The mAP value starting from 0.05% although the freeze block is set. Shouldn't it preserve the features of the base peoplenet model?

Please provide the following information when requesting support.

• Hardware: T4
• Network Type: Detectnet_v2
• TLT Version
• Training spec file
random_seed: 42
dataset_config {
data_sources {
tfrecords_path: “/shared/users/3ce93a7c-0889-537e-8991-7d76b64a7194/datasets/c71df972-db7b-45af-ae21-976bf2c34d8d/tfrecords/"
image_directory_path: “/shared/users/3ce93a7c-0889-537e-8991-7d76b64a7194/datasets/c71df972-db7b-45af-ae21-976bf2c34d8d/”
}
image_extension: “jpg”
target_class_mapping {
key: “face”
value: “face”
}
target_class_mapping {
key: “person”
value: “person”
}
validation_data_source {
tfrecords_path: "/shared/users/3ce93a7c-0889-537e-8991-7d76b64a7194/datasets/84096349-00b9-4ab5-93cc-3edcb12dda4d/tfrecords/

image_directory_path: “/shared/users/3ce93a7c-0889-537e-8991-7d76b64a7194/datasets/84096349-00b9-4ab5-93cc-3edcb12dda4d/”
}
}
augmentation_config {
preprocessing {
output_image_width: 960
output_image_height: 544
min_bbox_width: 1.0
min_bbox_height: 1.0
output_image_channel: 3
}
spatial_augmentation {
hflip_probability: 0.5
zoom_min: 1.0
zoom_max: 1.0
translate_max_x: 8.0
translate_max_y: 8.0
}
color_augmentation {
hue_rotation_max: 25.0
saturation_shift_max: 0.20000000298023224
contrast_scale_max: 0.10000000149011612
contrast_center: 0.5
}
}
postprocessing_config {
target_class_config {
key: “face”
value {
clustering_config {
coverage_threshold: 0.007499999832361937
minimum_bounding_box_height: 20
dbscan_eps: 0.23000000417232513
dbscan_min_samples: 1
dbscan_confidence_threshold: 0.7200000286102295
nms_iou_threshold: 0.20000000298023224
}
}
}
target_class_config {
key: “person”
value {
clustering_config {
coverage_threshold: 0.007499999832361937
minimum_bounding_box_height: 20
dbscan_eps: 0.23000000417232513
dbscan_min_samples: 1
dbscan_confidence_threshold: 0.10000000149011612
nms_iou_threshold: 0.20000000298023224
}
}
}
}
model_config {
pretrained_model_file: “/shared/users/3ce93a7c-0889-537e-8991-7d76b64a7194/models/0870c02c-f79d-4bad-bf92-3a1a5f77d464/peoplenet_vtrainable_v2.6/resnet34_peoplenet.tlt”
num_layers: 34
use_batch_norm: true
objective_set {
bbox {
scale: 35.0
offset: 0.5
}
cov {
}
}
freeze_blocks: 0.0
freeze_blocks: 1.0
freeze_blocks: 2.0
freeze_blocks: 3.0
freeze_blocks: 4.0
arch: “resnet”
}
evaluation_config {
validation_period_during_training: 1
first_validation_epoch: 1
minimum_detection_ground_truth_overlap {
key: “face”
value: 0.5
}
minimum_detection_ground_truth_overlap {
key: “person”
value: 0.5
}
evaluation_box_config {
key: “face”
value {
minimum_height: 20
maximum_height: 9999
minimum_width: 10
maximum_width: 9999
}
}
evaluation_box_config {
key: “person”
value {
minimum_height: 20
maximum_height: 9999
minimum_width: 10
maximum_width: 9999
}
}
}
cost_function_config {
target_classes {
name: “person”
class_weight: 4.0
coverage_foreground_weight: 0.05000000074505806
objectives {
name: “cov”
initial_weight: 1.0
weight_target: 1.0
}
objectives {
name: “bbox”
initial_weight: 10.0
weight_target: 10.0
}
}
target_classes {
name: “face”
class_weight: 4.0
coverage_foreground_weight: 0.05000000074505806
objectives {
name: “cov”
initial_weight: 1.0
weight_target: 1.0
}
objectives {
name: “bbox”
initial_weight: 10.0
weight_target: 10.0
}
}
enable_autoweighting: true
max_objective_weight: 0.9998999834060669
min_objective_weight: 9.999999747378752e-05
}
training_config {
batch_size_per_gpu: 4
num_epochs: 100
learning_rate {
soft_start_annealing_schedule {
min_learning_rate: 3.5999998999614036e-06
max_learning_rate: 0.0005000000237487257
soft_start: 1.0000000116860974e-07
annealing: 1.0000000116860974e-07
}
}
regularizer {
type: L2
weight: 3.000000026176508e-09
}
optimizer {
adam {
epsilon: 9.99999993922529e-09
beta1: 0.8999999761581421
beta2: 0.9990000128746033
}
}
cost_scaling {
initial_exponent: 20.0
increment: 0.005
decrement: 1.0
}
checkpoint_interval: 1
visualizer {
num_images: 3
infrequent_logging_frequency: 5
}
}
bbox_rasterizer_config {
target_class_config {
key: “face”
value {
cov_center_x: 0.5
cov_center_y: 0.5
cov_radius_x: 1.0
cov_radius_y: 1.0
bbox_min_radius: 1.0
}
}
target_class_config {
key: “person”
value {
cov_center_x: 0.5
cov_center_y: 0.5
cov_radius_x: 1.0
cov_radius_y: 1.0
bbox_min_radius: 1.0
}
}
deadzone_radius: 0.4000000059604645
}

Results logs:
2024-03-20 12:32:35,976 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2c_bn_2 weights set from pre-trained model.
2024-03-20 12:32:35,977 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_6 weights set from pre-trained model.
2024-03-20 12:32:35,977 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2c_relu weights set from pre-trained model.
2024-03-20 12:32:36,451 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_conv_1 weights set from pre-trained model.
2024-03-20 12:32:36,934 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_bn_1 weights set from pre-trained model.
2024-03-20 12:32:36,935 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_relu_1 weights set from pre-trained model.
2024-03-20 12:32:37,405 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_conv_2 weights set from pre-trained model.
2024-03-20 12:32:37,882 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_bn_2 weights set from pre-trained model.
2024-03-20 12:32:37,882 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_7 weights set from pre-trained model.
2024-03-20 12:32:37,882 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_2d_relu weights set from pre-trained model.
2024-03-20 12:32:38,361 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_conv_1 weights set from pre-trained model.
2024-03-20 12:32:38,845 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_bn_1 weights set from pre-trained model.
2024-03-20 12:32:38,845 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_relu_1 weights set from pre-trained model.
2024-03-20 12:32:39,325 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_conv_2 weights set from pre-trained model.
2024-03-20 12:32:39,792 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_conv_shortcut weights set from pre-trained model.
2024-03-20 12:32:40,282 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_bn_2 weights set from pre-trained model.
2024-03-20 12:32:40,765 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_bn_shortcut weights set from pre-trained model.
2024-03-20 12:32:40,765 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_8 weights set from pre-trained model.
2024-03-20 12:32:40,765 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3a_relu weights set from pre-trained model.
2024-03-20 12:32:41,240 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_conv_1 weights set from pre-trained model.
2024-03-20 12:32:41,722 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_bn_1 weights set from pre-trained model.
2024-03-20 12:32:41,722 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_relu_1 weights set from pre-trained model.
2024-03-20 12:32:42,202 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_conv_2 weights set from pre-trained model.
2024-03-20 12:32:42,682 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_bn_2 weights set from pre-trained model.
2024-03-20 12:32:42,682 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_9 weights set from pre-trained model.
2024-03-20 12:32:42,682 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3b_relu weights set from pre-trained model.
2024-03-20 12:32:43,166 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_conv_1 weights set from pre-trained model.
2024-03-20 12:32:43,649 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_bn_1 weights set from pre-trained model.
2024-03-20 12:32:43,649 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_relu_1 weights set from pre-trained model.
2024-03-20 12:32:44,133 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_conv_2 weights set from pre-trained model.
2024-03-20 12:32:44,624 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_bn_2 weights set from pre-trained model.
2024-03-20 12:32:44,624 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_10 weights set from pre-trained model.
2024-03-20 12:32:44,624 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3c_relu weights set from pre-trained model.
2024-03-20 12:32:45,106 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_conv_1 weights set from pre-trained model.
2024-03-20 12:32:45,614 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_bn_1 weights set from pre-trained model.
2024-03-20 12:32:45,614 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_relu_1 weights set from pre-trained model.
2024-03-20 12:32:46,104 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_conv_2 weights set from pre-trained model.
2024-03-20 12:32:46,593 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_bn_2 weights set from pre-trained model.
2024-03-20 12:32:46,593 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_11 weights set from pre-trained model.
2024-03-20 12:32:46,594 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3d_relu weights set from pre-trained model.
2024-03-20 12:32:47,080 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_conv_1 weights set from pre-trained model.
2024-03-20 12:32:47,566 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_bn_1 weights set from pre-trained model.
2024-03-20 12:32:47,566 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_relu_1 weights set from pre-trained model.
2024-03-20 12:32:48,051 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_conv_2 weights set from pre-trained model.
2024-03-20 12:32:48,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_bn_2 weights set from pre-trained model.
2024-03-20 12:32:48,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_12 weights set from pre-trained model.
2024-03-20 12:32:48,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3e_relu weights set from pre-trained model.
2024-03-20 12:32:49,038 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_conv_1 weights set from pre-trained model.
2024-03-20 12:32:49,527 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_bn_1 weights set from pre-trained model.
2024-03-20 12:32:49,527 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_relu_1 weights set from pre-trained model.
2024-03-20 12:32:50,010 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_conv_2 weights set from pre-trained model.
2024-03-20 12:32:50,510 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_bn_2 weights set from pre-trained model.
2024-03-20 12:32:50,510 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_13 weights set from pre-trained model.
2024-03-20 12:32:50,510 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_3f_relu weights set from pre-trained model.
2024-03-20 12:32:51,007 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_conv_1 weights set from pre-trained model.
2024-03-20 12:32:51,518 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_bn_1 weights set from pre-trained model.
2024-03-20 12:32:51,518 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_relu_1 weights set from pre-trained model.
2024-03-20 12:32:52,013 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_conv_2 weights set from pre-trained model.
2024-03-20 12:32:52,509 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_conv_shortcut weights set from pre-trained model.
2024-03-20 12:32:53,014 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_bn_2 weights set from pre-trained model.
2024-03-20 12:32:53,519 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_bn_shortcut weights set from pre-trained model.
2024-03-20 12:32:53,519 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_14 weights set from pre-trained model.
2024-03-20 12:32:53,519 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4a_relu weights set from pre-trained model.
2024-03-20 12:32:54,015 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_conv_1 weights set from pre-trained model.
2024-03-20 12:32:54,521 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_bn_1 weights set from pre-trained model.
2024-03-20 12:32:54,521 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_relu_1 weights set from pre-trained model.
2024-03-20 12:32:55,021 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_conv_2 weights set from pre-trained model.
2024-03-20 12:32:55,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_bn_2 weights set from pre-trained model.
2024-03-20 12:32:55,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_15 weights set from pre-trained model.
2024-03-20 12:32:55,544 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4b_relu weights set from pre-trained model.
2024-03-20 12:32:56,056 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_conv_1 weights set from pre-trained model.
2024-03-20 12:32:56,561 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_bn_1 weights set from pre-trained model.
2024-03-20 12:32:56,561 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_relu_1 weights set from pre-trained model.
2024-03-20 12:32:57,069 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_conv_2 weights set from pre-trained model.
2024-03-20 12:32:57,566 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_bn_2 weights set from pre-trained model.
2024-03-20 12:32:57,566 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer add_16 weights set from pre-trained model.
2024-03-20 12:32:57,566 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.model.detectnet_model 142: Layer block_4c_relu weights set from pre-trained model.
2024-03-20 12:32:57,700 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.objectives.bbox_objective 78: Default L1 loss function will be used.


Layer (type) Output Shape Param # Connected to

input_1 (InputLayer) (None, 3, 544, 960) 0


conv1 (Conv2D) (None, 64, 272, 480) 9472 input_1[0][0]


bn_conv1 (BatchNormalization) (None, 64, 272, 480) 256 conv1[0][0]


activation_1 (Activation) (None, 64, 272, 480) 0 bn_conv1[0][0]


block_1a_conv_1 (Conv2D) (None, 64, 136, 240) 36928 activation_1[0][0]


block_1a_bn_1 (BatchNormalizati (None, 64, 136, 240) 256 block_1a_conv_1[0][0]


block_1a_relu_1 (Activation) (None, 64, 136, 240) 0 block_1a_bn_1[0][0]


block_1a_conv_2 (Conv2D) (None, 64, 136, 240) 36928 block_1a_relu_1[0][0]


block_1a_conv_shortcut (Conv2D) (None, 64, 136, 240) 4160 activation_1[0][0]


block_1a_bn_2 (BatchNormalizati (None, 64, 136, 240) 256 block_1a_conv_2[0][0]


block_1a_bn_shortcut (BatchNorm (None, 64, 136, 240) 256 block_1a_conv_shortcut[0][0]


add_1 (Add) (None, 64, 136, 240) 0 block_1a_bn_2[0][0]
block_1a_bn_shortcut[0][0]


block_1a_relu (Activation) (None, 64, 136, 240) 0 add_1[0][0]


block_1b_conv_1 (Conv2D) (None, 64, 136, 240) 36928 block_1a_relu[0][0]


block_1b_bn_1 (BatchNormalizati (None, 64, 136, 240) 256 block_1b_conv_1[0][0]


block_1b_relu_1 (Activation) (None, 64, 136, 240) 0 block_1b_bn_1[0][0]


block_1b_conv_2 (Conv2D) (None, 64, 136, 240) 36928 block_1b_relu_1[0][0]


block_1b_bn_2 (BatchNormalizati (None, 64, 136, 240) 256 block_1b_conv_2[0][0]


add_2 (Add) (None, 64, 136, 240) 0 block_1b_bn_2[0][0]
block_1a_relu[0][0]


block_1b_relu (Activation) (None, 64, 136, 240) 0 add_2[0][0]


block_1c_conv_1 (Conv2D) (None, 64, 136, 240) 36928 block_1b_relu[0][0]


block_1c_bn_1 (BatchNormalizati (None, 64, 136, 240) 256 block_1c_conv_1[0][0]


block_1c_relu_1 (Activation) (None, 64, 136, 240) 0 block_1c_bn_1[0][0]


block_1c_conv_2 (Conv2D) (None, 64, 136, 240) 36928 block_1c_relu_1[0][0]


block_1c_bn_2 (BatchNormalizati (None, 64, 136, 240) 256 block_1c_conv_2[0][0]


add_3 (Add) (None, 64, 136, 240) 0 block_1c_bn_2[0][0]
block_1b_relu[0][0]


block_1c_relu (Activation) (None, 64, 136, 240) 0 add_3[0][0]


block_2a_conv_1 (Conv2D) (None, 128, 68, 120) 73856 block_1c_relu[0][0]


block_2a_bn_1 (BatchNormalizati (None, 128, 68, 120) 512 block_2a_conv_1[0][0]


block_2a_relu_1 (Activation) (None, 128, 68, 120) 0 block_2a_bn_1[0][0]


block_2a_conv_2 (Conv2D) (None, 128, 68, 120) 147584 block_2a_relu_1[0][0]


block_2a_conv_shortcut (Conv2D) (None, 128, 68, 120) 8320 block_1c_relu[0][0]


block_2a_bn_2 (BatchNormalizati (None, 128, 68, 120) 512 block_2a_conv_2[0][0]


block_2a_bn_shortcut (BatchNorm (None, 128, 68, 120) 512 block_2a_conv_shortcut[0][0]


add_4 (Add) (None, 128, 68, 120) 0 block_2a_bn_2[0][0]
block_2a_bn_shortcut[0][0]


block_2a_relu (Activation) (None, 128, 68, 120) 0 add_4[0][0]


block_2b_conv_1 (Conv2D) (None, 128, 68, 120) 147584 block_2a_relu[0][0]


block_2b_bn_1 (BatchNormalizati (None, 128, 68, 120) 512 block_2b_conv_1[0][0]


block_2b_relu_1 (Activation) (None, 128, 68, 120) 0 block_2b_bn_1[0][0]


block_2b_conv_2 (Conv2D) (None, 128, 68, 120) 147584 block_2b_relu_1[0][0]


block_2b_bn_2 (BatchNormalizati (None, 128, 68, 120) 512 block_2b_conv_2[0][0]


add_5 (Add) (None, 128, 68, 120) 0 block_2b_bn_2[0][0]
block_2a_relu[0][0]


block_2b_relu (Activation) (None, 128, 68, 120) 0 add_5[0][0]


block_2c_conv_1 (Conv2D) (None, 128, 68, 120) 147584 block_2b_relu[0][0]


block_2c_bn_1 (BatchNormalizati (None, 128, 68, 120) 512 block_2c_conv_1[0][0]


block_2c_relu_1 (Activation) (None, 128, 68, 120) 0 block_2c_bn_1[0][0]


block_2c_conv_2 (Conv2D) (None, 128, 68, 120) 147584 block_2c_relu_1[0][0]


block_2c_bn_2 (BatchNormalizati (None, 128, 68, 120) 512 block_2c_conv_2[0][0]


add_6 (Add) (None, 128, 68, 120) 0 block_2c_bn_2[0][0]
block_2b_relu[0][0]


block_2c_relu (Activation) (None, 128, 68, 120) 0 add_6[0][0]


block_2d_conv_1 (Conv2D) (None, 128, 68, 120) 147584 block_2c_relu[0][0]


block_2d_bn_1 (BatchNormalizati (None, 128, 68, 120) 512 block_2d_conv_1[0][0]


block_2d_relu_1 (Activation) (None, 128, 68, 120) 0 block_2d_bn_1[0][0]


block_2d_conv_2 (Conv2D) (None, 128, 68, 120) 147584 block_2d_relu_1[0][0]


block_2d_bn_2 (BatchNormalizati (None, 128, 68, 120) 512 block_2d_conv_2[0][0]


add_7 (Add) (None, 128, 68, 120) 0 block_2d_bn_2[0][0]
block_2c_relu[0][0]


block_2d_relu (Activation) (None, 128, 68, 120) 0 add_7[0][0]


block_3a_conv_1 (Conv2D) (None, 256, 34, 60) 295168 block_2d_relu[0][0]


block_3a_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3a_conv_1[0][0]


block_3a_relu_1 (Activation) (None, 256, 34, 60) 0 block_3a_bn_1[0][0]


block_3a_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3a_relu_1[0][0]


block_3a_conv_shortcut (Conv2D) (None, 256, 34, 60) 33024 block_2d_relu[0][0]


block_3a_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3a_conv_2[0][0]


block_3a_bn_shortcut (BatchNorm (None, 256, 34, 60) 1024 block_3a_conv_shortcut[0][0]


add_8 (Add) (None, 256, 34, 60) 0 block_3a_bn_2[0][0]
block_3a_bn_shortcut[0][0]


block_3a_relu (Activation) (None, 256, 34, 60) 0 add_8[0][0]


block_3b_conv_1 (Conv2D) (None, 256, 34, 60) 590080 block_3a_relu[0][0]


block_3b_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3b_conv_1[0][0]


block_3b_relu_1 (Activation) (None, 256, 34, 60) 0 block_3b_bn_1[0][0]


block_3b_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3b_relu_1[0][0]


block_3b_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3b_conv_2[0][0]


add_9 (Add) (None, 256, 34, 60) 0 block_3b_bn_2[0][0]
block_3a_relu[0][0]


block_3b_relu (Activation) (None, 256, 34, 60) 0 add_9[0][0]


block_3c_conv_1 (Conv2D) (None, 256, 34, 60) 590080 block_3b_relu[0][0]


block_3c_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3c_conv_1[0][0]


block_3c_relu_1 (Activation) (None, 256, 34, 60) 0 block_3c_bn_1[0][0]


block_3c_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3c_relu_1[0][0]


block_3c_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3c_conv_2[0][0]


add_10 (Add) (None, 256, 34, 60) 0 block_3c_bn_2[0][0]
block_3b_relu[0][0]


block_3c_relu (Activation) (None, 256, 34, 60) 0 add_10[0][0]


block_3d_conv_1 (Conv2D) (None, 256, 34, 60) 590080 block_3c_relu[0][0]


block_3d_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3d_conv_1[0][0]


block_3d_relu_1 (Activation) (None, 256, 34, 60) 0 block_3d_bn_1[0][0]


block_3d_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3d_relu_1[0][0]


block_3d_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3d_conv_2[0][0]


add_11 (Add) (None, 256, 34, 60) 0 block_3d_bn_2[0][0]
block_3c_relu[0][0]


block_3d_relu (Activation) (None, 256, 34, 60) 0 add_11[0][0]


block_3e_conv_1 (Conv2D) (None, 256, 34, 60) 590080 block_3d_relu[0][0]


block_3e_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3e_conv_1[0][0]


block_3e_relu_1 (Activation) (None, 256, 34, 60) 0 block_3e_bn_1[0][0]


block_3e_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3e_relu_1[0][0]


block_3e_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3e_conv_2[0][0]


add_12 (Add) (None, 256, 34, 60) 0 block_3e_bn_2[0][0]
block_3d_relu[0][0]


block_3e_relu (Activation) (None, 256, 34, 60) 0 add_12[0][0]


block_3f_conv_1 (Conv2D) (None, 256, 34, 60) 590080 block_3e_relu[0][0]


block_3f_bn_1 (BatchNormalizati (None, 256, 34, 60) 1024 block_3f_conv_1[0][0]


block_3f_relu_1 (Activation) (None, 256, 34, 60) 0 block_3f_bn_1[0][0]


block_3f_conv_2 (Conv2D) (None, 256, 34, 60) 590080 block_3f_relu_1[0][0]


block_3f_bn_2 (BatchNormalizati (None, 256, 34, 60) 1024 block_3f_conv_2[0][0]


add_13 (Add) (None, 256, 34, 60) 0 block_3f_bn_2[0][0]
block_3e_relu[0][0]


block_3f_relu (Activation) (None, 256, 34, 60) 0 add_13[0][0]


block_4a_conv_1 (Conv2D) (None, 512, 34, 60) 1180160 block_3f_relu[0][0]


block_4a_bn_1 (BatchNormalizati (None, 512, 34, 60) 2048 block_4a_conv_1[0][0]


block_4a_relu_1 (Activation) (None, 512, 34, 60) 0 block_4a_bn_1[0][0]


block_4a_conv_2 (Conv2D) (None, 512, 34, 60) 2359808 block_4a_relu_1[0][0]


block_4a_conv_shortcut (Conv2D) (None, 512, 34, 60) 131584 block_3f_relu[0][0]


block_4a_bn_2 (BatchNormalizati (None, 512, 34, 60) 2048 block_4a_conv_2[0][0]


block_4a_bn_shortcut (BatchNorm (None, 512, 34, 60) 2048 block_4a_conv_shortcut[0][0]


add_14 (Add) (None, 512, 34, 60) 0 block_4a_bn_2[0][0]
block_4a_bn_shortcut[0][0]


block_4a_relu (Activation) (None, 512, 34, 60) 0 add_14[0][0]


block_4b_conv_1 (Conv2D) (None, 512, 34, 60) 2359808 block_4a_relu[0][0]


block_4b_bn_1 (BatchNormalizati (None, 512, 34, 60) 2048 block_4b_conv_1[0][0]


block_4b_relu_1 (Activation) (None, 512, 34, 60) 0 block_4b_bn_1[0][0]


block_4b_conv_2 (Conv2D) (None, 512, 34, 60) 2359808 block_4b_relu_1[0][0]


block_4b_bn_2 (BatchNormalizati (None, 512, 34, 60) 2048 block_4b_conv_2[0][0]
2024-03-20 12:32:57,725 [TAO Toolkit] [INFO] root 2102: DetectNet V2 model built.
2024-03-20 12:32:57,725 [TAO Toolkit] [INFO] root 2102: Building rasterizer.
2024-03-20 12:32:57,726 [TAO Toolkit] [INFO] root 2102: Rasterizers built.
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/training_proto_utilities.py:102: The name tf.train.get_or_create_global_step is deprecated. Please use tf.compat.v1.train.get_or_create_global_step instead.

2024-03-20 12:32:57,726 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/training_proto_utilities.py:102: The name tf.train.get_or_create_global_step is deprecated. Please use tf.compat.v1.train.get_or_create_global_step instead.

WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/scripts/train.py:718: The name tf.summary.scalar is deprecated. Please use tf.compat.v1.summary.scalar instead.

2024-03-20 12:32:57,741 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/scripts/train.py:718: The name tf.summary.scalar is deprecated. Please use tf.compat.v1.summary.scalar instead.

2024-03-20 12:32:57,742 [TAO Toolkit] [INFO] root 2102: Building training graph.
2024-03-20 12:32:57,744 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 175: Serial augmentation enabled = False
2024-03-20 12:32:57,744 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 177: Pseudo sharding enabled = False
2024-03-20 12:32:57,744 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 269: Max Image Dimensions (all sources): (0, 0)
2024-03-20 12:32:57,744 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 380: number of cpus: 48, io threads: 96, compute threads: 48, buffered batches: 4
2024-03-20 12:32:57,744 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 387: total dataset size 3799, number of sources: 1, batch size per gpu: 4, steps: 950
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/tensorflow_core/python/autograph/converters/directives.py:119: The name tf.set_random_seed is deprecated. Please use tf.compat.v1.set_random_seed instead.

2024-03-20 12:32:57,784 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/tensorflow_core/python/autograph/converters/directives.py:119: The name tf.set_random_seed is deprecated. Please use tf.compat.v1.set_random_seed instead.

2024-03-20 12:32:59,766 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.dataloader.default_dataloader 546: Bounding box coordinates were detected in the input specification! Bboxes will be automatically converted to polygon coordinates.
2024-03-20 12:33:02,487 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 409: shuffle: True - shard 0 of 1
2024-03-20 12:33:02,491 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 479: sampling 1 datasets with weights:
2024-03-20 12:33:02,491 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 481: source: 0 weight: 1.000000
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/tensorflow_core/python/autograph/converters/directives.py:119: The name tf.image.resize_images is deprecated. Please use tf.image.resize instead.

2024-03-20 12:33:03,254 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/tensorflow_core/python/autograph/converters/directives.py:119: The name tf.image.resize_images is deprecated. Please use tf.image.resize instead.

2024-03-20 12:33:04,034 [TAO Toolkit] [INFO] main 536: Found 3799 samples in training set
2024-03-20 12:33:04,035 [TAO Toolkit] [INFO] root 2102: Rasterizing tensors.
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/rasterizers/bbox_rasterizer.py:348: The name tf.bincount is deprecated. Please use tf.math.bincount instead.

2024-03-20 12:33:04,136 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/rasterizers/bbox_rasterizer.py:348: The name tf.bincount is deprecated. Please use tf.math.bincount instead.

2024-03-20 12:33:04,249 [TAO Toolkit] [INFO] root 2102: Tensors rasterized.
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/training_proto_utilities.py:49: The name tf.train.AdamOptimizer is deprecated. Please use tf.compat.v1.train.AdamOptimizer instead.

2024-03-20 12:33:04,249 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/training_proto_utilities.py:49: The name tf.train.AdamOptimizer is deprecated. Please use tf.compat.v1.train.AdamOptimizer instead.

WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/cost_function/cost_functions.py:29: The name tf.log is deprecated. Please use tf.math.log instead.

2024-03-20 12:33:04,500 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/cost_function/cost_functions.py:29: The name tf.log is deprecated. Please use tf.math.log instead.

WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/cost_function/cost_auto_weight_hook.py:250: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.

2024-03-20 12:33:04,524 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/cost_function/cost_auto_weight_hook.py:250: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.

2024-03-20 12:33:05,958 [TAO Toolkit] [INFO] root 2102: Training graph built.
2024-03-20 12:33:05,958 [TAO Toolkit] [INFO] root 2102: Building validation graph.
2024-03-20 12:33:05,959 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 175: Serial augmentation enabled = False
2024-03-20 12:33:05,959 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 177: Pseudo sharding enabled = False
2024-03-20 12:33:05,959 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 269: Max Image Dimensions (all sources): (0, 0)
2024-03-20 12:33:05,959 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 380: number of cpus: 48, io threads: 96, compute threads: 48, buffered batches: 4
2024-03-20 12:33:05,959 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 387: total dataset size 898, number of sources: 1, batch size per gpu: 4, steps: 225
2024-03-20 12:33:05,989 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.dataloader.default_dataloader 546: Bounding box coordinates were detected in the input specification! Bboxes will be automatically converted to polygon coordinates.
2024-03-20 12:33:06,232 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 409: shuffle: False - shard 0 of 1
2024-03-20 12:33:06,236 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 479: sampling 1 datasets with weights:
2024-03-20 12:33:06,236 [TAO Toolkit] [INFO] nvidia_tao_tf1.blocks.multi_source_loader.data_loader 481: source: 0 weight: 1.000000
2024-03-20 12:33:06,476 [TAO Toolkit] [INFO] main 591: Found 898 samples in validation set
2024-03-20 12:33:06,476 [TAO Toolkit] [INFO] root 2102: Rasterizing tensors.
2024-03-20 12:33:06,683 [TAO Toolkit] [INFO] root 2102: Tensors rasterized.
2024-03-20 12:33:07,324 [TAO Toolkit] [INFO] root 2102: Validation graph built.
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/tfhooks/validation_hook.py:58: The name tf.summary.FileWriterCache is deprecated. Please use tf.compat.v1.summary.FileWriterCache instead.

2024-03-20 12:33:07,325 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/tfhooks/validation_hook.py:58: The name tf.summary.FileWriterCache is deprecated. Please use tf.compat.v1.summary.FileWriterCache instead.

2024-03-20 12:33:08,825 [TAO Toolkit] [INFO] root 2102: Running training loop.
2024-03-20 12:33:08,825 [TAO Toolkit] [INFO] main 135: Checkpoint interval: 1
2024-03-20 12:33:08,825 [TAO Toolkit] [INFO] main 175: Scalars logged at every 95 steps
2024-03-20 12:33:08,826 [TAO Toolkit] [INFO] main 180: Images logged at every 0 steps
INFO:tensorflow:Create CheckpointSaverHook.
2024-03-20 12:33:08,828 [TAO Toolkit] [INFO] tensorflow 541: Create CheckpointSaverHook.
WARNING:tensorflow:From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/utilities.py:154: The name tf.train.SingularMonitoredSession is deprecated. Please use tf.compat.v1.train.SingularMonitoredSession instead.

2024-03-20 12:33:08,830 [TAO Toolkit] [WARNING] tensorflow 137: From /usr/local/lib/python3.8/dist-packages/nvidia_tao_tf1/cv/detectnet_v2/training/utilities.py:154: The name tf.train.SingularMonitoredSession is deprecated. Please use tf.compat.v1.train.SingularMonitoredSession instead.

INFO:tensorflow:Graph was finalized.
2024-03-20 12:33:11,750 [TAO Toolkit] [INFO] tensorflow 240: Graph was finalized.
INFO:tensorflow:Running local_init_op.
2024-03-20 12:33:15,183 [TAO Toolkit] [INFO] tensorflow 500: Running local_init_op.
INFO:tensorflow:Done running local_init_op.
2024-03-20 12:33:16,051 [TAO Toolkit] [INFO] tensorflow 502: Done running local_init_op.
INFO:tensorflow:Saving checkpoints for step-0.
2024-03-20 12:33:25,855 [TAO Toolkit] [INFO] tensorflow 81: Saving checkpoints for step-0.
INFO:tensorflow:epoch = 0.0, learning_rate = 1.6191649e-06, loss = 0.11848829, step = 0
2024-03-20 12:34:17,568 [TAO Toolkit] [INFO] tensorflow 262: epoch = 0.0, learning_rate = 1.6191649e-06, loss = 0.11848829, step = 0
2024-03-20 12:34:17,570 [TAO Toolkit] [INFO] root 2102: None
2024-03-20 12:34:17,570 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.tfhooks.task_progress_monitor_hook 149: Epoch 0/50: loss: 0.11849 learning rate: 1.6191649e-06 Time taken: 0:00:00 ETA: 0:00:00
2024-03-20 12:34:17,570 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 0.237
INFO:tensorflow:epoch = 0.002105263157894737, learning_rate = 1.6199589e-06, loss = 0.120211534, step = 2 (6.014 sec)
2024-03-20 12:34:23,582 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.002105263157894737, learning_rate = 1.6199589e-06, loss = 0.120211534, step = 2 (6.014 sec)
INFO:tensorflow:epoch = 0.02526315789473684, learning_rate = 1.628708e-06, loss = 0.11866148, step = 24 (5.432 sec)
2024-03-20 12:34:29,014 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.02526315789473684, learning_rate = 1.628708e-06, loss = 0.11866148, step = 24 (5.432 sec)
2024-03-20 12:34:29,014 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 3.530
INFO:tensorflow:epoch = 0.04842105263157895, learning_rate = 1.6375061e-06, loss = 0.11982238, step = 46 (5.361 sec)
2024-03-20 12:34:34,375 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.04842105263157895, learning_rate = 1.6375061e-06, loss = 0.11982238, step = 46 (5.361 sec)
2024-03-20 12:34:35,085 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 16.471
INFO:tensorflow:epoch = 0.07263157894736842, learning_rate = 1.6467552e-06, loss = 0.11932705, step = 69 (5.445 sec)
2024-03-20 12:34:39,820 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.07263157894736842, learning_rate = 1.6467552e-06, loss = 0.11932705, step = 69 (5.445 sec)
2024-03-20 12:34:41,006 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 16.889
INFO:tensorflow:epoch = 0.0968421052631579, learning_rate = 1.6560549e-06, loss = 0.11662158, step = 92 (5.448 sec)
2024-03-20 12:34:45,268 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.0968421052631579, learning_rate = 1.6560549e-06, loss = 0.11662158, step = 92 (5.448 sec)
INFO:tensorflow:global_step/sec: 3.34396
2024-03-20 12:34:45,979 [TAO Toolkit] [INFO] tensorflow 692: global_step/sec: 3.34396
2024-03-20 12:34:46,928 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 16.890
INFO:tensorflow:epoch = 0.12105263157894737, learning_rate = 1.6654088e-06, loss = 0.1136293, step = 115 (5.455 sec)
2024-03-20 12:34:50,723 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.12105263157894737, learning_rate = 1.6654088e-06, loss = 0.1136293, step = 115 (5.455 sec)
2024-03-20 12:34:52,863 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 16.849
INFO:tensorflow:epoch = 0.14526315789473684, learning_rate = 1.6748138e-06, loss = 0.114168674, step = 138 (5.463 sec)
2024-03-20 12:34:56,186 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.14526315789473684, learning_rate = 1.6748138e-06, loss = 0.114168674, step = 138 (5.463 sec)

2024-03-20 12:38:04,583 [TAO Toolkit] [INFO] nvidia_tao_tf1.core.hooks.sample_counter_hook 76: Train Samples / sec: 16.626
INFO:tensorflow:epoch = 0.9810526315789474, learning_rate = 2.0342488e-06, loss = 0.09080703, step = 932 (5.287 sec)
2024-03-20 12:38:06,502 [TAO Toolkit] [INFO] tensorflow 260: epoch = 0.9810526315789474, learning_rate = 2.0342488e-06, loss = 0.09080703, step = 932 (5.287 sec)
INFO:tensorflow:Saving checkpoints for step-950.
2024-03-20 12:38:10,589 [TAO Toolkit] [INFO] tensorflow 81: Saving checkpoints for step-950.
2024-03-20 12:38:14,192 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 0 / 224, 0.00s/step
2024-03-20 12:38:25,496 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 10 / 224, 1.13s/step
2024-03-20 12:38:35,179 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 20 / 224, 0.97s/step
2024-03-20 12:38:44,881 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 30 / 224, 0.97s/step
2024-03-20 12:38:54,526 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 40 / 224, 0.96s/step
2024-03-20 12:39:04,197 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 50 / 224, 0.97s/step
2024-03-20 12:39:14,048 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 60 / 224, 0.99s/step
2024-03-20 12:39:23,754 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 70 / 224, 0.97s/step
2024-03-20 12:39:33,639 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 80 / 224, 0.99s/step
2024-03-20 12:39:43,352 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 90 / 224, 0.97s/step
2024-03-20 12:39:53,099 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 100 / 224, 0.97s/step
2024-03-20 12:40:02,808 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 110 / 224, 0.97s/step
2024-03-20 12:40:12,443 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 120 / 224, 0.96s/step
2024-03-20 12:40:22,069 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 130 / 224, 0.96s/step
2024-03-20 12:40:31,620 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 140 / 224, 0.96s/step
2024-03-20 12:40:41,173 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 150 / 224, 0.96s/step
2024-03-20 12:40:50,750 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 160 / 224, 0.96s/step
2024-03-20 12:41:00,370 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 170 / 224, 0.96s/step
2024-03-20 12:41:10,069 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 180 / 224, 0.97s/step
2024-03-20 12:41:19,690 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 190 / 224, 0.96s/step
2024-03-20 12:41:29,325 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 200 / 224, 0.96s/step
2024-03-20 12:41:38,933 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 210 / 224, 0.96s/step
2024-03-20 12:41:48,513 [TAO Toolkit] [INFO] nvidia_tao_tf1.cv.detectnet_v2.evaluation.evaluation 130: step 220 / 224, 0.96s/step


add_15 (Add) (None, 512, 34, 60) 0 block_4b_bn_2[0][0]
block_4a_relu[0][0]


block_4b_relu (Activation) (None, 512, 34, 60) 0 add_15[0][0]


block_4c_conv_1 (Conv2D) (None, 512, 34, 60) 2359808 block_4b_relu[0][0]


block_4c_bn_1 (BatchNormalizati (None, 512, 34, 60) 2048 block_4c_conv_1[0][0]


block_4c_relu_1 (Activation) (None, 512, 34, 60) 0 block_4c_bn_1[0][0]


block_4c_conv_2 (Conv2D) (None, 512, 34, 60) 2359808 block_4c_relu_1[0][0]


block_4c_bn_2 (BatchNormalizati (None, 512, 34, 60) 2048 block_4c_conv_2[0][0]


add_16 (Add) (None, 512, 34, 60) 0 block_4c_bn_2[0][0]
block_4b_relu[0][0]


block_4c_relu (Activation) (None, 512, 34, 60) 0 add_16[0][0]


output_bbox (Conv2D) (None, 8, 34, 60) 4104 block_4c_relu[0][0]


output_cov (Conv2D) (None, 2, 34, 60) 1026 block_4c_relu[0][0]

Total params: 21,319,754
Trainable params: 22,282
Non-trainable params: 21,297,472


Matching predictions to ground truth, class 1/2.: 100%|██████████| 13152/13152 [00:00<00:00, 15996.90it/s]
Matching predictions to ground truth, class 2/2.: 100%|██████████| 461/461 [00:00<00:00, 36889.71it/s]2024-03-20 12:41:53,472 [TAO Toolkit] [INFO] root 2102: Evaluation metrics generated.

Epoch 1/50

Validation cost: 0.001752
Mean average_precision (in %): 0.0265

±-----------±-------------------------+
| class name | average precision (in %) |
±-----------±-------------------------+
| face | 0.0 |
| person | 0.05305001756385717 |
±-----------±-------------------------+

1 Like

Currently, finetuning peoplenet(based on detectnet_v2 network) without forgetting is not supported. But from https://developer.nvidia.com/blog/training-custom-pretrained-models-using-tlt/, with a frozen convolutional layer, the weights do not change in the frozen layer during loss update. This is especially helpful in transfer learning, where you can reuse the features provided by the pretrained weights and reduce training time.
From https://developer.nvidia.com/tao-toolkit-usecases-whitepaper/2-adapting-different-camera-types,
using a pretrained model can save on data labeling costs and training costs by reaching higher accuracy with a smaller dataset.

So, is the finetuning of PeopleNet not available in TAO toolkit API, or not available at all even with TLT and we can just finetune the DetectnetV2 (ResNet34 backbone) architecture with base coco weights?

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

You can finetune peoplenet with TAO. An example can be found in TAO Toolkit Use Cases - 2. Adapting to different | NVIDIA Developer.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.