_program(
+        pruned_program, train_opt_loss_name)
+
+
+    train_info_dict = {'compile_program':train_compile_program,\
+        'train_program':pruned_program,\
+        'reader':train_loader,\
+        'fetch_name_list':train_fetch_name_list,\
+        'fetch_varname_list':train_fetch_varname_list}
+
+    eval_info_dict = {'program':pruned_val_program,\
+        'reader':eval_reader,\
+        'fetch_name_list':eval_fetch_name_list,\
+        'fetch_varname_list':eval_fetch_varname_list}
+
+    if alg in ['EAST', 'DB']:
+        program.train_eval_det_run(
+            config, exe, train_info_dict, eval_info_dict, is_pruning=True)
+    else:
+        program.train_eval_rec_run(config, exe, train_info_dict, eval_info_dict)
+
+
+if __name__ == '__main__':
+    parser = program.ArgsParser()
+    FLAGS = parser.parse_args()
+    main()

deploy/slim/prune/sensitivity_anal.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# 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.
+
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+
+import os
+import sys
+__dir__ = os.path.dirname(__file__)
+sys.path.append(__dir__)
+sys.path.append(os.path.join(__dir__, '..', '..', '..'))
+sys.path.append(os.path.join(__dir__, '..', '..', '..', 'tools'))
+
+import json
+import cv2
+from paddle import fluid
+import paddleslim as slim
+from copy import deepcopy
+from tools.eval_utils.eval_det_utils import eval_det_run
+
+from tools import program
+from ppocr.utils.utility import initial_logger
+from ppocr.data.reader_main import reader_main
+from ppocr.utils.save_load import init_model
+from ppocr.utils.character import CharacterOps
+from ppocr.utils.utility import create_module
+from ppocr.data.reader_main import reader_main
+
+logger = initial_logger()
+
+
+def get_pruned_params(program):
+    params = []
+    for param in program.global_block().all_parameters():
+        if len(
+                param.shape
+        ) == 4 and 'depthwise' not in param.name and 'transpose' not in param.name:
+            params.append(param.name)
+    return params
+
+
+def eval_function(eval_args, mode='eval'):
+    exe = eval_args['exe']
+    config = eval_args['config']
+    eval_info_dict = eval_args['eval_info_dict']
+    metrics = eval_det_run(exe, config, eval_info_dict, mode=mode)
+    return metrics['hmean']
+
+
+def main():
+    config = program.load_config(FLAGS.config)
+    program.merge_config(FLAGS.opt)
+    logger.info(config)
+
+    # check if set use_gpu=True in paddlepaddle cpu version
+    use_gpu = config['Global']['use_gpu']
+    program.check_gpu(use_gpu)
+
+    alg = config['Global']['algorithm']
+    assert alg in ['EAST', 'DB', 'Rosetta', 'CRNN', 'STARNet', 'RARE']
+    if alg in ['Rosetta', 'CRNN', 'STARNet', 'RARE']:
+        config['Global']['char_ops'] = CharacterOps(config['Global'])
+
+    place = fluid.CUDAPlace(0) if use_gpu else fluid.CPUPlace()
+    startup_prog = fluid.Program()
+    eval_program = fluid.Program()
+    eval_build_outputs = program.build(
+        config, eval_program, startup_prog, mode='test')
+    eval_fetch_name_list = eval_build_outputs[1]
+    eval_fetch_varname_list = eval_build_outputs[2]
+    eval_program = eval_program.clone(for_test=True)
+    exe = fluid.Executor(place)
+    exe.run(startup_prog)
+
+    init_model(config, eval_program, exe)
+
+    eval_reader = reader_main(config=config, mode="eval")
+    eval_info_dict = {'program':eval_program,\
+        'reader':eval_reader,\
+        'fetch_name_list':eval_fetch_name_list,\
+        'fetch_varname_list':eval_fetch_varname_list}
+    eval_args = dict()
+    eval_args = {'exe': exe, 'config': config, 'eval_info_dict': eval_info_dict}
+    metrics = eval_function(eval_args)
+    print("Baseline: {}".format(metrics))
+
+    params = get_pruned_params(eval_program)
+    print('Start to analyze')
+    sens_0 = slim.prune.sensitivity(
+        eval_program,
+        place,
+        params,
+        eval_function,
+        sensitivities_file="sensitivities_0.data",
+        pruned_ratios=[0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8],
+        eval_args=eval_args,
+        criterion='geometry_median')
+
+
+if __name__ == '__main__':
+    parser = program.ArgsParser()
+    FLAGS = parser.parse_args()
+    main()

tools/program.py

@@ -241,7 +241,11 @@ def create_multi_devices_program(program, loss_var_name, for_quant=False):
     return compile_program
 
 
-def train_eval_det_run(config, exe, train_info_dict, eval_info_dict):
+def train_eval_det_run(config,
+                       exe,
+                       train_info_dict,
+                       eval_info_dict,
+                       is_pruning=False):
     train_batch_id = 0
     log_smooth_window = config['Global']['log_smooth_window']
     epoch_num = config['Global']['epoch_num']
@@ -297,7 +301,14 @@ def train_eval_det_run(config, exe, train_info_dict, eval_info_dict):
                         best_batch_id = train_batch_id
                         best_epoch = epoch
                         save_path = save_model_dir + "/best_accuracy"
-                        save_model(train_info_dict['train_program'], save_path)
+                        if is_pruning:
+                            import paddleslim as slim
+                            slim.prune.save_model(
+                                exe, train_info_dict['train_program'],
+                                save_path)
+                        else:
+                            save_model(train_info_dict['train_program'],
+                                       save_path)
                     strs = 'Test iter: {}, metrics:{}, best_hmean:{:.6f}, best_epoch:{}, best_batch_id:{}'.format(
                         train_batch_id, metrics, best_eval_hmean, best_epoch,
                         best_batch_id)
@@ -308,10 +319,20 @@ def train_eval_det_run(config, exe, train_info_dict, eval_info_dict):
             train_loader.reset()
         if epoch == 0 and save_epoch_step == 1:
             save_path = save_model_dir + "/iter_epoch_0"
-            save_model(train_info_dict['train_program'], save_path)
+            if is_pruning:
+                import paddleslim as slim
+                slim.prune.save_model(exe, train_info_dict['train_program'],
+                                      save_path)
+            else:
+                save_model(train_info_dict['train_program'], save_path)
         if epoch > 0 and epoch % save_epoch_step == 0:
             save_path = save_model_dir + "/iter_epoch_%d" % (epoch)
-            save_model(train_info_dict['train_program'], save_path)
+            if is_pruning:
+                import paddleslim as slim
+                slim.prune.save_model(exe, train_info_dict['train_program'],
+                                      save_path)
+            else:
+                save_model(train_info_dict['train_program'], save_path)
     return