jacobian

commit 384b86c4270d0f4d32242d90a530919f5c17218e
parent e802136589583b2f9e3766a46f85b3a94f496ef4
Author: David Freifeld <freifeld.david@gmail.com>
Date:   Fri, 25 Sep 2020 16:59:13 -0700

Incorporating faster file read into BPNN

Diffstat:
M
example.cpp
 | 
3
++-
M
src/bpnn.cpp
 | 
7
++++---
M
src/data.cpp
 | 
99
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-----

3 files changed, 99 insertions(+), 10 deletions(-)
diff --git a/example.cpp b/example.cpp
@@ -19,7 +19,7 @@
 double bench(int batch_sz, int epochs)
 {
     auto start = std::chrono::high_resolution_clock::now();
-    Network net ("./data_banknote_authentication.txt", batch_sz, 0.0155, 0.03, L2, 0, 0.9);
+    Network net ("./data_banknote_authentication.bak.bin", batch_sz, 0.0155, 0.03, L2, 0, 0.9);
     net.add_layer(4, "linear", linear, linear_deriv);
     net.add_layer(5, "lecun_tanh", lecun_tanh, lecun_tanh_deriv);
     net.add_layer(2, "linear", linear, linear_deriv);
@@ -34,6 +34,7 @@ double bench(int batch_sz, int epochs)
 int main(int argc, char** argv)
 {
     if (argc < 2) {
+        prep();
         std::cout << "Invalid command! Either pass a special option or pass two integers - batch_size and epochs (in that order)."  << "\n";
         exit(1);
     }
diff --git a/src/bpnn.cpp b/src/bpnn.cpp
@@ -74,8 +74,8 @@ Network::Network(char* path, int batch_sz, float learn_rate, float bias_rate, Re
 {
     int total_instances = prep_file(path, SHUFFLED_PATH);
     val_instances = split_file(SHUFFLED_PATH, total_instances, ratio);
-    data = open(TRAIN_PATH, O_RDONLY & O_NONBLOCK);
-    val_data = open(VAL_PATH, O_RDONLY & O_NONBLOCK);
+    data = open(TRAIN_PATH, O_RDONLY | O_NONBLOCK);
+    val_data = open(VAL_PATH, O_RDONLY | O_NONBLOCK);
     instances = total_instances - val_instances;
     assert(batch_size > 0 || batch_size < instances);
     decay = [this]() -> void {};
@@ -84,6 +84,7 @@ Network::Network(char* path, int batch_sz, float learn_rate, float bias_rate, Re
     };
 }
 
+
 void Network::init_decay(char* type, ...)
 {
     va_list args;
@@ -368,7 +369,7 @@ void Network::train()
     validate(VAL_PATH);
     if (silenced == false) printf("Epoch %i complete - cost %f - acc %f - val_cost %f - val_acc %f\n", epochs, epoch_cost, epoch_acc, val_cost, val_acc);
     batches=1;
-    data = open(TRAIN_PATH, O_RDONLY & O_NONBLOCK);
+    data = open(TRAIN_PATH, O_RDONLY | O_NONBLOCK);
     decay();
     epochs++;
 }
diff --git a/src/data.cpp b/src/data.cpp
@@ -12,18 +12,105 @@ inline float scan(char **p)
     return n*neg;
 }
 
-int Network::next_batch()
+void prep()
 {
-    uintmax_t lines = 0;
-    while(size_t bytes_read = read(data, buf, BUFFER_SIZE))
+    FILE* wptr = fopen("../../data_banknote_authentication.bin", "wb");
+    static const auto BUFFER_SIZE = 512*1024;
+    int fd = open("../../data_banknote_authentication.txt", O_RDONLY | O_NONBLOCK);
+    if(fd == -1) {
+        printf("fd == -1\n");
+    }
+    float tmp;
+    char buf[BUFFER_SIZE + 1];
+    while(size_t bytes_read = read(fd, buf, BUFFER_SIZE))
     {
+        if(bytes_read == (size_t)-1) {
+            printf("bytes_read == (size_t)-1\n");
+        }
         if (!bytes_read) break;
-        for(char *p = buf; lines < 10;) {
+        for(char *p = buf;;) {
             char* bound = (char*) memchr(p, '\n', (buf + bytes_read) - p);
             if (bound - p < 0) break; // Stop.
-            for (int i=0; i<layers[0].contents->cols(); ++i) (*layers[0].contents)(lines,i) = scan(&p);
-            (*labels)(lines,0) = scan(&p);
+            for (int i=0; i<5; ++i) {
+                tmp = scan(&p);
+                fwrite((void*)&tmp, sizeof(float), 1, wptr);
+            }
             p = bound + 1;
+        }
+    }
+}
+
+float mmap_read()
+{
+    static const auto BUFFER_SIZE = 600*1024;
+    int fd = open("../../data_banknote_authentication.bin", O_RDONLY | O_NONBLOCK);
+    if(fd == -1) {
+        printf("Cannot open file.");
+        exit(1);
+    }
+    int rc, ii;
+    struct stat st;
+    size_t size;
+    rc = fstat(fd, &st);
+    size=st.st_size;
+    float* ptr = (float*) mmap(0, size, PROT_READ, MAP_SHARED, fd, 0);
+    madvise(ptr, size, POSIX_MADV_SEQUENTIAL);
+    char buf[BUFFER_SIZE];
+    uintmax_t lines = 0;
+    float tmp;
+    for (ii=0; ii < size/sizeof *ptr; ii++) {
+        // Nothin.
+    }
+    rc = munmap(ptr, size);
+    close(fd);
+    return tmp;
+}
+
+float vec_read()
+{
+#define CHUNK_SZ (200*1024)
+#define BUFFER_SZ (600*1024)
+#define NUM_CHUNKS (BUFFER_SZ/CHUNK_SZ)
+    int fd = open("../../data_banknote_authentication.bin", O_RDONLY | O_NONBLOCK);
+    char rawbuf[BUFFER_SZ];
+    char* buf = (char*) rawbuf;
+    iovec iovecs[NUM_CHUNKS];
+    for (int i = 0; i < BUFFER_SZ; i+=CHUNK_SZ) {
+        iovecs[i/CHUNK_SZ].iov_base = buf + i;
+        iovecs[i/CHUNK_SZ].iov_len = CHUNK_SZ;
+    }
+    while(size_t bytes_read = readv(fd, iovecs, NUM_CHUNKS))
+    {
+        if(bytes_read == (size_t)-1) {
+            printf("\n%zu\n", bytes_read);
+            printf("bytes_read == (size_t)-1\n");
+            exit(1);
+        }
+        if (!bytes_read) break;
+    }
+    return 0;
+}
+
+float std_read(int BUFFER_SIZE)
+{
+    int fd = open("../../data_banknote_authentication.bin", O_RDONLY | O_NONBLOCK);
+    fcntl(fd, F_RDADVISE);
+    if(fd == -1) {
+
+
+int Network::next_batch()
+{
+    uintmax_t lines = 0;
+    while(size_t bytes_read = read(data, buf, BUFFER_SIZE))
+    {
+        if (!bytes_read) break;
+        for(char *p = buf; p < buf+BUFFER_SIZE && lines < 10;) {
+            for (int i=0; i<<layers[0].contents->cols(); ++i) {
+                (*layers[0].contents)(lines,i) = *((float*)p);
+                p += sizeof(float);
+            }
+            (*labels)(lines,0) = *((float*)p);
+            p += sizeof(float);
             ++lines;
         }
     }