Questions tagged [loss]
599 questions
0
votes
1 answer
softmax cross entropy return value
What does it mean if this is the return value for tf.losses.softmax_cross_entropy_loss?
Does the fact that is states value:0 mean and shape=() mean that nothing was computed?
haxtar
- 1,962
- 3
- 20
- 44
0
votes
1 answer
Loss function representing the euclidean distance from prediction to nearest groundtruth in images?
Is there a loss function that calculates the euclidean distance between a prediction pixel and the nearest groundtruth pixel? Specifically, this is the location distance, not the intensity distance.
This would be on binary predictions and binary…
user135237
- 389
- 1
- 3
- 12
0
votes
0 answers
the loss doesn't change for the network with a linear layer after softmax
# Inputs and Placeholders
x = tf.placeholder(tf.float32, shape=[None, 30])
y_ = tf.placeholder(tf.float32)
# Inference
W_1 = tf.Variable(tf.zeros([30,50]))
b_1 = tf.Variable(tf.zeros([50]))
layer_1 = tf.sigmoid(tf.add(tf.matmul(x, W_1), b_1))
W_2…
Luca
- 59
- 4
0
votes
1 answer
Modifying Torch criterion
I want to create a custom loss function in Torch which is a modification of ClassNLLCriterion. Concretely, ClassNLLCriterion loss is:
loss(x, class) = -x[class]
I want to modify this to be:
loss(x, class) = -x[class]*K
where K is a function of the…
braindead
- 97
- 2
- 7
0
votes
0 answers
custom loss function different than default
I am trying to understand how to build a custom loss function and the first thing I've tried is to reimplement the binary_crossentropy function in keras.
In my code if I do:
model.compile(Adam(lr=learning_rate), loss=losses.binary_crossentropy,…
Angel Lopez
- 29
- 1
- 4
0
votes
1 answer
Modifying ClassNLLCriterion in Torch
I am new to Torch and I want to create a custom loss function in Torch which is a modification of ClassNLLCriterion. Concretely, ClassNLLCriterion loss is:
loss(x, class) = -x[class]
I want to modify this to be:
loss(x, class) = -x[class] +…
braindead
- 97
- 2
- 7
0
votes
1 answer
Recovering a checkpoint after reaching NaN loss?
I'm training an RNN and sometime overnight the loss function reached NaN. I've been reading that a solution to this is to decrease the learning rate. When attempting to restart training from the (only) checkpoint I have and using a smaller learning…
hate5six
- 11
- 3
0
votes
1 answer
why is my keras custom metric not working?
Why does this code work fine for the loss function but the metrics fail after one iteration with "ValueError: operands could not be broadcast together with shapes (32,) (24,) (32,)"?
If I use "categorical_crossentropy" in quotes then it works. And…
simon
- 2,561
- 16
- 26
0
votes
1 answer
Multiple executions of the evaluation gives different losses in TensorFlow
I'm getting started with TensorFlow.
https://www.tensorflow.org/get_started/
While I was evaluating multiple times seeing how to feed the data,
I found that the loss changes with executions.
eval_input_fn =…
Change-the-world
- 329
- 3
- 9
0
votes
1 answer
Torch, how to get a tensor of loss values during batch optimization
I am training a network with batch optimization over my training set, and I would like to get a loss vector containing the loss of each of my training examples.
More specifically I am using images (of size 3x64x64) in a batch of size 64. Therefore…
fonfonx
- 1,475
- 21
- 30
0
votes
1 answer
caffe output the negative loss value with SoftmaxWithLoss layer?
Below is my last layer in training net:
layer {
name: "loss"
type: "SoftmaxWithLoss"
bottom: "final"
bottom: "label"
top: "loss"
loss_param {
ignore_label: 255
normalization: VALID
}
}
Note I adopt a softmax_loss layer. Since…
huangh12
- 11
- 3
0
votes
1 answer
Accuracy goes down during epoch keras
I've tried to write a neural network but the accuracy doesn't change each epoch. I'm using keras and I can watch the accuracy change as each epoch is evaluated per se and it will start low, go up a bit, then drop back down to the exact same value…
touchmywillis
- 1
- 1
0
votes
1 answer
Amibroker: Daily Loss Limit
I want to implement an afl code to find Daily Loss Limit in intraday trading. I will use the code for backtesting around 200 days.
I have the following code but it is with mistakes.
// identify new day
dn = DateNum();
newDay = dn != Ref( dn,-1);
//…
user4329531
- 11
- 2
0
votes
0 answers
caffe training loss does not converge
I'm getting the problem of non-converged training loss. (batchsize: 16, average loss:10).
I have tried with the following methods
+ Vary the learning rate lr (initial lr = 0.002 cause very high loss, around e+10). Then with lr = e-6, the loss…
Huynh Vu
- 1
0
votes
1 answer
Loss functions in MATLAB
I want to know how to interpret the loss functions results in MATLAB?
On other words, for example if I got 0.3247 as a results of kfoldLoss() function, is this mean that it is 32.47% error or it is a 0.3247%, or how correctly can I define/interpret…
FatmaZahraaAhmed
- 3
- 1
- 5