MethodError: Cannot convert an object of type Params to an object of type Float64

[paritosh5feb]

This is my model, a CNN for Image segmentation. I wrote after an article from this:- https://dev.to/yunshu67/lenet-implementation-for-cell-segmentation-105e

But I am getting an error as mentioned after the code.

`function leNet5()
return Chain(
convBlock(10,10),
MeanPool((5,5)),
convBlock(20,20),
MeanPool((5,5)),
flatten,
Dense(20,40),
Dense(40,80),
Dense(80,100))
end

m = leNet5()
p = params(m)
loss(x, y) = crossentropy(m(x), y)
opt = ADAM(p)
epochs = 10`

Error:-

MethodError: Cannot convert an object of type Params to an object of type Float64
Closest candidates are:
convert(::Type{Float64}, ::LLVM.ConstantFP) at /home/jupyterlab/.julia/packages/LLVM/KITdB/src/core/value/constant.jl:85
convert(::Type{T<:AbstractFloat}, ::Ratios.SimpleRatio{S}) where {T<:AbstractFloat, S} at /home/jupyterlab/.julia/packages/Ratios/uRs4y/src/Ratios.jl:14
convert(::Type{T<:Number}, ::T<:Number) where T<:Number at number.jl:6
...

Stacktrace:
[1] ADAM(::Params, ::Tuple{Float64,Float64}, ::IdDict{Any,Any}) at /home/jupyterlab/.julia/packages/Flux/NpkMm/src/optimise/optimisers.jl:159
[2] ADAM(::Params, ::Tuple{Float64,Float64}) at /home/jupyterlab/.julia/packages/Flux/NpkMm/src/optimise/optimisers.jl:164 (repeats 2 times)
[3] top-level scope at

[CarloLucibello]

Usage questions on the package are better asked on discourse instead of here.
The problem in your example is that you don't have to feed the params to the optimizer.
Please check the documentation.