model_main.py faster-rcnn CUDA_ERROR_OUT_OF_MEMORY

描述：

我能够使用 legacy/train.py 训练更快的 rcnn 模型，但是当我尝试使用相同的配置设置使用 model_main.py 进行训练时，它遇到了如下问题。图像分辨率： 1920x1080

tensorflow/stream_executor/cuda/cuda_driver.cc:890] failed to alloc 8589934592 bytes on host: CUDA_ERROR_OUT_OF_MEMORY: out of memory
.tensorflow/core/common_runtime/gpu/gpu_host_allocator.h:44] could not allocate pinned host memory of size: 8589934592
tensorflow/core/common_runtime/bfc_allocator.cc:764] Bin (256):     Total Chunks: 4753, Chunks in use: 4753. 1.16MiB allocated for chunks. 1.16MiB in use in bin. 144.3KiB client-requested in use in bin.
tensorflow/core/common_runtime/bfc_allocator.cc:800] InUse at 0000000203800000 next 1 of size 256

我尝试过的：

将批大小设置为 1
使用内存增长

配置 = tf。配置原型((
config.gpu_options.allow_growth = True
会话 = TF。会话(配置=配置(

或

session_config = tf。配置原型((
session_config.gpu_options.allow_growth = True
config = tf.estimator.RunConfig(model_dir=FLAGS.model_dir， session_config=session_config， log_step_count_steps=10， save_summary_steps=20， keep_checkpoint_max=20， save_checkpoints_steps=100(

不要分配整个 GPU 内存

配置 = tf。配置原型((
config.gpu_options.per_process_gpu_memory_fraction = 0.6
会话 = TF。会话(配置=配置(

或

session_config = tf。配置原型((
session_config.gpu_options.per_process_gpu_memory_fraction = 0.6
config = tf.estimator.RunConfig(model_dir=FLAGS.model_dir， session_config=session_config， log_step_count_steps=10， save_summary_steps=20， keep_checkpoint_max=20， save_checkpoints_steps=100(

张量流CUDA_ERROR_OUT_OF_MEMORY

设置queue_capacity、min_after_dequeue、num_readers、batch_queue_capacity、num_batch_queue_threads prefetch_queue_capacity

在大图像上训练时内存不足

min_dimension，max_dimension降低到 270， 480

这些都不适合我。

环境：

操作系统平台和分发：Win 10 专业版：1909
TensorFlow 安装自：pip tensorflow-gpu
TensorFlow 版本 1.14
对象检测：0.1 CUDA/cuDNN 版本：Cuda 10.0、Cudnn 10.0
GPU型号和内存：NVIDIA GeForce RTX 2070 SUPER，内存8 G
系统内存：32G

我的配置：

# Faster R-CNN with Inception v2, configured for Oxford-IIIT Pets Dataset.
# Users should configure the fine_tune_checkpoint field in the train config as
# well as the label_map_path and input_path fields in the train_input_reader and
# eval_input_reader. Search for "PATH_TO_BE_CONFIGURED" to find the fields that
# should be configured.
model {
faster_rcnn {
num_classes: 2
image_resizer {
keep_aspect_ratio_resizer {
min_dimension: 1080
max_dimension: 1920
}
}
feature_extractor {
type: 'faster_rcnn_inception_v2'
first_stage_features_stride: 16
}
first_stage_anchor_generator {
grid_anchor_generator {
scales: [0.25, 0.5, 1.0, 2.0]
aspect_ratios: [0.5, 1.0, 2.0]
height_stride: 16
width_stride: 16
}
}
first_stage_box_predictor_conv_hyperparams {
op: CONV
regularizer {
l2_regularizer {
weight: 0.0
}
}
initializer {
truncated_normal_initializer {
stddev: 0.01
}
}
}
first_stage_nms_score_threshold: 0.0
first_stage_nms_iou_threshold: 0.7
first_stage_max_proposals: 300
first_stage_localization_loss_weight: 2.0
first_stage_objectness_loss_weight: 1.0
initial_crop_size: 14
maxpool_kernel_size: 2
maxpool_stride: 2
second_stage_box_predictor {
mask_rcnn_box_predictor {
use_dropout: false
dropout_keep_probability: 1.0
fc_hyperparams {
op: FC
regularizer {
l2_regularizer {
weight: 0.0
}
}
initializer {
variance_scaling_initializer {
factor: 1.0
uniform: true
mode: FAN_AVG
}
}
}
}
}
second_stage_post_processing {
batch_non_max_suppression {
score_threshold: 0.0
iou_threshold: 0.6
max_detections_per_class: 100
max_total_detections: 300
}
score_converter: SOFTMAX
}
second_stage_localization_loss_weight: 2.0
second_stage_classification_loss_weight: 1.0
}
}
train_config: {
batch_size: 1
optimizer {
momentum_optimizer: {
learning_rate: {
manual_step_learning_rate {
initial_learning_rate: 0.0002
schedule {
step: 900000
learning_rate: .00002
}
schedule {
step: 1200000
learning_rate: .000002
}
}
}
momentum_optimizer_value: 0.9
}
use_moving_average: false
}
gradient_clipping_by_norm: 10.0
fine_tune_checkpoint: ""
from_detection_checkpoint: true
load_all_detection_checkpoint_vars: true
# Note: The below line limits the training process to 200K steps, which we
# empirically found to be sufficient enough to train the pets dataset. This
# effectively bypasses the learning rate schedule (the learning rate will
# never decay). Remove the below line to train indefinitely.
num_steps: 200000
data_augmentation_options {
random_horizontal_flip {
}
}
batch_queue_capacity: 60
num_batch_queue_threads: 30
prefetch_queue_capacity: 40
}

train_input_reader: {
tf_record_input_reader {
input_path: "D:\object_detection\train_data\train.record"
}
label_map_path: "D:\object_detection\pascal_label_map.pbtxt"
queue_capacity: 2
min_after_dequeue: 1
num_readers: 1
}
eval_config: {
metrics_set: "coco_detection_metrics"
num_examples: 1101
}
eval_input_reader: {
tf_record_input_reader {
input_path: "D:\object_detection\eval_data\eval.record"
}
label_map_path: "D:\object_detection\pascal_label_map.pbtxt"
shuffle: false
num_readers: 1
}

如果有其他解决方案，我将非常感谢您。

对象检测模型会消耗大量内存。这是因为它们的工作方式以及它们为查找框而生成的大量定位点。

你做得很好，但你的GPU不足以训练这些模型。您可以执行的操作：

减小图像大小，例如 720x512
使用 SGD 作为优化器，而不是其他优化器(如 Adam(。SGD 消耗的内存大约是 Adam 的 3 倍。

还值得一提的是，您在处理 1 个实例的小批量时做得很好。如果我没记错的话，FasterRCNN 每批只有 2 张图像进行训练

我刚刚发现，如果我将batch_size设置为 3，那么它可以正常工作。当我batch_size设置回 1 时，它遇到了 OOM 问题。

这很奇怪，我仍然不知道为什么，因为它应该始终以较小的批量大小节省内存。

如果您遇到相同的情况，可以尝试稍微增加批量大小，但我不能保证它会起作用。

相关内容

最新更新

热门标签：