1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138

//! Provides common utility functions
use std::sync::{Arc, RwLock};
use co::prelude::*;
use coblas::plugin::*;
use conn;
use num::traits::{NumCast, cast};

/// Shared Lock used for our tensors
pub type ArcLock<T> = Arc<RwLock<T>>;

/// Create a simple native backend.
///
/// This is handy when you need to sync data to host memory to read/write it.
pub fn native_backend() -> Backend<Native> {
    let framework = Native::new();
    let hardwares = &framework.hardwares().to_vec();
    let backend_config = BackendConfig::new(framework, hardwares);
    Backend::new(backend_config).unwrap()
}

/// Write into a native Collenchyma Memory.
pub fn write_to_memory<T: NumCast + ::std::marker::Copy>(mem: &mut MemoryType, data: &[T]) {
    write_to_memory_offset(mem, data, 0);
}

/// Write into a native Collenchyma Memory with a offset.
pub fn write_to_memory_offset<T: NumCast + ::std::marker::Copy>(mem: &mut MemoryType, data: &[T], offset: usize) {
    match mem {
        &mut MemoryType::Native(ref mut mem) => {
            let mut mem_buffer = mem.as_mut_slice::<f32>();
            for (index, datum) in data.iter().enumerate() {
                // mem_buffer[index + offset] = *datum;
                mem_buffer[index + offset] = cast(*datum).unwrap();
            }
        },
        #[cfg(any(feature = "opencl", feature = "cuda"))]
        _ => {}
    }
}

/// Write the `i`th sample of a batch into a SharedTensor.
///
/// The size of a single sample is infered through
/// the first dimension of the SharedTensor, which
/// is asumed to be the batchsize.
///
/// Allocates memory on a Native Backend if neccessary.
pub fn write_batch_sample<T: NumCast + ::std::marker::Copy>(tensor: &mut SharedTensor<f32>, data: &[T], i: usize) {
    let native_backend = native_backend();

    let batch_size = tensor.desc().size();
    let sample_size = batch_size / tensor.desc()[0];

    let _ = tensor.add_device(native_backend.device());
    tensor.sync(native_backend.device()).unwrap();
    write_to_memory_offset(tensor.get_mut(native_backend.device()).unwrap(), &data, i * sample_size);
}

/// Create a Collenchyma SharedTensor for a scalar value.
pub fn native_scalar<T: NumCast + ::std::marker::Copy>(scalar: T) -> SharedTensor<T> {
    let native = native_backend();
    let mut shared_scalar = SharedTensor::<T>::new(native.device(), &vec![1]).unwrap();
    write_to_memory(shared_scalar.get_mut(native.device()).unwrap(), &[scalar]);

    shared_scalar
}

/// Casts a Vec<usize> to as Vec<i32>
pub fn cast_vec_usize_to_i32(input: Vec<usize>) -> Vec<i32> {
    let mut out = Vec::new();
    for i in input.iter() {
        out.push(*i as i32);
    }
    out
}

/// Extends IBlas with Axpby
pub trait Axpby<F> : Axpy<F> + Scal<F> {
    /// Performs the operation y := a*x + b*y .
    ///
    /// Consists of a scal(b, y) followed by a axpby(a,x,y).
    fn axpby(&self, a: &mut SharedTensor<F>, x: &mut SharedTensor<F>, b: &mut SharedTensor<F>, y: &mut SharedTensor<F>) -> Result<(), ::co::error::Error> {
        try!(self.scal(b, y));
        try!(self.axpy(a, x, y));
        Ok(())
    }

    /// Performs the operation y := a*x + b*y .
    ///
    /// Consists of a scal(b, y) followed by a axpby(a,x,y).
    fn axpby_plain(&self, a: &SharedTensor<F>, x: &SharedTensor<F>, b: &SharedTensor<F>, y: &mut SharedTensor<F>) -> Result<(), ::co::error::Error> {
        try!(self.scal_plain(b, y));
        try!(self.axpy_plain(a, x, y));
        Ok(())
    }
}

impl<T: Axpy<f32> + Scal<f32>> Axpby<f32> for T {}

/// Encapsulates all traits required by Solvers.
// pub trait SolverOps<F> : Axpby<F> + Dot<F> + Copy<F> {}
//
// impl<T: Axpby<f32> + Dot<f32> + Copy<f32>> SolverOps<f32> for T {}
pub trait SolverOps<F> : LayerOps<F> + Axpby<F> + Dot<F> + Copy<F> {}

impl<T: LayerOps<f32> + Axpby<f32> + Dot<f32> + Copy<f32>> SolverOps<f32> for T {}

/// Encapsulates all traits used in Layers.
#[cfg(all(feature="cuda", not(feature="native")))]
pub trait LayerOps<F> : conn::Convolution<F>
                      + conn::Pooling<F>
                      + conn::Relu<F> + conn::ReluPointwise<F>
                      + conn::Sigmoid<F> + conn::SigmoidPointwise<F>
                      + conn::Tanh<F> + conn::TanhPointwise<F>
                      + conn::Softmax<F> + conn::LogSoftmax<F>
                      + Gemm<F> {}
#[cfg(feature="native")]
/// Encapsulates all traits used in Layers.
pub trait LayerOps<F> : conn::Relu<F>
                      + conn::Sigmoid<F>
                      + conn::Tanh<F>
                      + conn::Softmax<F> + conn::LogSoftmax<F>
                      + Gemm<F> {}

#[cfg(all(feature="cuda", not(feature="native")))]
impl<T: conn::Convolution<f32>
      + conn::Pooling<f32>
      + conn::Relu<f32> + conn::ReluPointwise<f32>
      + conn::Sigmoid<f32> + conn::SigmoidPointwise<f32>
      + conn::Tanh<f32> + conn::TanhPointwise<f32>
      + conn::Softmax<f32> + conn::LogSoftmax<f32>
      + Gemm<f32>> LayerOps<f32> for T {}
#[cfg(feature="native")]
impl<T: conn::Relu<f32>
      + conn::Sigmoid<f32>
      + conn::Tanh<f32>
      + conn::Softmax<f32> + conn::LogSoftmax<f32>
      + Gemm<f32>> LayerOps<f32> for T {}