mirror of https://github.com/alibaba/EasyCV.git
117 lines
3.6 KiB
Python
117 lines
3.6 KiB
Python
# Copyright (c) Alibaba, Inc. and its affiliates.
|
|
import copy
|
|
import logging
|
|
import os
|
|
import sys
|
|
import tempfile
|
|
import unittest
|
|
|
|
import torch
|
|
from mmcv import Config
|
|
from tests.ut_config import (COMPRESSION_TEST_DATA,
|
|
PRETRAINED_MODEL_YOLOX_COMPRESSION)
|
|
|
|
from easycv.file import io
|
|
from easycv.utils.test_util import run_in_subprocess
|
|
|
|
sys.path.append(os.path.dirname(os.path.realpath(__file__)))
|
|
|
|
logging.basicConfig(level=logging.INFO)
|
|
|
|
SMALL_IMAGENET_DATA_ROOT = COMPRESSION_TEST_DATA.rstrip('/') + '/'
|
|
_QUANTIZE_OPTIONS = {
|
|
'total_epochs': 1,
|
|
'data.imgs_per_gpu': 16,
|
|
}
|
|
|
|
_PRUNE_OPTIONS = {
|
|
'total_epochs': 11,
|
|
'data.imgs_per_gpu': 1,
|
|
}
|
|
|
|
TRAIN_CONFIGS = [
|
|
{
|
|
'config_file': 'configs/edge_models/yolox_edge.py',
|
|
'cfg_options': {
|
|
**_QUANTIZE_OPTIONS, 'data.train.data_source.ann_file':
|
|
SMALL_IMAGENET_DATA_ROOT + 'annotations/instances_train2017.json',
|
|
'data.train.data_source.img_prefix':
|
|
SMALL_IMAGENET_DATA_ROOT + 'images',
|
|
'data.val.data_source.ann_file':
|
|
SMALL_IMAGENET_DATA_ROOT + 'annotations/instances_train2017.json',
|
|
'data.val.data_source.img_prefix':
|
|
SMALL_IMAGENET_DATA_ROOT + 'images'
|
|
}
|
|
},
|
|
{
|
|
'config_file': 'configs/edge_models/yolox_edge.py',
|
|
'cfg_options': {
|
|
**_PRUNE_OPTIONS, 'img_scale': (128, 128),
|
|
'data.train.data_source.ann_file':
|
|
SMALL_IMAGENET_DATA_ROOT + 'annotations/instances_train2017.json',
|
|
'data.train.data_source.img_prefix':
|
|
SMALL_IMAGENET_DATA_ROOT + 'images',
|
|
'data.val.data_source.ann_file':
|
|
SMALL_IMAGENET_DATA_ROOT + 'annotations/instances_train2017.json',
|
|
'data.val.data_source.img_prefix':
|
|
SMALL_IMAGENET_DATA_ROOT + 'images'
|
|
}
|
|
},
|
|
]
|
|
|
|
|
|
class ModelPruneTest(unittest.TestCase):
|
|
|
|
def setUp(self):
|
|
print(('Testing %s.%s' % (type(self).__name__, self._testMethodName)))
|
|
|
|
def tearDown(self):
|
|
super().tearDown()
|
|
|
|
def _base_prune(self, train_cfgs):
|
|
print(train_cfgs)
|
|
cfg_file = train_cfgs.pop('config_file')
|
|
cfg_options = train_cfgs.pop('cfg_options', None)
|
|
work_dir = train_cfgs.pop('work_dir', None)
|
|
if not work_dir:
|
|
work_dir = tempfile.TemporaryDirectory().name
|
|
|
|
cfg = Config.fromfile(cfg_file)
|
|
if cfg_options is not None:
|
|
cfg.merge_from_dict(cfg_options)
|
|
cfg.eval_pipelines[0].data = dict(**cfg.data.val)
|
|
|
|
tmp_cfg_file = tempfile.NamedTemporaryFile(suffix='.py').name
|
|
cfg.dump(tmp_cfg_file)
|
|
|
|
ckpt_path = PRETRAINED_MODEL_YOLOX_COMPRESSION
|
|
pruning_class = 'AGP'
|
|
pruning_algorithm = 'taylorfo'
|
|
|
|
args_str = ' '.join(
|
|
['='.join((str(k), str(v))) for k, v in train_cfgs.items()])
|
|
|
|
cmd = 'python tools/prune.py %s %s --work_dir=%s --pruning_class=%s --pruning_algorithm=%s %s' % \
|
|
(tmp_cfg_file, ckpt_path, work_dir, pruning_class, pruning_algorithm, args_str)
|
|
|
|
logging.info('run command: %s' % cmd)
|
|
run_in_subprocess(cmd)
|
|
|
|
output_files = io.listdir(work_dir)
|
|
self.assertIn('{}_traced_model.pt'.format(cfg.model.type),
|
|
output_files)
|
|
|
|
io.remove(work_dir)
|
|
io.remove(tmp_cfg_file)
|
|
|
|
@unittest.skipIf(torch.__version__ < '1.8.0',
|
|
'model compression need pytorch version >= 1.8.0')
|
|
def test_model_prune(self):
|
|
train_cfgs = copy.deepcopy(TRAIN_CONFIGS[1])
|
|
|
|
self._base_prune(train_cfgs)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|