commit 127b1d82c9bdc3792ada3a77df60963f4a1458ad
parent a932a0d864b832d5cfc92de680346f05adcfed7c
Author: David Freifeld <freifeld.david@gmail.com>
Date: Sun, 21 Jun 2020 13:02:04 -0700
Trains in parallel (with bugs)
Diffstat:
3 files changed, 34 insertions(+), 38 deletions(-)
diff --git a/bpnn.cpp b/bpnn.cpp
@@ -134,9 +134,9 @@ float Network::accuracy()
float correct = 0;
int total = 0;
for (int i = 0; i < layers[length-1].contents->rows(); i++) {
- printf("%i vs %f\n", (int)(*labels)(i, 0), (*layers[length-1].contents)(i, 0));
+ // printf("%i vs %f\n", (int)(*labels)(i, 0), (*layers[length-1].contents)(i, 0));
if ((*labels)(i, 0) == round((*layers[length-1].contents)(i, 0))) {
- printf("Correct!\n");
+ //printf("Correct!\n");
correct += 1;
}
total = i;
@@ -152,8 +152,6 @@ void Network::backpropagate()
std::vector<Eigen::MatrixXd> gradients;
std::vector<Eigen::MatrixXd> deltas;
Eigen::MatrixXd error = ((*layers[length-1].contents) - (*labels)).cwiseProduct(((*layers[length-1].contents) - (*labels)));
- error = error.cwiseProduct(((*layers[length-1].contents) - (*labels)).cwiseProduct(((*layers[length-1].contents) - (*labels))));
- error = error.cwiseProduct(((*layers[length-1].contents) - (*labels)).cwiseProduct(((*layers[length-1].contents) - (*labels))));
gradients.push_back(error.cwiseProduct(*layers[length-1].dZ));
deltas.push_back((*layers[length-2].contents).transpose() * gradients[0]);
int counter = 1;
@@ -272,7 +270,6 @@ float Network::test(char* path)
void demo(int total_epochs)
{
- auto begin = std::chrono::high_resolution_clock::now();
// std::cout << "\n\n\n";
int linecount = prep_file("./extra.txt", "./shuffled.txt");
Network net ("./shuffled.txt", 4, 1, 1, 5, 10, 1);
@@ -324,6 +321,4 @@ void demo(int total_epochs)
}
net.list_net();
printf("Test accuracy: %f\n", net.test("./test.txt"));
- auto end = std::chrono::high_resolution_clock::now();
- std::cout <<std::chrono::duration_cast<std::chrono::nanoseconds>(end-begin).count() << "ns aka " << (double) std::chrono::duration_cast<std::chrono::nanoseconds>(end-begin).count() / pow(10,9) << "s" << std::endl;
}
diff --git a/kerasdemo.py b/kerasdemo.py
@@ -17,21 +17,21 @@ from keras.models import Sequential
from keras.layers import Dense
init = time.time()
# load the dataset
-dataset = loadtxt('data_banknote_authentication.txt', delimiter=',')
+dataset = loadtxt('extra.txt', delimiter=',')
# split into input (X) and output (y) variables
X = dataset[:,0:4]
y = dataset[:,4]
# define the keras model
model = Sequential()
-model.add(Dense(4, input_dim=4, activation='sigmoid'))
+model.add(Dense(4, input_dim=4, activation='linear'))
model.add(Dense(5, activation='sigmoid'))
model.add(Dense(5, activation='sigmoid'))
-model.add(Dense(1, activation='relu'))
+model.add(Dense(1, activation='sigmoid'))
# compile the keras model
opt = keras.optimizers.SGD(lr=1)
model.compile(loss='mse', optimizer=opt, metrics=['accuracy'])
# fit the keras model on the dataset
-model.fit(X, y, epochs=50, batch_size=1)
+model.fit(X, y, epochs=50, batch_size=10)
# evaluate the keras model
_, accuracy = model.evaluate(X, y)
print('Accuracy: %.2f' % (accuracy*100))
diff --git a/mr_bpnn_2.cpp b/mr_bpnn_2.cpp
@@ -11,52 +11,51 @@ struct pair* map (struct pair input_pair)
float epoch_cost = 1000;
float epoch_accuracy = -1;
int epochs = 0;
+ int total_epochs = 50;
net->batches= 0;
// net.feedforward();
// net.backpropagate();
// std::cout << net.cost() << "\n";
printf("Beginning train on %i instances for %i epochs...\n", linecount, 50);
- while (epochs < 1) {
+ while (epochs < total_epochs) {
auto ep_begin = std::chrono::high_resolution_clock::now();
+ // int linecount = prep_file("./data_banknote_authentication.txt");
float cost_sum = 0;
float acc_sum = 0;
- // int linecount = prep_file("./data_banknote_authentication.txt", "./shuffled.txt");
double times[5] = {0};
for (int i = 0; i <= linecount-net->batch_size; i+=net->batch_size) {
- auto feed_begin = std::chrono::high_resolution_clock::now();
+ // auto feed_begin = std::chrono::high_resolution_clock::now();
net->feedforward();
- auto back_begin = std::chrono::high_resolution_clock::now();
+ // auto back_begin = std::chrono::high_resolution_clock::now();
net->backpropagate();
- auto cost_begin = std::chrono::high_resolution_clock::now();
+ // auto cost_begin = std::chrono::high_resolution_clock::now();
cost_sum += net->cost();
// std::cout << acc_sum << " "<< net.accuracy() << " " << net.batch_size << "\n";
- auto acc_begin = std::chrono::high_resolution_clock::now();
- if (i < 1*net->batch_size) {
- printf("Batch accuracy: %f\n", net->accuracy());
- }
+ // auto acc_begin = std::chrono::high_resolution_clock::now();
acc_sum += net->accuracy();
// std::cout << net.cost() << " as it is " << net.labels[0] << " vs " << *net.layers[net.length-1].contents << "\n";
- auto batch_begin = std::chrono::high_resolution_clock::now();
+ // auto batch_begin = std::chrono::high_resolution_clock::now();
- int exit = net->next_batch(net->fpath);
- auto loop_end = std::chrono::high_resolution_clock::now();
- times[0] += std::chrono::duration_cast<std::chrono::nanoseconds>(back_begin - feed_begin).count() / pow(10,9);
- times[1] += std::chrono::duration_cast<std::chrono::nanoseconds>(cost_begin - back_begin).count() / pow(10,9);
- times[2] += std::chrono::duration_cast<std::chrono::nanoseconds>(acc_begin - cost_begin).count() / pow(10,9);
- times[3] += std::chrono::duration_cast<std::chrono::nanoseconds>(batch_begin - acc_begin).count() / pow(10,9);
- times[4] += std::chrono::duration_cast<std::chrono::nanoseconds>(loop_end - batch_begin).count() / pow(10,9);
- net->batches++;
- if (exit == -1) {
- break;
+ if (i != linecount-net->batch_size) { // Don't try to advance batch on final batch.
+ net->next_batch(net->fpath);
}
+ net->batches++;
+ // auto loop_end = std::chrono::high_resolution_clock::now();
+ // times[0] += std::chrono::duration_cast<std::chrono::nanoseconds>(back_begin -feed_begin).count() / pow(10,9);
+ // times[1] += std::chrono::duration_cast<std::chrono::nanoseconds>(cost_begin - back_begin).count() / pow(10,9);
+ // times[2] += std::chrono::duration_cast<std::chrono::nanoseconds>(acc_begin - cost_begin).count() / pow(10,9);
+ // times[3] += std::chrono::duration_cast<std::chrono::nanoseconds>(batch_begin - acc_begin).count() / pow(10,9);
+ // times[4] += std::chrono::duration_cast<std::chrono::nanoseconds>(loop_end - batch_begin).count() / pow(10,9);
}
- printf("Avg time spent across %i batches: %lf on feedforward, %lf on backprop, %lf on cost, %lf on acc, %lf on next batch\n", net->batches, times[0]/net->batches, times[1]/net->batches, times[2]/net->batches, times[3]/net->batches, times[4]/net->batches);
- net->batches=1;
epoch_accuracy = 1.0/((float) linecount/net->batch_size) * acc_sum;
epoch_cost = 1.0/((float) linecount/net->batch_size) * cost_sum;
auto ep_end = std::chrono::high_resolution_clock::now();
- printf("Epoch %i/%i - time %f - cost %f - acc %f\n", epochs+1, 50, (double) std::chrono::duration_cast<std::chrono::nanoseconds>(ep_end-ep_begin).count() / pow(10,9), epoch_cost, epoch_accuracy);
+ double epochtime = (double) std::chrono::duration_cast<std::chrono::nanoseconds>(ep_end-ep_begin).count() / pow(10,9);
+ printf("Epoch %i/%i - time %f - cost %f - acc %f\n", epochs+1, total_epochs, epochtime, epoch_cost, epoch_accuracy);
+ // printf("Avg time spent across %i batches: %lf on feedforward, %lf on backprop, %lf on cost, %lf on acc, %lf on next batch.\n", net.batches, times[0]/net.batches, times[1]/net.batches, times[2]/net.batches, times[3]/net.batches, times[4]/net.batches);
+ // printf("Time spent across epoch: %lf on feedforward, %lf on backprop, %lf on cost, %lf on acc, %lf on next batch, %lf other.\n", times[0], times[1], times[2], times[3], times[4], epochtime-times[0]-times[1]-times[2]-times[3]-times[4]);
+ net->batches=1;
epochs++;
}
struct pair* output = new struct pair;
@@ -64,13 +63,11 @@ struct pair* map (struct pair input_pair)
strcpy(key, path);
output[0].key = key;
output[0].value = net;
- printf("%p %p VALS\n", output[0].key, output[0].value);
return output;
}
struct pair* reduce (struct pair* input_pairs)
{
- printf("%p %p VALS\n", input_pairs[0].key, input_pairs[0].value);
struct pair* output = new struct pair[6];
for (int i = 0; input_pairs[i].key != 0x0; i++) {
Network net = *(Network*)input_pairs[i].value;
@@ -78,7 +75,6 @@ struct pair* reduce (struct pair* input_pairs)
*acc = net.test("./test.txt");
output[i].key = input_pairs[i].key;
output[i].value = acc;
- printf("%s %f\n", (char*)output[i].key, *(float*)output[i].value);
}
return output;
}
@@ -107,6 +103,11 @@ void translate(char* path)
int main(int argc, char** argv)
{
- // begin(argv[2], map, reduce, translate, strtol(argv[1], NULL, 10), 1, argv[3], strtol(argv[4], NULL, 10));
+ auto prog_begin = std::chrono::high_resolution_clock::now();
+ //begin(argv[2], map, reduce, translate, strtol(argv[1], NULL, 10), 1, argv[3], strtol(argv[4], NULL, 10));
demo(50);
+ auto prog_end = std::chrono::high_resolution_clock::now();
+ std::cout << "Time: " << std::chrono::duration_cast<std::chrono::nanoseconds>(prog_end-prog_begin).count() / pow(10,9) << "\n";
+
+ // demo(50);
}
