TensorFlowModels/train_vm_.py at main · PurdueDualityLab/TensorFlowModels · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
# Lint as: python3
# pylint: skip-file
# Copyright 2020 The TensorFlow 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.
# ==============================================================================
"""TensorFlow Model Garden Vision training driver."""
from yolo.utils.run_utils import prep_gpu

from absl import app
from absl import flags
import gin
import sys

from official.core import train_utils
# pylint: disable=unused-import
from yolo.common import registry_imports
# pylint: enable=unused-import
from official.common import distribute_utils
from official.common import flags as tfm_flags
from official.core import task_factory
from official.core import train_lib
from official.modeling import performance
import tensorflow as tf

FLAGS = flags.FLAGS
"""
get the cache file:
scp -i <keyfile> cache.zip  purdue@<ip>:~/

tensorboard:
on the vm:
nohup tensorboard --logdir ../checkpoints/yolov4-model --port 6006  >> temp.log

on your device:
ssh -i <keyfile> -N -f -L localhost:16006:localhost:6006 purdue@<ip>

get the checkpoint from device:
scp -i <keyfile> purdue@<ip>:<path to>/checkpoints/<checkpoint>.zip .


train darknet:
python3.8 -m yolo.train_vm --mode=train_and_eval --experiment=darknet_classification --model_dir=../checkpoints/darknet53 --config_file=yolo/configs/experiments/darknet/darknet53.yaml
python3 -m yolo.train_vm --mode=train_and_eval --experiment=darknet_classification --model_dir=../checkpoints/dilated_darknet53 --config_file=yolo/configs/experiments/dilated_darknet53.yaml

finetune darknet:
nohup python3 -m yolo.train_vm --mode=train_and_eval --experiment=darknet_classification --model_dir=../checkpoints/darknet53_remap_fn --config_file=yolo/configs/experiments/darknet53_leaky_fn_tune.yaml >> darknet53.log & tail -f darknet53.log

train yolo-v4:
nohup python3 -m yolo.train_vm --mode=train_and_eval --experiment=yolo_custom --model_dir=../checkpoints/yolov4-model --config_file=yolo/configs/experiments/yolov4.yaml  >> yolov4.log & tail -f yolov4.log
nohup python3 -m yolo.train_vm --mode=train_and_eval --experiment=yolo_custom --model_dir=../checkpoints/yolov4- --config_file=yolo/configs/experiments/yolov4-1gpu.yaml  >> yolov4-1gpu.log & tail -f yolov4-1gpu.log


evalaute Yolo:
nohup python3 -m yolo.train_vm --mode=train_and_eval --experiment=yolo_custom --model_dir=../checkpoints/yolov4- --config_file=yolo/configs/experiments/yolov4-eval.yaml  >> yolov4-eval.log & tail -f yolov4-eval.log
"""


def subdivison_adjustment(params):
  if params.task.model.filter.nms_type == "greedy":
    tf.config.set_soft_device_placement(True)
  return params


def main(_):
  gin.parse_config_files_and_bindings(FLAGS.gin_file, FLAGS.gin_params)
  params = train_utils.parse_configuration(FLAGS)

  subdivison_adjustment(params)

  model_dir = FLAGS.model_dir
  if 'train' in FLAGS.mode and model_dir != None:
    # Pure eval modes do not output yaml files. Otherwise continuous eval job
    # may race against the train job for writing the same file.
    train_utils.serialize_config(params, model_dir)

  # Sets mixed_precision policy. Using 'mixed_float16' or 'mixed_bfloat16'
  # can have significant impact on model speeds by utilizing float16 in case of
  # GPUs, and bfloat16 in the case of TPUs. loss_scale takes effect only when
  # dtype is float16
  if params.runtime.mixed_precision_dtype:
    performance.set_mixed_precision_policy(params.runtime.mixed_precision_dtype,
                                           params.runtime.loss_scale)
  if params.runtime.worker_hosts != '' and params.runtime.worker_hosts is not None:
    num_workers = distribute_utils.configure_cluster(
        worker_hosts=params.runtime.worker_hosts,
        task_index=params.runtime.task_index)
    print(num_workers)
  distribution_strategy = distribute_utils.get_distribution_strategy(
      distribution_strategy=params.runtime.distribution_strategy,
      all_reduce_alg=params.runtime.all_reduce_alg,
      num_gpus=params.runtime.num_gpus,
      tpu_address=params.runtime.tpu)
  with distribution_strategy.scope():
    task = task_factory.get_task(params.task, logging_dir=model_dir)

  train_lib.run_experiment(
      distribution_strategy=distribution_strategy,
      task=task,
      mode=FLAGS.mode,
      params=params,
      model_dir=model_dir)


if __name__ == '__main__':
  import datetime

  a = datetime.datetime.now()
  tfm_flags.define_flags()
  app.run(main)
  b = datetime.datetime.now()

  print('\n\n\n\n\n\n\n {b - a}')