I have a CNN code that was written using tensorflow library:
x_img = tf.placeholder(tf.float32)
y_label = tf.placeholder(tf.float32)
def convnet_3d(x_img, W):
conv_3d_layer = tf.nn.conv3d(x_img, W, strides=[1,1,1,1,1], padding='VALID')
return conv_3d_layer
def maxpool_3d(x_img):
maxpool_3d_layer = tf.nn.max_pool3d(x_img, ksize=[1,2,2,2,1], strides=[1,2,2,2,1], padding='VALID')
return maxpool_3d_layer
def convolutional_neural_network(x_img):
weights = {'W_conv1_layer':tf.Variable(tf.random_normal([3,3,3,1,32])),
'W_conv2_layer':tf.Variable(tf.random_normal([3,3,3,32,64])),
'W_fc_layer':tf.Variable(tf.random_normal([409600,1024])),
'W_out_layer':tf.Variable(tf.random_normal([1024, num_classes]))}
biases = {'b_conv1_layer':tf.Variable(tf.random_normal([32])),
'b_conv2_layer':tf.Variable(tf.random_normal([64])),
'b_fc_layer':tf.Variable(tf.random_normal([1024])),
'b_out_layer':tf.Variable(tf.random_normal([num_classes]))}
x_img = tf.reshape(x_img, shape=[-1, img_x, img_y, img_z, 1])
conv1_layer = tf.nn.relu(convnet_3d(x_img, weights['W_conv1_layer']) + biases['b_conv1_layer'])
conv1_layer = maxpool_3d(conv1_layer)
conv2_layer = tf.nn.relu(convnet_3d(conv1_layer, weights['W_conv2_layer']) + biases['b_conv2_layer'])
conv2_layer = maxpool_3d(conv2_layer)
fc_layer = tf.reshape(conv2_layer,[-1, 409600])
fc_layer = tf.nn.relu(tf.matmul(fc_layer, weights['W_fc_layer'])+biases['b_fc_layer'])
fc_layer = tf.nn.dropout(fc_layer, keep_rate)
output_layer = tf.matmul(fc_layer, weights['W_out_layer'])+biases['b_out_layer']
return output_layer
my input image x_img is 25x25x25(3d image), I have some questions about the code:
1- is [3,3,3,1,32] in 'W_conv1_layer' means [width x height x depth x channel x number of filters]?
2- in 'W_conv2_layer' weights are [3,3,3,32,64], why the output is 64? I know that 3x3x3 is filter size and 32 is input come from first layer.
3- in 'W_fc_layer' weights are [409600,1024], 1024 is number of nodes in FC layer, but where this magic number '409600' come from?
4- before the image get into the conv layers why we need to reshape the image
x_img = tf.reshape(x_img, shape=[-1, img_x, img_y, img_z, 1])
All the answers can be found in the official doc of conv3d.
The weights should be [filter_depth, filter_height, filter_width, in_channels, out_channels]
The numbers 32 and 64 are chosen because it works simply they are just hyperparameters
409600 comes from reshaping the output of maxpool3d (it is probably a mistake the real size should be 4096 see comments)
Because tensorflow expects certain layouts for its input
Your should try implementing a simple convnet on images before moving to more complicated stuff.
Related
I trained a neural network on MNIST using PyTorch:
class MnistCNN(nn.Module):
def __init__(self):
super().__init__()
self.conv1 = nn.Conv2d( 1, 16, 3, stride = 1, padding = 2)
self.pool1 = nn.MaxPool2d(kernel_size = 2)
self.conv2 = nn.Conv2d(16, 32, 3, stride = 1, padding = 2)
self.pool2 = nn.MaxPool2d(kernel_size = 2)
self.dropout = nn.Dropout(0.5)
self.lin = nn.Linear(32 * 8 * 8, 10)
def forward(self, x):
# conv block
x = F.relu(self.conv1(x))
x = self.pool1(x)
# conv block
x = F.relu(self.conv2(x))
x = self.pool2(x)
# dense block
x = x.view(x.size(0), -1)
x = self.dropout(x)
return self.lin(x)
I would like to implement vanilla Gradient Visualization (see reference below) on my model.
Simonyan, K., Vedaldi, A., Zisserman, A.
Deep inside convolutional networks: Visualising image classification models and saliency maps.
arXiv preprint arXiv:1312.6034 (2013)
Question: How can I implement this method in PyTorch?
If I understand correctly, vanilla gradient visualization consists in computing the partial derivatives of the loss of my model w.r.t all the pixels in my input image. So to make it short, I need to tweek my self.conv1 layer so that it computes the gradient over its input pixels instead of the gradient over its weights.
Please correct me if I'm wrong.
You do not need to change anything about your conv layer. Each layer computes gradients both w.r.t. parameters (for updates) and w.r.t. inputs (for "downstream" gradients by the chain rule). Therefore, all you need is to set your input image's x gradient property to true:
x, y = ... # get one image from MNIST
x.requires_grad_(True) # indicate to pytorch that you would like to look at these gradients
pred = model(x)
loss = criterion(pred, y)
loss.backward() # propagate gradients
x.grad # <- here you should have the gradients of the loss w.r.t pixels
I started working with Pytorch recently so my understanding of it isn't quite strong. I previously had a 1 layer CNN but wanted to extend it to 2 layers, but the input and output channels have been throwing errors I can seem to decipher. Why does it expect 192 channels? Can someone give me a pointer to help me understand this better? I have seen several related problems on here, but I don't understand those solutions either.
import numpy as np
import pandas as pd
import torch
import torch.nn as nn
from transformers import BertConfig, BertModel, BertTokenizer
import math
from transformers import AdamW, get_linear_schedule_with_warmup
def pad_sents(sents, pad_token): # Pad list of sentences according to the longest sentence in the batch.
sents_padded = []
max_len = max(len(s) for s in sents)
for s in sents:
padded = [pad_token] * max_len
padded[:len(s)] = s
sents_padded.append(padded)
return sents_padded
def sents_to_tensor(tokenizer, sents, device):
tokens_list = [tokenizer.tokenize(str(sent)) for sent in sents]
sents_lengths = [len(tokens) for tokens in tokens_list]
tokens_list_padded = pad_sents(tokens_list, '[PAD]')
sents_lengths = torch.tensor(sents_lengths, device=device)
masks = []
for tokens in tokens_list_padded:
mask = [0 if token == '[PAD]' else 1 for token in tokens]
masks.append(mask)
masks_tensor = torch.tensor(masks, dtype=torch.long, device=device)
tokens_id_list = [tokenizer.convert_tokens_to_ids(tokens) for tokens in tokens_list_padded]
sents_tensor = torch.tensor(tokens_id_list, dtype=torch.long, device=device)
return sents_tensor, masks_tensor, sents_lengths
class ConvModel(nn.Module):
def __init__(self, device, dropout_rate, n_class, out_channel=16):
super(ConvModel, self).__init__()
self.bert_config = BertConfig.from_pretrained('bert-base-uncased', output_hidden_states=True)
self.dropout_rate = dropout_rate
self.n_class = n_class
self.out_channel = out_channel
self.bert = BertModel.from_pretrained('bert-base-uncased', config=self.bert_config)
self.out_channels = self.bert.config.num_hidden_layers * self.out_channel
self.tokenizer = BertTokenizer.from_pretrained('bert-base-uncased', config=self.bert_config)
self.conv = nn.Conv2d(in_channels=self.bert.config.num_hidden_layers,
out_channels=self.out_channels,
kernel_size=(3, self.bert.config.hidden_size),
groups=self.bert.config.num_hidden_layers)
self.conv1 = nn.Conv2d(in_channels=self.out_channels,
out_channels=48,
kernel_size=(3, self.bert.config.hidden_size),
groups=self.bert.config.num_hidden_layers)
self.hidden_to_softmax = nn.Linear(self.out_channels, self.n_class, bias=True)
self.dropout = nn.Dropout(p=self.dropout_rate)
self.device = device
def forward(self, sents):
sents_tensor, masks_tensor, sents_lengths = sents_to_tensor(self.tokenizer, sents, self.device)
encoded_layers = self.bert(input_ids=sents_tensor, attention_mask=masks_tensor)
hidden_encoded_layer = encoded_layers[2]
hidden_encoded_layer = hidden_encoded_layer[0]
hidden_encoded_layer = torch.unsqueeze(hidden_encoded_layer, dim=1)
hidden_encoded_layer = hidden_encoded_layer.repeat(1, 12, 1, 1)
conv_out = self.conv(hidden_encoded_layer) # (batch_size, channel_out, some_length, 1)
conv_out = self.conv1(conv_out)
conv_out = torch.squeeze(conv_out, dim=3) # (batch_size, channel_out, some_length)
conv_out, _ = torch.max(conv_out, dim=2) # (batch_size, channel_out)
pre_softmax = self.hidden_to_softmax(conv_out)
return pre_softmax
def batch_iter(data, batch_size, shuffle=False, bert=None):
batch_num = math.ceil(data.shape[0] / batch_size)
index_array = list(range(data.shape[0]))
if shuffle:
data = data.sample(frac=1)
for i in range(batch_num):
indices = index_array[i * batch_size: (i + 1) * batch_size]
examples = data.iloc[indices]
sents = list(examples.train_BERT_tweet)
targets = list(examples.train_label.values)
yield sents, targets # list[list[str]] if not bert else list[str], list[int]
def train():
label_name = ['Yes', 'Maybe', 'No']
device = torch.device("cpu")
df_train = pd.read_csv('trainn.csv') # , index_col=0)
train_label = dict(df_train.train_label.value_counts())
label_max = float(max(train_label.values()))
train_label_weight = torch.tensor([label_max / train_label[i] for i in range(len(train_label))], device=device)
model = ConvModel(device=device, dropout_rate=0.2, n_class=len(label_name))
optimizer = AdamW(model.parameters(), lr=1e-3, correct_bias=False)
scheduler = get_linear_schedule_with_warmup(optimizer, num_warmup_steps=100, num_training_steps=1000) # changed the last 2 arguments to old ones
model = model.to(device)
model.train()
cn_loss = torch.nn.CrossEntropyLoss(weight=train_label_weight, reduction='mean')
train_batch_size = 16
for epoch in range(1):
for sents, targets in batch_iter(df_train, batch_size=train_batch_size, shuffle=True): # for each epoch
optimizer.zero_grad()
pre_softmax = model(sents)
loss = cn_loss(pre_softmax, torch.tensor(targets, dtype=torch.long, device=device))
loss.backward()
optimizer.step()
scheduler.step()
TrainingModel = train()
Here's a snippet of data https://github.com/Kosisochi/DataSnippet
It seems that the original version of the code you had in this question behaved differently. The final version of the code you have here gives me a different error from what you posted, more specifically - this:
RuntimeError: Calculated padded input size per channel: (20 x 1). Kernel size: (3 x 768). Kernel size can't be greater than actual input size
I apologize if I misunderstood the situation, but it seems to me that your understanding of what exactly nn.Conv2d layer does is not 100% clear and that is the main source of your struggle. I interpret the part "detailed explanation on 2 layer CNN in Pytorch" you requested as an ask to explain in detail on how that layer works and I hope that after this is done there will be no problem applying it 1 time, 2 times or more.
You can find all the documentation about the layer here, but let me give you a recap which hopefully will help to understand more the errors you're getting.
First of all nn.Conv2d inputs are 4-d tensors of the shape (BatchSize, ChannelsIn, Height, Width) and outputs are 4-d tensors of the shape (BatchSize, ChannelsOut, HeightOut, WidthOut). The simplest way to think about nn.Conv2d is of something applied to 2d images with pixel grid of size Height x Width and having ChannelsIn different colors or features per pixel. Even if your inputs have nothing to do with actual images the behavior of the layer is still the same. Simplest situation is when the nn.Conv2d is not using padding (as in your code). In that case the kernel_size=(kernel_height, kernel_width) argument specifies the rectangle which you can imagine sweeping through Height x Width rectangle of your inputs and producing one pixel for each valid position. Without padding the coordinate of the rectangle's point can be any pair of indicies (x, y) with x between 0 and Height - kernel_height and y between 0 and Width - kernel_width. Thus the output will look like a 2d image of size (Height - kernel_height + 1) x (Width - kernel_width + 1) and will have as many output channels as specified to nn.Conv2d constructor, so the output tensor will be of shape (BatchSize, ChannelsOut, Height - kernel_height + 1, Width - kernel_width + 1).
The parameter groups is not affecting how shapes are changed by the layer - it is only controlling which input channels are used as inputs for the output channels (groups=1 means that every input channel is used as input for every output channel, otherwise input and output channels are divided into corresponding number of groups and only input channels from group i are used as inputs for the output channels from group i).
Now in your current version of the code you have BatchSize = 16 and the output of pre-trained model is (BatchSize, DynamicSize, 768) with DynamicSize depending on the input, e.g. 22. You then introduce additional dimension as axis 1 with unsqueeze and repeat the values along that dimension transforming the tensor of shape (16, 22, 768) into (16, 12, 22, 768). Effectively you are using the output of the pre-trained model as 12-channel (with each channel having same values as others) 2-d images here of size (22, 768), where 22 is not fixed (depends on the batch). Then you apply a nn.Conv2d with kernel size (3, 768) - which means that there is no "wiggle room" for width and output 2-d images will be of size (20, 1) and since your layer has 192 channels final size of the output of first convolution layer has shape (16, 192, 20, 1). Then you try to apply second layer of convolution on top of that with kernel size (3, 768) again, but since your 2-d "image" is now just (20 x 1) there is no valid position to fit (3, 768) kernel rectangle inside a rectangle (20 x 1) which leads to the error message Kernel size can't be greater than actual input size.
Hope this explanation helps. Now to the choices you have to avoid the issue:
(a) is to add padding in such a way that the size of the output is not changing comparing to input (I won't go into details here,
because I don't think this is what you need)
(b) Use smaller kernel on both first and/or second convolutions (e.g. if you don't change first convolution the only valid width for
the second kernel would be 1).
(c) Looking at what you're trying to do my guess is that you actually don't want to use 2d convolution, you want 1d convolution (on the sequence) with every position described by 768 values. When you're using one convolution layer with 768 width kernel (and same 768 width input) you're effectively doing exactly same thing as 1d convolution with 768 input channels, but then if you try to apply second one you have a problem. You can specify kernel width as 1 for the next layer(s) and that will work for you, but a more correct way would be to transpose pre-trained model's output tensor by switching the last dimensions - getting shape (16, 768, DynamicSize) from (16, DynamicSize, 768) and then apply nn.Conv1d layer with 768 input channels and arbitrary ChannelsOut as output channels and 1d kernel_size=3 (meaning you look at 3 consecutive elements of the sequence for convolution). If you do that than without padding input shape of (16, 768, DynamicSize) will become (16, ChannelsOut, DynamicSize-2), and after you apply second Conv1d with e.g. the same settings as first one you'll get a tensor of shape (16, ChannelsOut, DynamicSize-4), etc. (each time the 1d length will shrink by kernel_size-1). You can always change number of channels/kernel_size for each subsequent convolution layer too.
I'm working on a project where I have to predict the future states of a 1D vector with y entries. I'm trying to do this using an ANN setup with LSTM units in combination with a convolution layer. The method I'm using is based on the method they used in a (pre-release paper). The suggested setup is as follows:
In the picture c is the 1D vector with y entries. The ANN gets the n previous states as an input and produces o next states as an output.
Currently, my ANN setup looks like this:
inputLayer = Input(shape = (n, y))
encoder = LSTM(200)(inputLayer)
x = RepeatVector(1)(encoder)
decoder = LSTM(200, return_sequences=True)(x)
x = Conv1D(y, 4, activation = 'linear', padding = 'same')(decoder)
model = Model(inputLayer, x)
Here n is the length of the input sequences and y is the length of the state array. As can be seen I'm repeating the d vector only 1 time, as I'm trying to predict only 1 time step in the future. Is this the way to setup the above mentioned network?
Furthermore, I have a numpy array (data) with a shape of (Sequences, Time Steps, State Variables) to train with. I was trying to divide this in randomly selected batches with a generator like this:
def BatchGenerator(batch_size, n, y, data):
# Infinite loop.
while True:
# Allocate a new array for the batch of input-signals.
x_shape = (batch_size, n, y)
x_batch = np.zeros(shape=x_shape, dtype=np.float16)
# Allocate a new array for the batch of output-signals.
y_shape = (batch_size, 1, y)
y_batch = np.zeros(shape=y_shape, dtype=np.float16)
# Fill the batch with random sequences of data.
for i in range(batch_size):
# Select a random sequence
seq_idx = np.random.randint(data.shape[0])
# Get a random start-index.
# This points somewhere into the training-data.
start_idx = np.random.randint(data.shape[1] - n)
# Copy the sequences of data starting at this
# Each batch inside x_batch has a shape of [n, y]
x_batch[i,:,:] = data[seq_idx, start_idx:start_idx+n, :]
# Each batch inside y_batch has a shape of [1, y] (as we predict only 1 time step in advance)
y_batch[i,:,:] = data[seq_idx, start_idx+n, :]
yield (x_batch, y_batch)
The problem is that it gives an error if I'm using a batch_size of more than 1. Could anyone help me to set this data up in a way that it can be used optimally to train my neural network?
The model is now trained using:
generator = BatchGenerator(batch_size, n, y, data)
model.fit_generator(generator = generator, steps_per_epoch = steps_per_epoch, epochs = epochs)
Thanks in advance!
I'm working on a RNN architecture which does speech enhancement. The dimensions of the input is [XX, X, 1024] where XX is the batch size and X is the variable sequence length.
The input to the network is positive valued data and the output is masked binary data(IBM) which is later used to construct enhanced signal.
For instance, if the input to network is [10, 65, 1024] the output will be [10,65,1024] tensor with binary values. I'm using Tensorflow with mean squared error as loss function. But I'm not sure which activation function to use here(which keeps the outputs either zero or one), Following is the code I've come up with so far
tf.reset_default_graph()
num_units = 10 #
num_layers = 3 #
dropout = tf.placeholder(tf.float32)
cells = []
for _ in range(num_layers):
cell = tf.contrib.rnn.LSTMCell(num_units)
cell = tf.contrib.rnn.DropoutWrapper(cell, output_keep_prob = dropout)
cells.append(cell)
cell = tf.contrib.rnn.MultiRNNCell(cells)
X = tf.placeholder(tf.float32, [None, None, 1024])
Y = tf.placeholder(tf.float32, [None, None, 1024])
output, state = tf.nn.dynamic_rnn(cell, X, dtype=tf.float32)
out_size = Y.get_shape()[2].value
logit = tf.contrib.layers.fully_connected(output, out_size)
prediction = (logit)
flat_Y = tf.reshape(Y, [-1] + Y.shape.as_list()[2:])
flat_logit = tf.reshape(logit, [-1] + logit.shape.as_list()[2:])
loss_op = tf.losses.mean_squared_error(labels=flat_Y, predictions=flat_logit)
#adam optimizier as the optimization function
optimizer = tf.train.AdamOptimizer(learning_rate=0.001) #
train_op = optimizer.minimize(loss_op)
#extract the correct predictions and compute the accuracy
correct_pred = tf.equal(tf.argmax(prediction, 1), tf.argmax(Y, 1))
accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
Also my reconstruction isn't good. Can someone suggest on improving the model?
If you want your outputs to be either 0 or 1, to me it seems a good idea to turn this into a classification problem. To this end, I would use a sigmoidal activation and cross entropy:
...
prediction = tf.nn.sigmoid(logit)
loss_op = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(labels=Y, logits=logit))
...
In addition, from my point of view the hidden dimensionality (10) of your stacked RNNs seems quite small for such a big input dimensionality (1024). However this is just a guess, and it is something that needs to be tuned.
I'll preface this by saying this is my first posted question on SO. I've just recently started working with Tensorflow, and have been attempting to apply a convolutional-neural network model approach for classification of .csv records in a file representing images from scans of microarray data. (FYI: Microarrays are a grid of spotted DNA on a glass slide, representing specific DNA target sequences for determining the presence of those DNA targets in a sample. The individual pixels represent fluorescence intensity value from 0-1). The file has ~200,000 records in total. Each record (image) has 10816 pixels that represent DNA sequences from known viruses, and one index label which identifies the virus species. The pixels create a pattern which is unique to each of the different viruses. There are 2165 different viruses in total represented within the 200,000 records. I have trained the network on images of labeled microarray datasets, but when I try to pass a new dataset through to classify it/them as one of the 2165 different viruses and determine predicted values and probabilities, I don't seem to be having much luck. This is the code that I am currently using for this:
import tensorflow as tf
import numpy as np
import csv
def extract_data(filename):
print("extracting data...")
NUM_LABELS = 2165
NUM_FEATURES = 10816
labels = []
fvecs = []
rowCount = 0
#iterate over the rows, split the label from the features
#convert the labels to integers and features to floats
for line in open(filename):
rowCount = rowCount + 1
row = line.split(',')
labels.append(row[3])#(int(row[7])) #<<<IT ALWAYS PREDICTS THIS VALUE!
for x in row [4:10820]:
fvecs.append(float(x))
#convert the array of float arrasy into a numpy float matrix
fvecs_np = np.matrix(fvecs).astype(np.float32)
#convert the array of int lables inta a numpy array
labels_np = np.array(labels).astype(dtype=np.uint8)
#convert the int numpy array into a one-hot matrix
labels_onehot = (np.arange(NUM_LABELS) == labels_np[:, None]).astype(np.float32)
print("arrays converted")
return fvecs_np, labels_onehot
def TestModels():
fvecs_np, labels_onehot = extract_data("MicroarrayTestData.csv")
print('RESTORING NN MODEL')
weights = {}
biases = {}
sess=tf.Session()
init = tf.global_variables_initializer()
#Load meta graph and restore weights
ModelID = "MicroarrayCNN_Data-1000.meta"
print("RESTORING:::", ModelID)
saver = tf.train.import_meta_graph(ModelID)
saver.restore(sess,tf.train.latest_checkpoint('./'))
graph = tf.get_default_graph()
x = graph.get_tensor_by_name("x:0")
y = graph.get_tensor_by_name("y:0")
keep_prob = tf.placeholder(tf.float32)
y_ = tf.placeholder("float", shape=[None, 2165])
wc1 = graph.get_tensor_by_name("wc1:0")
wc2 = graph.get_tensor_by_name("wc2:0")
wd1 = graph.get_tensor_by_name("wd1:0")
Wout = graph.get_tensor_by_name("Wout:0")
bc1 = graph.get_tensor_by_name("bc1:0")
bc2 = graph.get_tensor_by_name("bc2:0")
bd1 = graph.get_tensor_by_name("bd1:0")
Bout = graph.get_tensor_by_name("Bout:0")
weights = {wc1, wc2, wd1, Wout}
biases = {bc1, bc2, bd1, Bout}
print("NEXTArgmax")
prediction=tf.argmax(y,1)
probabilities = y
predY = prediction.eval(feed_dict={x: fvecs_np, y: labels_onehot}, session=sess)
probY = probabilities.eval(feed_dict={x: fvecs_np, y: labels_onehot}, session=sess)
accuracy = tf.reduce_mean(tf.cast(prediction, "float"))
print(sess.run(accuracy, feed_dict={x: fvecs_np, y: labels_onehot}))
print("%%%%%%%%%%%%%%%%%%%%%%%%%%")
print("Predicted::: ", predY, accuracy)
print("%%%%%%%%%%%%%%%%%%%%%%%%%%")
feed_dictTEST = {y: labels_onehot}
probabilities=probY
print("probabilities", probabilities.eval(feed_dict={x: fvecs_np}, session=sess))
########## Run Analysis ###########
TestModels()
So, when I run this code I get the correct prediction for the test set, although I am not sure I believe it, because it appears that whatever value I append in line 14 (see below) is the output it predicts:
labels.append(row[3])#<<<IT ALWAYS PREDICTS THIS VALUE!
I don't understand this, and it makes me suspicious that I've set up the CNN incorrectly, as I would have expected it to ignore my input label and determine a bast match from the trained network based on the trained patterns. The only thing I can figure is that when I pass the value through for the prediction; it is instead training the model on this data as well, and then predicting itself. Is this a correct assumption, or am I misinterpreting how Tensorflow works?
The other issue is that when I try to use code that (based on other tutorials) which is supposed to output the probabilities of all of the 2165 possible outputs, I get the error:
InvalidArgumentError (see above for traceback): Shape [-1,2165] has negative dimensions
[[Node: y = Placeholder[dtype=DT_FLOAT, shape=[?,2165], _device="/job:localhost/replica:0/task:0/cpu:0"]()]]
To me, it looks like it is the correct layer based on the 2165 value in the Tensor shape, but I don't understand the -1 value. So, to wrap up the summary, my questions are:
Based on the fact that I get the value that I have in the label of the input data, is this the correct method to make a classification using this model?
Am I missing a layer or have I configured the model incorrectly in order to extract the probabilities of all of the possible output classes, or am I using the wrong code to extract the information? I try to print out the accuracy to see if that would work, but instead it outputs the description of a tensor, so clearly that is incorrect as well.
(ADDITIONAL INFORMATION)
As requested, I'm also including the original code that was used to train the model, which is now below. You can see I do sort of a piece meal training of a limited number of related records at a time by their taxonomic relationships as I iterate through the file. This is mostly because the Mac that I'm training on (Mac Pro w/ 64GB ram) tends to give me the "Killed -9" error due to overuse of resources if I don't do it this way. There may be a better way to do it, but this seems to work.
Original Author: Aymeric Damien
Project: https://github.com/aymericdamien/TensorFlow-Examples/
from __future__ import print_function
import tensorflow as tf
import numpy as np
import csv
import random
# Parameters
num_epochs = 2
train_size = 1609
learning_rate = 0.001 #(larger >speed, lower >accuracy)
training_iters = 5000 # How much do you want to train (more = better trained)
batch_size = 32 #How many samples to train on, size of the training batch
display_step = 10 # How often to diplay what is going on during training
# Network Parameters
n_input = 10816 # MNIST data input (img shape: 28*28)...in my case 104x104 = 10816(rough array size)
n_classes = 2165 #3280 #2307 #787# Switched to 100 taxa/training set, dynamic was too wonky.
dropout = 0.75 # Dropout, probability to keep units. Jeffery Hinton's group developed it, that prevents overfitting to find new paths. More generalized model.
# Functions
def extract_data(filename):
print("extracting data...")
# arrays to hold the labels and feature vectors.
NUM_LABELS = 2165
NUM_FEATURES = 10826
taxCount = 0
taxCurrent = 0
labels = []
fvecs = []
rowCount = 0
#iterate over the rows, split the label from the features
#convert the labels to integers and features to floats
print("entering CNN loop")
for line in open(filename):
rowCount = rowCount + 1
row = line.split(',')
taxCurrent = row[3]
print("profile:", row[0:12])
labels.append(int(row[3]))
fvecs.append([float(x) for x in row [4:10820]])
#convert the array of float arrasy into a numpy float matrix
fvecs_np = np.matrix(fvecs).astype(np.float32)
#convert the array of int lables inta a numpy array
labels_np = np.array(labels).astype(dtype=np.uint8)
#convert the int numpy array into a one-hot matrix
labels_onehot = (np.arange(NUM_LABELS) == labels_np[:, None]).astype(np.float32)
print("arrays converted")
return fvecs_np, labels_onehot
# Create some wrappers for simplicity
def conv2d(x, W, b, strides=1): #Layer 1 : Convolutional layer
# Conv2D wrapper, with bias and relu activation
print("conv2d")
x = tf.nn.conv2d(x, W, strides=[1, strides, strides, 1], padding='SAME') # Strides are the tensors...list of integers. Tensors=data
x = tf.nn.bias_add(x, b) #bias is the tuning knob
return tf.nn.relu(x) #rectified linear unit (activation function)
def maxpool2d(x, k=2): #Layer 2 : Takes samples from the image. (This is a 4D tensor)
print("maxpool2d")
# MaxPool2D wrapper
return tf.nn.max_pool(x, ksize=[1, k, k, 1], strides=[1, k, k, 1],
padding='SAME')
# Create model
def conv_net(x, weights, biases, dropout):
print("conv_net setup")
# Reshape input picture
x = tf.reshape(x, shape=[-1, 104, 104, 1]) #-->52x52 , -->26x26x64
# Convolution Layer
conv1 = conv2d(x, weights['wc1'], biases['bc1']) #defined above already
# Max Pooling (down-sampling)
conv1 = maxpool2d(conv1, k=2)
print(conv1.get_shape)
# Convolution Layer
conv2 = conv2d(conv1, weights['wc2'], biases['bc2']) #wc2 and bc2 are just placeholders...could actually skip this layer...maybe
# Max Pooling (down-sampling)
conv2 = maxpool2d(conv2, k=2)
print(conv2.get_shape)
# Fully connected layer
# Reshape conv2 output to fit fully connected layer input
fc1 = tf.reshape(conv2, [-1, weights['wd1'].get_shape().as_list()[0]])
fc1 = tf.add(tf.matmul(fc1, weights['wd1']), biases['bd1'])
fc1 = tf.nn.relu(fc1) #activation function for the NN
# Apply Dropout
fc1 = tf.nn.dropout(fc1, dropout)
# Output, class prediction
out = tf.add(tf.matmul(fc1, weights['Wout']), biases['Bout'])
return out
def Train_Network(Txid_IN, Sess_File_Name):
import tensorflow as tf
tf.reset_default_graph()
x,y = 0,0
weights = {}
biases = {}
# tf Graph input
print("setting placeholders")
x = tf.placeholder(tf.float32, [None, n_input], name="x") #Gateway for data (images)
y = tf.placeholder(tf.float32, [None, n_classes], name="y") # Gateway for data (labels)
keep_prob = tf.placeholder(tf.float32) #dropout # Gateway for dropout(keep probability)
# Store layers weight & bias
#CREATE weights
weights = {
# 5x5 conv, 1 input, 32 outputs
'wc1': tf.Variable(tf.random_normal([5, 5, 1, 32]),name="wc1"), #
# 5x5 conv, 32 inputs, 64 outputs
'wc2': tf.Variable(tf.random_normal([5, 5, 32, 64]),name="wc2"),
# fully connected, 7*7*64 inputs, 1024 outputs
'wd1': tf.Variable(tf.random_normal([26*26*64, 1024]),name="wd1"),
# 1024 inputs, 10 outputs (class prediction)
'Wout': tf.Variable(tf.random_normal([1024, n_classes]),name="Wout")
}
biases = {
'bc1': tf.Variable(tf.random_normal([32]), name="bc1"),
'bc2': tf.Variable(tf.random_normal([64]), name="bc2"),
'bd1': tf.Variable(tf.random_normal([1024]), name="bd1"),
'Bout': tf.Variable(tf.random_normal([n_classes]), name="Bout")
}
# Construct model
print("constructing model")
pred = conv_net(x, weights, biases, keep_prob)
print(pred)
# Define loss(cost) and optimizer
#cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(pred, y)) Deprecated version of the statement
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = pred, labels=y)) #added reduce_mean 6/27
optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)
# Evaluate model
correct_pred = tf.equal(tf.argmax(pred, 1), tf.argmax(y, 1))
accuracy = tf.reduce_mean(tf.cast(correct_pred, tf.float32))
print("%%%%%%%%%%%%%%%%%%%%")
print ("%% ", correct_pred)
print ("%% ", accuracy)
print("%%%%%%%%%%%%%%%%%%%%")
# Initializing the variables
#init = tf.initialize_all_variables()
init = tf.global_variables_initializer()
saver = tf.train.Saver()
fvecs_np, labels_onehot = extract_data("MicroarrayDataOUT.csv") #CHAGE TO PICORNAVIRUS!!!!!AHHHHHH!!!
print("starting session")
# Launch the graph
FitStep = 0
with tf.Session() as sess: #graph is encapsulated by its session
sess.run(init)
step = 1
# Keep training until reach max iterations (training_iters)
while step * batch_size < training_iters:
if FitStep >= 5:
break
else:
#iterate and train
print(step)
print(fvecs_np, labels_onehot)
for step in range(num_epochs * train_size // batch_size):
sess.run(optimizer, feed_dict={x: fvecs_np, y: labels_onehot, keep_prob:dropout}) #no dropout???...added Keep_prob:dropout
if FitStep >= 5:
break
#else:
###batch_x, batch_y = mnist.train.next_batch(batch_size)
# Run optimization op (backprop)
###sess.run(optimizer, feed_dict={x: batch_x, y: batch_y,
### keep_prob: dropout}) <<<<SOMETHING IS WRONG IN HERE?!!!
if step % display_step == 0:
# Calculate batch loss and accuracy
loss, acc = sess.run([cost, accuracy], feed_dict={x: fvecs_np,
y: labels_onehot,
keep_prob: 1.})
print("Iter " + str(step*batch_size) + ", Minibatch Loss= " + \
"{:.6f}".format(np.mean(loss)) + ", Training Accuracy= " + \
"{:.5f}".format(acc))
TrainAcc = float("{:.5f}".format(acc))
#print("******", TrainAcc)
if TrainAcc >= .99: #Changed from .95 temporarily
print(FitStep)
FitStep = FitStep+1
saver.save(sess, Sess_File_Name, global_step=1000) #
print("Saved Session:", Sess_File_Name)
step += 1
print("Optimization Finished!")
print("Testing Accuracy:", \
sess.run(accuracy, feed_dict={x: fvecs_np[:256],
y: labels_onehot[:256],
keep_prob: 1.}))
#feed_dictTEST = {x: fvecs_np[50]}
#prediction=tf.argmax(y,1)
#print(prediction)
#best = sess.run([prediction],feed_dictTEST)
#print(best)
print("DONE")
sess.close()
def Tax_Iterator(CSV_inFile, CSV_outFile): #Deprecate
#Need to copy *.csv file to MySQL for sorting
resultFileINIT = open(CSV_outFile,'w')
resultFileINIT.close()
TaxCount = 0
TaxThreshold = 2165
ThresholdStep = 2165
PrevTax = 0
linecounter = 0
#Open all GenBank profile list
for line in open(CSV_inFile):
linecounter = linecounter+1
print(linecounter)
resultFile = open(CSV_outFile,'a')
wr = csv.writer(resultFile, dialect='excel')
# Check for new TXID
row = line.split(',')
print(row[7], "===", PrevTax)
if row[7] != PrevTax:
print("X1")
TaxCount = TaxCount+1
PrevTax = row[7]
#Check it current Tax count is < or > threshold
# < threshold
print(TaxCount,"=+=", TaxThreshold)
if TaxCount<=3300:
print("X2")
CurrentTax= row[7]
CurrTxCount = CurrentTax
print("TaxCount=", TaxCount)
print( "Add to CSV")
print("row:", CurrentTax, "***", row[0:15])
wr.writerow(row[0:-1])
# is > threshold
else:
print("X3")
# but same TXID....
print(row[7], "=-=", CurrentTax)
if row[7]==CurrentTax:
print("X4")
CurrentTax= row[7]
print("TaxCount=", TaxCount)
print( "Add to CSV")
print("row:", CurrentTax, "***", row[0:15])
wr.writerow(row[0:-1])
# but different TXID...
else:
print(row[7], "=*=", CurrentTax)
if row[7]>CurrentTax:
print("X5")
TaxThreshold=TaxThreshold+ThresholdStep
resultFile.close()
Sess_File_Name = "CNN_VirusIDvSPECIES_XXALL"+ str(TaxThreshold-ThresholdStep)
print("<<<< Start Training >>>>"
print("Training on :: ", CurrTxCount, "Taxa", TaxCount, "data points.")
Train_Network(CurrTxCount, Sess_File_Name)
print("Training complete")
resultFileINIT = open(CSV_outFile,'w')
resultFileINIT.close()
CurrentTax= row[7]
#reset tax count
CurrTxCount = 0
TaxCount = 0
resultFile.close()
Sess_File_Name = "MicroarrayCNN_Data"+ str(TaxThreshold+ThresholdStep)
print("<<<< Start Training >>>>")
print("Training on :: ", CurrTxCount, "Taxa", TaxCount, "data points.")
Train_Network(CurrTxCount, Sess_File_Name)
resultFileINIT = open(CSV_outFile,'w')
resultFileINIT.close()
CurrentTax= row[7]
Tax_Iterator("MicroarrayInput.csv", "MicroarrayOutput.csv")
You defined prediction as prediction=tf.argmax(y,1). And in both feed_dict, you feed labels_onehot for y. Consequently, your "prediction" is always equal to the labels.
As you didn't post the code you used to train your network, I can't tell you what exactly you need to change.
Edit: I have isses understanding the underlying problem you're trying to solve - based on your code, you're trying to train a neural network with 2165 different classes using 1609 training examples. How is this even possible? If each example had a different class, there would still be some classes without any training example. Or does one image belong to many classes? From your statement at the beginning of your question, I had assumed you're trying to output a real-valued number between 0-1.
I'm actually surprised that the code actually worked as it looks like you're adding only a single number to your labels list, but your model expects a list with length 2165 for each training example.