Skip to content

Conversation

@fliingelephant
Copy link
Contributor

Try fixing #98

  • Still FAIL to pass the "belief propagation on circle" test (ComplexF64 version), where BP fails to converge.
  • There are four parts involving cost_and_gradient and I only edited belief.jl, mar.jl and leave those in map.jl and mmap.jl remained the same.
  • I am not sure if it is correct to conjugate tensors when the datatype is neither Real nor Complex.

@fliingelephant fliingelephant marked this pull request as draft July 2, 2025 02:31
Copy link
Member

@GiggleLiu GiggleLiu left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There is an extra gauge degree of freedom for complex numbers, so the message can be different by a phase.

n = 5
chi = 3
mps_uai = TensorInference.random_tensor_train_uai(Float64, n, chi)
mps_uai = TensorInference.random_tensor_train_uai(ComplexF64, n, chi)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please test both Float64 and ComplexF64.

Signed-off-by: 周唤海 <[email protected]>
@GiggleLiu
Copy link
Member

GiggleLiu commented Jul 2, 2025

close in favor of #100

@GiggleLiu GiggleLiu closed this Jul 2, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants