m4nuc / aifiddle-issues Goto Github PK

Hello,

Using AI fiddle to create fully connected nets and single layer conv nets converges OK, however none of my attempts to use 2-3 convolution layers seem to get much above 50% accuracy.

To get a reference point I re implemented in AI Fiddle the simple conv net here: https://github.com/keras-team/keras/blob/master/examples/mnist_cnn.py and also ran it in a colaboratory notebook. The notebook was at 92% accuracy after the first epoch. In AIFiddle I get to around 50% with 1 epoch and a batch size of 32. Moving to a batch size of 128 (as used by the code sample) sees this crashing down to 10%.

Do you have any idea what could cause this?

https://beta.aifiddle.io/ doesn't respond

Right now, if the tab is closed while a training procedure is going on, the application doesn't provide any alert. I think that the best behavior would be to alert the user about the ongoing training and ask for a confirmation (like Facebook does when you have an unsent message/comment). The objective of that change is to avoid an accidental closing.

You've made the world a better place.

Hi,
I created an account and probably mistyped my password. I can't find a way to reset it.

It would be nice to have an export-feature for each sort of layer that can generate sample code for the main NN libraries, e.g.,

pytorch

    model = torch.nn.Sequential()
    model.add_module("linear_1", torch.nn.Linear(input_dim, 128))
    model.add_module("relu_1", torch.nn.ReLU())
    model.add_module("linear_2", torch.nn.Linear(128, 64))
    model.add_module("relu_1", torch.nn.ReLU())
    model.add_module("linear_3", torch.nn.Linear(64, output_dim))

keras

???

tf

    # Hidden 1
    with tf.name_scope('hidden1'):
        weights = tf.Variable(
            tf.truncated_normal([IMAGE_PIXELS, hidden1_units],
                                stddev=1.0 / math.sqrt(float(IMAGE_PIXELS))), name='weights')
        biases = tf.Variable(tf.zeros([hidden1_units]), name='biases')
        hidden1 = tf.nn.relu(tf.matmul(images, weights) + biases)
    # Hidden 2
    with tf.name_scope('hidden2'):
        weights = tf.Variable(
            tf.truncated_normal([hidden1_units, hidden2_units],
                                stddev=1.0 / math.sqrt(float(hidden1_units))), name='weights')
        biases = tf.Variable(tf.zeros([hidden2_units]),name='biases')
        hidden2 = tf.nn.relu(tf.matmul(hidden1, weights) + biases)
    # Linear
    with tf.name_scope('softmax_linear'):
        weights = tf.Variable(
            tf.truncated_normal([hidden2_units, NUM_CLASSES],
                                stddev=1.0 / math.sqrt(float(hidden2_units))), name='weights')
        biases = tf.Variable(tf.zeros([NUM_CLASSES]),name='biases')
        logits = tf.matmul(hidden2, weights) + biases
    #
    return logits

mxnet

data = mx.symbol.Variable('data')
fc1  = mx.symbol.FullyConnected(data = data, name='fc1', num_hidden=128)
act1 = mx.symbol.Activation(data = fc1, name='relu1', act_type="relu")
fc2  = mx.symbol.FullyConnected(data = act1, name = 'fc2', num_hidden = 64)
act2 = mx.symbol.Activation(data = fc2, name='relu2', act_type="relu")
fc3  = mx.symbol.FullyConnected(data = act2, name='fc3', num_hidden=10)
mlp  = mx.symbol.SoftmaxOutput(data = fc3, name = 'softmax')

I just realized that, while training, if I switch to another tab to do something else, the training stops. When I switch back, it continues normally. The only thing that continues if the tab loses focus is the time counter.

I think the expected behavior would be to continue the training even if the tab doesn't have the focus.

I'm using:
Windows 10
Firefox 64.0.2 (64-bit)
Dozens of tabs opened.

Would be nice to be able to restart training with reinitialised weights, quickly and easily. Not sure how best to do this right now.

Hello,

On both Chrome and Firefox Fashion MNIST display a "Can not load dataset" error at the bottom right of the screen.

Thank you

I'm a little slow with reading/understanding the tooltips. It would be helpful if the tooltip that pops up in the bottom right sticks around until the user is done. It would help with debugging.

Can not update Conv2D_irzspz20l1 –– casual is not a valid PaddingMode. Valid values are valid,same,causal or null/undefined.

It seems like there is a typo in the padding options

When chosing the RMSprop optimizer, the webpage become totally blank.
Others optimizer seem fine.