C++ cuda_make_array函数代码示例

void time_ongpu(int TA, int TB, int m, int k, int n)
{
    int iter = 10;
    float *a = random_matrix(m,k);
    float *b = random_matrix(k,n);

    int lda = (!TA)?k:m;
    int ldb = (!TB)?n:k;

    float *c = random_matrix(m,n);

    float *a_cl = cuda_make_array(a, m*k);
    float *b_cl = cuda_make_array(b, k*n);
    float *c_cl = cuda_make_array(c, m*n);

    int i;
    clock_t start = clock(), end;
    for(i = 0; i<iter; ++i){
        gemm_ongpu(TA,TB,m,n,k,1,a_cl,lda,b_cl,ldb,1,c_cl,n);
        cudaThreadSynchronize();
    }
    double flop = ((double)m)*n*(2.*k + 2.)*iter;
    double gflop = flop/pow(10., 9);
    end = clock();
    double seconds = sec(end-start);
    printf("Matrix Multiplication %dx%d * %dx%d, TA=%d, TB=%d: %lf s, %lf GFLOPS\n",m,k,k,n, TA, TB, seconds, gflop/seconds);
    cuda_free(a_cl);
    cuda_free(b_cl);
    cuda_free(c_cl);
    free(a);
    free(b);
    free(c);
}

layer make_shortcut_layer(int batch, int index, int w, int h, int c, int w2, int h2, int c2)
{
    fprintf(stderr,"Shortcut Layer: %d\n", index);
    layer l = {0};
    l.type = SHORTCUT;
    l.batch = batch;
    l.w = w2;
    l.h = h2;
    l.c = c2;
    l.out_w = w;
    l.out_h = h;
    l.out_c = c;
    l.outputs = w*h*c;
    l.inputs = l.outputs;

    l.index = index;

    l.delta =  calloc(l.outputs*batch, sizeof(float));
    l.output = calloc(l.outputs*batch, sizeof(float));;

    l.forward = forward_shortcut_layer;
    l.backward = backward_shortcut_layer;
    #ifdef GPU
    l.forward_gpu = forward_shortcut_layer_gpu;
    l.backward_gpu = backward_shortcut_layer_gpu;

    l.delta_gpu =  cuda_make_array(l.delta, l.outputs*batch);
    l.output_gpu = cuda_make_array(l.output, l.outputs*batch);
    #endif
    return l;
}

maxpool_layer make_maxpool_layer(int batch, int h, int w, int c, int size, int stride, int padding)
{
    maxpool_layer l = {};

    l.type = MAXPOOL;
    l.batch = batch;
    l.h = h;
    l.w = w;
    l.c = c;
    l.pad = padding;
    l.out_w = (w + 2*padding)/stride;
    l.out_h = (h + 2*padding)/stride;
    l.out_c = c;
    l.outputs = l.out_h * l.out_w * l.out_c;
    l.inputs = h*w*c;
    l.size = size;
    l.stride = stride;
    int output_size = l.out_h * l.out_w * l.out_c * batch;
    l.indexes = (int*)calloc(output_size, sizeof(int));
    l.output =  (float*)calloc(output_size, sizeof(float));
    l.delta =   (float*)calloc(output_size, sizeof(float));
    l.forward = forward_maxpool_layer;
    l.backward = backward_maxpool_layer;
    #ifdef GPU
    l.forward_gpu = forward_maxpool_layer_gpu;
    l.backward_gpu = backward_maxpool_layer_gpu;
    l.indexes_gpu = cuda_make_int_array(output_size);
    l.output_gpu  = cuda_make_array(l.output, output_size);
    l.delta_gpu   = cuda_make_array(l.delta, output_size);
    #endif
    fprintf(stderr, "max          %d x %d / %d  %4d x%4d x%4d   ->  %4d x%4d x%4d\n", size, size, stride, w, h, c, l.out_w, l.out_h, l.out_c);
    return l;
}

layer make_compact_layer(int batch, int splits, int method, int w, int h, int c)
{
    fprintf(stderr,"Compact Layer: Split and merge channels in %d groups. %d %d %d\n", splits,w ,h ,c);
    layer l = {0};
    l.type = COMPACT;
    l.batch = batch;
    l.w = w;
    l.h = h;
    l.c = c;
    l.out_w = w;
    l.out_h = h;
    if (method<10)
    {
        l.out_c = c/splits;
    } else
    {
        l.out_c = c;
    }
    l.outputs = l.out_w * l.out_h * l.out_c;
    l.inputs = w*h*c;

    l.index = splits;
    l.method = method;

    l.delta =  calloc(l.outputs*batch, sizeof(float));
    l.output = calloc(l.outputs*batch, sizeof(float));;
    l.indexes = calloc(l.outputs*batch, sizeof(int));
    #ifdef GPU
    l.delta_gpu =  cuda_make_array(l.delta, l.outputs*batch);
    l.output_gpu = cuda_make_array(l.output, l.outputs*batch);
    l.indexes_gpu = cuda_make_int_array(l.outputs*batch);
    #endif

    return l;
}

cost_layer make_cost_layer(int batch, int inputs, COST_TYPE cost_type, float scale)
{
    fprintf(stderr, "cost                                           %4d\n",  inputs);
    cost_layer l = {0};
    l.type = COST;

    l.scale = scale;
    l.batch = batch;
    l.inputs = inputs;
    l.outputs = inputs;
    l.cost_type = cost_type;
    l.delta = calloc(inputs*batch, sizeof(float));
    l.output = calloc(inputs*batch, sizeof(float));
    l.cost = calloc(1, sizeof(float));

    l.forward = forward_cost_layer;
    l.backward = backward_cost_layer;
    #ifdef GPU
    l.forward_gpu = forward_cost_layer_gpu;
    l.backward_gpu = backward_cost_layer_gpu;

    l.delta_gpu = cuda_make_array(l.output, inputs*batch);
    l.output_gpu = cuda_make_array(l.delta, inputs*batch);
    #endif
    return l;
}

crop_layer make_crop_layer(int batch, int h, int w, int c, int crop_height, int crop_width, int flip, float angle, float saturation, float exposure)
{
    fprintf(stderr, "Crop Layer: %d x %d -> %d x %d x %d image\n", h,w,crop_height,crop_width,c);
    crop_layer l = {0};
    l.type = CROP;
    l.batch = batch;
    l.h = h;
    l.w = w;
    l.c = c;
    l.flip = flip;
    l.angle = angle;
    l.saturation = saturation;
    l.exposure = exposure;
    l.crop_width = crop_width;
    l.crop_height = crop_height;
    l.out_w = crop_width;
    l.out_h = crop_height;
    l.out_c = c;
    l.inputs = l.w * l.h * l.c;
    l.outputs = l.out_w * l.out_h * l.out_c;
    l.output = calloc(crop_width*crop_height * c*batch, sizeof(float));
    #ifdef GPU
    l.output_gpu = cuda_make_array(l.output, crop_width*crop_height*c*batch);
    l.rand_gpu = cuda_make_array(0, l.batch*8);
    #endif
    return l;
}

route_layer make_route_layer(int batch, int n, int *input_layers, int *input_sizes)
{
    fprintf(stderr,"Route Layer:");
    route_layer l = {0};
    l.type = ROUTE;
    l.batch = batch;
    l.n = n;
    l.input_layers = input_layers;
    l.input_sizes = input_sizes;
    int i;
    int outputs = 0;
    for(i = 0; i < n; ++i){
        fprintf(stderr," %d", input_layers[i]);
        outputs += input_sizes[i];
    }
    fprintf(stderr, "\n");
    l.outputs = outputs;
    l.inputs = outputs;
    l.delta = calloc(outputs*batch, sizeof(float));
    l.output = calloc(outputs*batch, sizeof(float));;
    #ifdef GPU
    l.delta_gpu = cuda_make_array(0, outputs*batch);
    l.output_gpu = cuda_make_array(0, outputs*batch);
    #endif
    return l;
}

crop_layer make_crop_layer(int batch, int h, int w, int c, int crop_height,
		int crop_width, int flip, real_t angle, real_t saturation,
		real_t exposure) {
	fprintf(stderr, "Crop Layer: %d x %d -> %d x %d x %d image\n", h, w,
			crop_height, crop_width, c);
	crop_layer l = { 0 };
	l.type = CROP;
	l.batch = batch;
	l.h = h;
	l.w = w;
	l.c = c;
	l.scale = (real_t) crop_height / h;
	l.flip = flip;
	l.angle = angle;
	l.saturation = saturation;
	l.exposure = exposure;
	l.out_w = crop_width;
	l.out_h = crop_height;
	l.out_c = c;
	l.inputs = l.w * l.h * l.c;
	l.outputs = l.out_w * l.out_h * l.out_c;
	l.output = calloc(l.outputs * batch, sizeof(real_t));
	l.forward = forward_crop_layer;
	l.backward = backward_crop_layer;

#ifdef GPU
	l.forward_gpu = forward_crop_layer_gpu;
	l.backward_gpu = backward_crop_layer_gpu;
	l.output_gpu = cuda_make_array(l.output, l.outputs*batch);
	l.rand_gpu = cuda_make_array(0, l.batch*8);
#endif
	return l;
}

void resize_reorg_old_layer(layer *l, int w, int h)
{
    int stride = l->stride;
    int c = l->c;

    l->h = h;
    l->w = w;

    if(l->reverse){
        l->out_w = w*stride;
        l->out_h = h*stride;
        l->out_c = c/(stride*stride);
    }else{
        l->out_w = w/stride;
        l->out_h = h/stride;
        l->out_c = c*(stride*stride);
    }

    l->outputs = l->out_h * l->out_w * l->out_c;
    l->inputs = l->outputs;
    int output_size = l->outputs * l->batch;

    l->output = realloc(l->output, output_size * sizeof(float));
    l->delta = realloc(l->delta, output_size * sizeof(float));

#ifdef GPU
    cuda_free(l->output_gpu);
    cuda_free(l->delta_gpu);
    l->output_gpu  = cuda_make_array(l->output, output_size);
    l->delta_gpu   = cuda_make_array(l->delta,  output_size);
#endif
}

detection_layer make_detection_layer(int batch, int inputs, int n, int side, int classes, int coords, int rescore, const bool b_debug)
{
    detection_layer l = {0};
    l.type    = DETECTION;

    l.n       = n;
    l.batch   = batch;
    l.inputs  = inputs;
    l.classes = classes;
    l.coords  = coords;
    l.rescore = rescore;
    l.b_debug = b_debug;

    l.side = side;
    assert(side*side*((1 + l.coords)*l.n + l.classes) == inputs);
    l.cost = calloc(1, sizeof(float));
    l.outputs = l.inputs;
    l.truths = l.side*l.side*(1+l.coords+l.classes);
    l.output = calloc(batch*l.outputs, sizeof(float));
    l.delta = calloc(batch*l.outputs, sizeof(float));
#ifdef GPU
    l.output_gpu = cuda_make_array(l.output, batch*l.outputs);
    l.delta_gpu = cuda_make_array(l.delta, batch*l.outputs);
#endif

    fprintf(stderr, "Detection Layer\n");
    srand(0);

    return l;
}

void resize_convolutional_layer(convolutional_layer *l, int w, int h)
{
    l->w = w;
    l->h = h;
    int out_w = convolutional_out_width(*l);
    int out_h = convolutional_out_height(*l);

    l->out_w = out_w;
    l->out_h = out_h;

    l->outputs = l->out_h * l->out_w * l->out_c;
    l->inputs = l->w * l->h * l->c;

    l->col_image = realloc(l->col_image,
            out_h*out_w*l->size*l->size*l->c*sizeof(float));
    l->output = realloc(l->output,
            l->batch*out_h * out_w * l->n*sizeof(float));
    l->delta  = realloc(l->delta,
            l->batch*out_h * out_w * l->n*sizeof(float));

#ifdef GPU
    cuda_free(l->col_image_gpu);
    cuda_free(l->delta_gpu);
    cuda_free(l->output_gpu);

    l->col_image_gpu = cuda_make_array(l->col_image, out_h*out_w*l->size*l->size*l->c);
    l->delta_gpu =     cuda_make_array(l->delta, l->batch*out_h*out_w*l->n);
    l->output_gpu =    cuda_make_array(l->output, l->batch*out_h*out_w*l->n);
#endif
}

avgpool_layer make_avgpool_layer(int batch, int w, int h, int c) {
	fprintf(stderr, "avg                     %4d x%4d x%4d   ->  %4d\n", w, h,
			c, c);
	avgpool_layer l = { 0 };
	l.type = AVGPOOL;
	l.batch = batch;
	l.h = h;
	l.w = w;
	l.c = c;
	l.out_w = 1;
	l.out_h = 1;
	l.out_c = c;
	l.outputs = l.out_c;
	l.inputs = h * w * c;
	int output_size = l.outputs * batch;
	l.output = calloc(output_size, sizeof(real_t));
	l.delta = calloc(output_size, sizeof(real_t));
	l.forward = forward_avgpool_layer;
	l.backward = backward_avgpool_layer;
#ifdef GPU
	l.forward_gpu = forward_avgpool_layer_gpu;
	l.backward_gpu = backward_avgpool_layer_gpu;
	l.output_gpu = cuda_make_array(l.output, output_size);
	l.delta_gpu = cuda_make_array(l.delta, output_size);
#endif
	return l;
}

detection_layer make_detection_layer(int batch, int inputs, int classes, int coords, int joint, int rescore, int background, int objectness)
{
    detection_layer l = {0};
    l.type = DETECTION;
    
    l.batch = batch;
    l.inputs = inputs;
    l.classes = classes;
    l.coords = coords;
    l.rescore = rescore;
    l.objectness = objectness;
    l.background = background;
    l.joint = joint;
    l.cost = calloc(1, sizeof(float));
    l.does_cost=1;
    int outputs = get_detection_layer_output_size(l);
    l.outputs = outputs;
    l.output = calloc(batch*outputs, sizeof(float));
    l.delta = calloc(batch*outputs, sizeof(float));
    #ifdef GPU
    l.output_gpu = cuda_make_array(0, batch*outputs);
    l.delta_gpu = cuda_make_array(0, batch*outputs);
    #endif

    fprintf(stderr, "Detection Layer\n");
    srand(0);

    return l;
}

region_layer make_region_layer(int batch, int inputs, int n, int side, int classes, int coords, int rescore)
{
    region_layer l = {0};
    l.type = REGION;
    
    l.n = n;
    l.batch = batch;
    l.inputs = inputs;
    l.classes = classes;
    l.coords = coords;
    l.rescore = rescore;
    l.side = side;
    assert(side*side*l.coords*l.n == inputs);
    l.cost = calloc(1, sizeof(float));
    int outputs = l.n*5*side*side;
    l.outputs = outputs;
    l.output = calloc(batch*outputs, sizeof(float));
    l.delta = calloc(batch*inputs, sizeof(float));
    #ifdef GPU
    l.output_gpu = cuda_make_array(l.output, batch*outputs);
    l.delta_gpu = cuda_make_array(l.delta, batch*inputs);
#endif

    fprintf(stderr, "Region Layer\n");
    srand(0);

    return l;
}

void resize_maxpool_layer(maxpool_layer *l, int w, int h)
{
    int stride = l->stride;
    l->h = h;
    l->w = w;
    l->inputs = h*w*l->c;

    l->out_w = (w-1)/stride + 1;
    l->out_h = (h-1)/stride + 1;
    l->outputs = l->out_w * l->out_h * l->c;
    int output_size = l->outputs * l->batch;

    l->indexes = realloc(l->indexes, output_size * sizeof(int));
    l->output = realloc(l->output, output_size * sizeof(float));
    l->delta = realloc(l->delta, output_size * sizeof(float));

    #ifdef GPU
    cuda_free((float *)l->indexes_gpu);
    cuda_free(l->output_gpu);
    cuda_free(l->delta_gpu);
    l->indexes_gpu = cuda_make_int_array(output_size);
    l->output_gpu  = cuda_make_array(l->output, output_size);
    l->delta_gpu   = cuda_make_array(l->delta,  output_size);
    #endif
}