Mercurial > ift6266
view baseline/mlp/ratio_classes/mlp_nist_ratio.py @ 596:f6a3b28b002c
nips2010_submission.pdf
author | Yoshua Bengio <bengioy@iro.umontreal.ca> |
---|---|
date | Thu, 14 Oct 2010 15:52:02 -0400 |
parents | d8129a09ffb1 |
children |
line wrap: on
line source
# -*- coding: utf-8 -*- """ This tutorial introduces the multilayer perceptron using Theano. A multilayer perceptron is a logistic regressor where instead of feeding the input to the logistic regression you insert a intermidiate layer, called the hidden layer, that has a nonlinear activation function (usually tanh or sigmoid) . One can use many such hidden layers making the architecture deep. The tutorial will also tackle the problem of MNIST digit classification. .. math:: f(x) = G( b^{(2)} + W^{(2)}( s( b^{(1)} + W^{(1)} x))), References: - textbooks: "Pattern Recognition and Machine Learning" - Christopher M. Bishop, section 5 TODO: recommended preprocessing, lr ranges, regularization ranges (explain to do lr first, then add regularization) """ __docformat__ = 'restructedtext en' import setup_batches import pdb import numpy import theano import theano.tensor as T import time import theano.tensor.nnet import pylearn import theano,pylearn.version from pylearn.io import filetensor as ft data_path = '/data/lisa/data/nist/by_class/' class MLP(object): """Multi-Layer Perceptron Class A multilayer perceptron is a feedforward artificial neural network model that has one layer or more of hidden units and nonlinear activations. Intermidiate layers usually have as activation function thanh or the sigmoid function while the top layer is a softamx layer. """ def __init__(self, input, n_in, n_hidden, n_out,learning_rate, test_subclass): """Initialize the parameters for the multilayer perceptron :param input: symbolic variable that describes the input of the architecture (one minibatch) :param n_in: number of input units, the dimension of the space in which the datapoints lie :param n_hidden: number of hidden units :param n_out: number of output units, the dimension of the space in which the labels lie """ # initialize the parameters theta = (W1,b1,W2,b2) ; note that this # example contains only one hidden layer, but one can have as many # layers as he/she wishes, making the network deeper. The only # problem making the network deep this way is during learning, # backpropagation being unable to move the network from the starting # point towards; this is where pre-training helps, giving a good # starting point for backpropagation, but more about this in the # other tutorials # `W1` is initialized with `W1_values` which is uniformely sampled # from -6./sqrt(n_in+n_hidden) and 6./sqrt(n_in+n_hidden) # the output of uniform if converted using asarray to dtype # theano.config.floatX so that the code is runable on GPU W1_values = numpy.asarray( numpy.random.uniform( \ low = -numpy.sqrt(6./(n_in+n_hidden)), \ high = numpy.sqrt(6./(n_in+n_hidden)), \ size = (n_in, n_hidden)), dtype = theano.config.floatX) # `W2` is initialized with `W2_values` which is uniformely sampled # from -6./sqrt(n_hidden+n_out) and 6./sqrt(n_hidden+n_out) # the output of uniform if converted using asarray to dtype # theano.config.floatX so that the code is runable on GPU W2_values = numpy.asarray( numpy.random.uniform( low = -numpy.sqrt(6./(n_hidden+n_out)), \ high= numpy.sqrt(6./(n_hidden+n_out)),\ size= (n_hidden, n_out)), dtype = theano.config.floatX) self.W1 = theano.shared( value = W1_values ) self.b1 = theano.shared( value = numpy.zeros((n_hidden,), dtype= theano.config.floatX)) self.W2 = theano.shared( value = W2_values ) self.b2 = theano.shared( value = numpy.zeros((n_out,), dtype= theano.config.floatX)) #include the learning rate in the classifer so #we can modify it on the fly when we want lr_value=learning_rate self.lr=theano.shared(value=lr_value) # symbolic expression computing the values of the hidden layer self.hidden = T.tanh(T.dot(input, self.W1)+ self.b1) # symbolic expression computing the values of the top layer self.p_y_given_x= T.nnet.softmax(T.dot(self.hidden, self.W2)+self.b2) # compute prediction as class whose probability is maximal in # symbolic form #self.y_pred = T.argmax( self.p_y_given_x, axis =1) #self.y_pred_num = T.argmax( self.p_y_given_x[0:9], axis =1) self.test_subclass = test_subclass #if (self.test_subclass == "u"): # self.y_pred = T.argmax( self.p_y_given_x[10:35], axis =1) + 10 #elif (self.test_subclass == "l"): # self.y_pred = T.argmax( self.p_y_given_x[35:], axis =1) + 35 #elif (self.test_subclass == "d"): # self.y_pred = T.argmax( self.p_y_given_x[0:9], axis =1) #else: self.y_pred = T.argmax( self.p_y_given_x, axis =1) # L1 norm ; one regularization option is to enforce L1 norm to # be small self.L1 = abs(self.W1).sum() + abs(self.W2).sum() # square of L2 norm ; one regularization option is to enforce # square of L2 norm to be small self.L2_sqr = (self.W1**2).sum() + (self.W2**2).sum() def negative_log_likelihood(self, y): """Return the mean of the negative log-likelihood of the prediction of this model under a given target distribution. .. math:: \frac{1}{|\mathcal{D}|}\mathcal{L} (\theta=\{W,b\}, \mathcal{D}) = \frac{1}{|\mathcal{D}|}\sum_{i=0}^{|\mathcal{D}|} \log(P(Y=y^{(i)}|x^{(i)}, W,b)) \\ \ell (\theta=\{W,b\}, \mathcal{D}) :param y: corresponds to a vector that gives for each example the :correct label """ return -T.mean(T.log(self.p_y_given_x)[T.arange(y.shape[0]),y]) def errors(self, y): """Return a float representing the number of errors in the minibatch over the total number of examples of the minibatch """ # check if y has same dimension of y_pred if y.ndim != self.y_pred.ndim: raise TypeError('y should have the same shape as self.y_pred', ('y', target.type, 'y_pred', self.y_pred.type)) # check if y is of the correct datatype if y.dtype.startswith('int'): # the T.neq operator returns a vector of 0s and 1s, where 1 # represents a mistake in prediction return T.mean(T.neq(self.y_pred, y)) else: raise NotImplementedError() def mlp_full_nist( verbose = False,\ adaptive_lr = 1,\ train_data = 'all/all_train_data.ft',\ train_labels = 'all/all_train_labels.ft',\ test_data = 'all/all_test_data.ft',\ test_labels = 'all/all_test_labels.ft',\ learning_rate=0.5,\ L1_reg = 0.00,\ L2_reg = 0.0001,\ nb_max_exemples=1000000,\ batch_size=20,\ nb_hidden = 500,\ nb_targets = 26,\ tau=1e6,\ main_class="l",\ start_ratio=1,\ end_ratio=1): configuration = [learning_rate,nb_max_exemples,nb_hidden,adaptive_lr] #save initial learning rate if classical adaptive lr is used initial_lr=learning_rate total_validation_error_list = [] total_train_error_list = [] learning_rate_list=[] best_training_error=float('inf'); # set up batches batches = setup_batches.Batches() batches.set_batches(main_class, start_ratio,end_ratio,batch_size,verbose) train_batches = batches.get_train_batches() test_batches = batches.get_test_batches() validation_batches = batches.get_validation_batches() ishape = (32,32) # this is the size of NIST images # allocate symbolic variables for the data x = T.fmatrix() # the data is presented as rasterized images y = T.lvector() # the labels are presented as 1D vector of # [long int] labels if verbose==True: print 'finished parsing the data' # construct the logistic regression class classifier = MLP( input=x.reshape((batch_size,32*32)),\ n_in=32*32,\ n_hidden=nb_hidden,\ n_out=nb_targets,\ learning_rate=learning_rate,\ test_subclass=main_class) # the cost we minimize during training is the negative log likelihood of # the model plus the regularization terms (L1 and L2); cost is expressed # here symbolically cost = classifier.negative_log_likelihood(y) \ + L1_reg * classifier.L1 \ + L2_reg * classifier.L2_sqr # compiling a theano function that computes the mistakes that are made by # the model on a minibatch test_model = theano.function([x,y], classifier.errors(y)) # compute the gradient of cost with respect to theta = (W1, b1, W2, b2) g_W1 = T.grad(cost, classifier.W1) g_b1 = T.grad(cost, classifier.b1) g_W2 = T.grad(cost, classifier.W2) g_b2 = T.grad(cost, classifier.b2) # specify how to update the parameters of the model as a dictionary updates = \ { classifier.W1: classifier.W1 - classifier.lr*g_W1 \ , classifier.b1: classifier.b1 - classifier.lr*g_b1 \ , classifier.W2: classifier.W2 - classifier.lr*g_W2 \ , classifier.b2: classifier.b2 - classifier.lr*g_b2 } # compiling a theano function `train_model` that returns the cost, but in # the same time updates the parameter of the model based on the rules # defined in `updates` train_model = theano.function([x, y], cost, updates = updates ) n_minibatches = len(train_batches) #conditions for stopping the adaptation: #1) we have reached nb_max_exemples (this is rounded up to be a multiple of the train size) #2) validation error is going up twice in a row(probable overfitting) # This means we no longer stop on slow convergence as low learning rates stopped # too fast. # no longer relevant patience =nb_max_exemples/batch_size patience_increase = 2 # wait this much longer when a new best is # found improvement_threshold = 0.995 # a relative improvement of this much is # considered significant validation_frequency = n_minibatches/4 best_params = None best_validation_loss = float('inf') best_iter = 0 test_score = 0. start_time = time.clock() n_iter = nb_max_exemples/batch_size # nb of max times we are allowed to run through all exemples n_iter = n_iter/n_minibatches + 1 #round up n_iter=max(1,n_iter) # run at least once on short debug call time_n=0 #in unit of exemples if (main_class == "u"): class_offset = 10 elif (main_class == "l"): class_offset = 36 else: class_offset = 0 if verbose == True: print 'looping at most %d times through the data set' %n_iter for iter in xrange(n_iter* n_minibatches): # get epoch and minibatch index epoch = iter / n_minibatches minibatch_index = iter % n_minibatches if adaptive_lr==2: classifier.lr.value = tau*initial_lr/(tau+time_n) # get the minibatches corresponding to `iter` modulo # `len(train_batches)` x,y = train_batches[ minibatch_index ] y = y - class_offset # convert to float x_float = x/255.0 cost_ij = train_model(x_float,y) if (iter+1) % validation_frequency == 0: # compute zero-one loss on validation set this_validation_loss = 0. for x,y in validation_batches: # sum up the errors for each minibatch y = y - class_offset x_float = x/255.0 this_validation_loss += test_model(x_float,y) # get the average by dividing with the number of minibatches this_validation_loss /= len(validation_batches) #save the validation loss total_validation_error_list.append(this_validation_loss) #get the training error rate this_train_loss=0 for x,y in train_batches: # sum up the errors for each minibatch y = y - class_offset x_float = x/255.0 this_train_loss += test_model(x_float,y) # get the average by dividing with the number of minibatches this_train_loss /= len(train_batches) #save the validation loss total_train_error_list.append(this_train_loss) if(this_train_loss<best_training_error): best_training_error=this_train_loss if verbose == True: print('epoch %i, minibatch %i/%i, validation error %f, training error %f %%' % \ (epoch, minibatch_index+1, n_minibatches, \ this_validation_loss*100.,this_train_loss*100)) print 'learning rate = %f' %classifier.lr.value print 'time = %i' %time_n #save the learning rate learning_rate_list.append(classifier.lr.value) # if we got the best validation score until now if this_validation_loss < best_validation_loss: # save best validation score and iteration number best_validation_loss = this_validation_loss best_iter = iter # reset patience if we are going down again # so we continue exploring patience=nb_max_exemples/batch_size # test it on the test set test_score = 0. for x,y in test_batches: y = y - class_offset x_float=x/255.0 test_score += test_model(x_float,y) test_score /= len(test_batches) if verbose == True: print((' epoch %i, minibatch %i/%i, test error of best ' 'model %f %%') % (epoch, minibatch_index+1, n_minibatches, test_score*100.)) # if the validation error is going up, we are overfitting (or oscillating) # stop converging but run at least to next validation # to check overfitting or ocsillation # the saved weights of the model will be a bit off in that case elif this_validation_loss >= best_validation_loss: #calculate the test error at this point and exit # test it on the test set # however, if adaptive_lr is true, try reducing the lr to # get us out of an oscilliation if adaptive_lr==1: classifier.lr.value=classifier.lr.value/2.0 test_score = 0. #cap the patience so we are allowed one more validation error #calculation before aborting patience = iter+validation_frequency+1 for x,y in test_batches: y = y - class_offset x_float=x/255.0 test_score += test_model(x_float,y) test_score /= len(test_batches) if verbose == True: print ' validation error is going up, possibly stopping soon' print((' epoch %i, minibatch %i/%i, test error of best ' 'model %f %%') % (epoch, minibatch_index+1, n_minibatches, test_score*100.)) if iter>patience: print 'we have diverged' break time_n= time_n + batch_size end_time = time.clock() if verbose == True: print(('Optimization complete. Best validation score of %f %% ' 'obtained at iteration %i, with test performance %f %%') % (best_validation_loss * 100., best_iter, test_score*100.)) print ('The code ran for %f minutes' % ((end_time-start_time)/60.)) print iter #save the model and the weights numpy.savez('model.npy', config=configuration, W1=classifier.W1.value,W2=classifier.W2.value, b1=classifier.b1.value,b2=classifier.b2.value) numpy.savez('results.npy',config=configuration,total_train_error_list=total_train_error_list,total_validation_error_list=total_validation_error_list,\ learning_rate_list=learning_rate_list) return (best_training_error*100.0,best_validation_loss * 100.,test_score*100.,best_iter*batch_size,(end_time-start_time)/60) if __name__ == '__main__': mlp_full_nist(True) def jobman_mlp_full_nist(state,channel): (train_error,validation_error,test_error,nb_exemples,time)=mlp_full_nist(learning_rate=state.learning_rate,\ nb_hidden=state.nb_hidden,\ main_class=state.main_class,\ start_ratio=state.ratio,\ end_ratio=state.ratio) state.train_error=train_error state.validation_error=validation_error state.test_error=test_error state.nb_exemples=nb_exemples state.time=time return channel.COMPLETE