Problem with DecayOnValidation learning rate policy

DecayOnValidation_test.jl

use_cuda = false

using Mocha
srand(12345678)

############################################################
# Prepare Random Data
############################################################

srand(1234)

modeldir = "modeltest"

N = 1000
M = 20
P = 10

X = rand(M, N)
W = rand(M, P)
B = rand(P, 1)

Y = (W'*X .+ B)
Y = Y + 0.01*randn(size(Y))

############################################################
# Define network
############################################################
if use_cuda
  backend = GPUBackend()
else
  backend = CPUBackend()
end
init(backend)

data_layer = MemoryDataLayer(batch_size=500, data=Array[X,Y])
weight_layer = InnerProductLayer(name="ip",output_dim=P, tops=[:pred], bottoms=[:data])
loss_layer = SquareLossLayer(name="loss", bottoms=[:pred, :label])

net = Net("TEST", backend, [loss_layer, weight_layer, data_layer])
println(net)

############################################################
# Solve
############################################################

nepochs = 10

lr_policy = LRPolicy.DecayOnValidation(0.001, "loss-square-loss", 0.5)

params = SolverParameters(regu_coef=0.0005, mom_policy=MomPolicy.Fixed(0.9),
                          max_iter=N*nepochs, lr_policy=lr_policy,
                          load_from=modeldir)
solver = SGD(params)


setup_coffee_lounge(solver, save_into="$modeldir/statistics.jld", every_n_iter=1000)



val_performance = ValidationPerformance(net)

## add_coffee_break(solver, val_performance, every_n_epoch=2) # same at every_n_iter=2
add_coffee_break(solver, val_performance, every_n_iter=2000)

## register the listener to get notified on performance validation for lr_policy
setup(params.lr_policy, val_performance, solver)

## add_coffee_break(solver, Snapshot(modeldir), every_n_epoch=2) # same at every_n_iter=2
add_coffee_break(solver, Snapshot(modeldir), every_n_iter=2000)


add_coffee_break(solver, TrainingSummary(show_obj_val=true, show_lr=true), every_n_iter=500)

solve(solver, net)

shutdown(backend)

Results in the following error message:

ERROR: policy not defined
 in decay_on_validation_listener at /home/ich/.julia/v0.3/Mocha/src/solvers.jl:46
 in anonymous at .../.julia/v0.3/Mocha/src/solvers.jl:78
 in enjoy at .../.julia/v0.3/Mocha/src/coffee/validation-performance.jl:38
 in check_coffee_break at .../.julia/v0.3/Mocha/src/coffee-break.jl:129
 in solve at .../.julia/v0.3/Mocha/src/solvers.jl:356
 in include at /usr/bin/../lib/julia/sys.so
 in include_from_node1 at /usr/bin/../lib/julia/sys.so
while loading ..../test.jl, in expression starting on line 60

on Julia v"0.3.10", Manjaro linux

error message disapears with 0bbfc835e349d5729039a145362b612e3e8ef59c