precision test for Torch7 and new hardware

precision-network.lua

--[[ precision-test

Compare precision of hardware and software implementation
run with: qlua precision-network.lua

--]]

require 'nn'
require 'pl'
require 'image'
torch.setdefaulttensortype('torch.FloatTensor')


local iC = 1     -- input channel
local iH = 128   -- input size
local iW = iH
local kC = 4     -- nb kernels
local kH = 3     -- kernel size
local kW = kH
local pH = 2     -- pool size
local pW = pH


-- define network and weights
network = nn.Sequential()
network:add(nn.SpatialConvolutionMM(iC, kC, kH, kW))
-- network:add(nn.Threshold())
-- network:add(nn.SpatialMaxPooling(pH, pW, pH, pW))

for i = 1, kC do
   network.modules[1].weight[i]:fill(0.01*i)
end
network.modules[1].bias:fill(0)

-- use lena as an input
local lena_red  = image.lena()[1]
local lena_byte = image.scale(lena_red,iW,iH):resize(1,iH,iW):mul(256):byte()
local lena_sw   = image.scale(lena_red,iW,iH):resize(1,iH,iW):expand(iC,iH,iW)
local lena_hw   = torch.repeatTensor(lena_byte,iC,1,1)

-- parse network
local dst_sw = network:forward(lena_sw)
local dst_hw = network:forward(lena_sw) -- replace this with new library functions!


-- print output
print('==> Precision test')
local precision = 5
local coordinate = 20
local function trunc(x)
   return math.floor(x*math.pow(10,precision)+.5)/math.pow(10,precision)
end

for i = 1, kC do
   local sw = trunc(dst_sw[i][coordinate][coordinate])
   local hw = trunc(dst_hw[i][coordinate][coordinate])
   local diff = trunc(math.abs(sw-hw))
   print('output['..i..']: ', 'CPU = ', sw, 'FPGA = ', hw, 'DIFF = ', diff)
end

-- display output
image.display(dst_hw)