mirror of
https://github.com/alibaba/EasyCV.git
synced 2025-06-03 14:49:00 +08:00
9aaa600f79
* add attention layer and more loss function * add attention layer and various loss functions * add siou loss * add tah,various attention layers, and different loss functions * add asff sim, gsconv * blade utils fit faster * blade optimize for yolox static & fp16 * decode output for yolox control by cfg * add reparameterize_models for export * e2e trt_nms plugin export support and numeric test * split preprocess from end2end+blade, speedup from 17ms->7.2ms Co-authored-by: zouxinyi0625 <zouxinyi.zxy@alibaba-inc.com>
133 lines
4.2 KiB
Python
133 lines
4.2 KiB
Python
# Copyright (c) Alibaba, Inc. and its affiliates.
|
|
import copy
|
|
import glob
|
|
import json
|
|
import logging
|
|
import os
|
|
import sys
|
|
import tempfile
|
|
import unittest
|
|
|
|
import torch
|
|
from mmcv import Config
|
|
from tests.ut_config import (DET_DATA_MANIFEST_OSS, DET_DATA_SMALL_COCO_LOCAL,
|
|
PRETRAINED_MODEL_YOLOXS)
|
|
|
|
from easycv.file import io
|
|
from easycv.file.utils import get_oss_config
|
|
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_COCO_DATA_ROOT = DET_DATA_SMALL_COCO_LOCAL.rstrip('/') + '/'
|
|
SMALL_COCO_ITAG_DATA_ROOT = DET_DATA_MANIFEST_OSS.rstrip('/') + '/'
|
|
_COMMON_OPTIONS = {
|
|
'data.imgs_per_gpu': 1,
|
|
}
|
|
|
|
TRAIN_CONFIGS = [
|
|
# itag test
|
|
{
|
|
'config_file':
|
|
'configs/detection/yolox/yolox_s_8xb16_300e_coco_pai.py',
|
|
'cfg_options': {
|
|
**_COMMON_OPTIONS,
|
|
'data.train.data_source.path':
|
|
SMALL_COCO_ITAG_DATA_ROOT + 'train2017_20.manifest',
|
|
'data.val.data_source.path':
|
|
SMALL_COCO_ITAG_DATA_ROOT + 'val2017_20.manifest',
|
|
}
|
|
},
|
|
{
|
|
'config_file': 'configs/detection/yolox/yolox_s_8xb16_300e_coco.py',
|
|
'cfg_options': {
|
|
**_COMMON_OPTIONS, 'data.train.data_source.img_prefix':
|
|
SMALL_COCO_DATA_ROOT + 'train2017',
|
|
'data.val.data_source.img_prefix':
|
|
SMALL_COCO_DATA_ROOT + 'val2017',
|
|
'data.train.data_source.ann_file':
|
|
SMALL_COCO_DATA_ROOT + 'instances_train2017_20.json',
|
|
'data.val.data_source.ann_file':
|
|
SMALL_COCO_DATA_ROOT + 'instances_val2017_20.json'
|
|
}
|
|
},
|
|
]
|
|
|
|
|
|
class EvalTest(unittest.TestCase):
|
|
|
|
def setUp(self):
|
|
print(('Testing %s.%s' % (type(self).__name__, self._testMethodName)))
|
|
|
|
def tearDown(self):
|
|
super().tearDown()
|
|
|
|
def check_metric(self, work_dir):
|
|
json_file = glob.glob(os.path.join(work_dir, '*.json'))
|
|
with io.open(json_file[0], 'r') as f:
|
|
content = f.readlines()
|
|
res = json.loads(content[0])
|
|
self.assertAlmostEqual(
|
|
res['DetectionBoxes_Precision/mAP'], 0.450, delta=0.001)
|
|
self.assertAlmostEqual(
|
|
res['DetectionBoxes_Precision/mAP@.50IOU'],
|
|
0.6132,
|
|
delta=0.001)
|
|
self.assertAlmostEqual(
|
|
res['DetectionBoxes_Precision/mAP@.75IOU'], 0.490, delta=0.001)
|
|
|
|
def _base_eval(self, eval_cfgs, dist=False, dist_eval=False):
|
|
cfg_file = eval_cfgs.pop('config_file')
|
|
cfg_options = eval_cfgs.pop('cfg_options', None)
|
|
work_dir = eval_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) # imgs_per_gpu=1
|
|
cfg.eval_pipelines[0].dist_eval = dist_eval
|
|
|
|
tmp_cfg_file = tempfile.NamedTemporaryFile(suffix='.py').name
|
|
cfg.dump(tmp_cfg_file)
|
|
|
|
args_str = ' '.join(
|
|
['='.join((str(k), str(v))) for k, v in eval_cfgs.items()])
|
|
|
|
if dist:
|
|
nproc_per_node = 2
|
|
cmd = 'bash tools/dist_test.sh %s %s %s --eval --work_dir=%s %s ' % (
|
|
tmp_cfg_file, nproc_per_node, PRETRAINED_MODEL_YOLOXS,
|
|
work_dir, args_str)
|
|
else:
|
|
cmd = 'python tools/eval.py %s %s --eval --work_dir=%s %s' % (
|
|
tmp_cfg_file, PRETRAINED_MODEL_YOLOXS, work_dir, args_str)
|
|
|
|
logging.info('run command: %s' % cmd)
|
|
run_in_subprocess(cmd)
|
|
|
|
self.check_metric(work_dir)
|
|
|
|
io.remove(work_dir)
|
|
io.remove(tmp_cfg_file)
|
|
|
|
def test_eval(self):
|
|
eval_cfgs = copy.deepcopy(TRAIN_CONFIGS[1])
|
|
|
|
self._base_eval(eval_cfgs)
|
|
|
|
@unittest.skipIf(torch.cuda.device_count() <= 1, 'distributed unittest')
|
|
def test_eval_dist(self):
|
|
eval_cfgs = copy.deepcopy(TRAIN_CONFIGS[0])
|
|
eval_cfgs['cfg_options'].update(dict(oss_io_config=get_oss_config()))
|
|
|
|
self._base_eval(eval_cfgs, dist=True, dist_eval=True)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
unittest.main()
|