我正在使用FaceNet repository，我正在尝试使用<script type="text/javascript">
$(document).ready(function () {
$("#Productslist").change(function () {
debugger;
var selectedIndex = $(this).val();
var divH = document.getElementById(selectedIndex);
if (selectedIndex > 0) {
$.ajax(
{
url: '/Inventory/ViewProduct',
type: 'GET',
data: { ProductID: selectedIndex },
contentType: 'application/json; charset=utf-8',
success: function (data) {
///
},
error: function () {
alert("error");
}
});
}
else {
divH.style.visibility = 'hidden'; // Hide
}
});
});
</script>
恢复预训练模型的权重和偏差。
当我尝试恢复模型时，我总是遇到同样的错误：
=filter(A:A,countif(C:C,A:A)=0)似乎找到了正确的文件路径，但是当它尝试加载模型时，会出现一些权重和偏差的问题。
我使用的代码如下：
tf.train.Saver如您所见，张量通过NotFoundError (see above for traceback): Key InceptionResnetV1/Block8/Branch_0/Conv2d_1x1/biases not found in checkpoint
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
end_point打印，import tensorflow as tf
import cv2
import numpy as np
from model.inception_resnet_v1 import inception_resnet_v1 as resnet_bottleneck
# from model.inception_resnet_v1 import inference as resnet_logits
from tensorflow.python.tools import inspect_checkpoint as chkp
def main():
with tf.Graph().as_default():
# Load an image into a numpy array, and expand the dimension to the correct one, based on the
# inception_resNet model.
img = cv2.imread('/home/uc3m1/Documentos/BQ/modelos/facenet-master/src/datasets/dataset_prueba/Tiger_Woods'
'/Tiger_Woods_0002.png')
img = np.expand_dims(img, axis=0)
img = img.astype(np.float32)
# Based on FaceNet, extracting the logits and end_points
logits, end_points = resnet_bottleneck(img, 1)
# Initializers
init_global = tf.initializers.global_variables()
init_local = tf.initializers.local_variables()
# Checking if the checkpoints tensors
chkp.print_tensors_in_checkpoint_file('/home/uc3m1/PycharmProjects/siameseFaceNet/weights/model-20180408'
'-102900.ckpt-90', tensor_name='', all_tensors=False,
all_tensor_names=True)
# Create a saver
saver = tf.train.Saver(tf.global_variables())
with tf.Session() as sess:
sess.run(init_global)
sess.run(init_local)
# Restore the pretrained model from FaceNet
saver.restore(sess, '/home/uc3m1/PycharmProjects/siameseFaceNet/weights/model-20180408-102900.ckpt-90')
print(sess.run(end_points["PreLogitsFlatten"]).shape)
if __name__ == '__main__':
main()
确实未在模型中定义。我真的不知道如何添加这种偏见，或者如何正确加载它。
有谁知道原因？