Modelling Implementation

[guillaume-byte]

Model modelling is implemented and available; however:

• No direct render integration in the UI, only through agent query or Python functions (model.add_neurons(...)
• Test and fix model graph generation (currently supports Torch and ONNX but has a limited scope)
• Refactor modelling code Refactor index mapping in neuron operations #6 Refactor indices operation functions + - into one function, operating on indices #7 Fix hardcoded code for norms & transposed conv & dependency flag & shape conv. #8

TODO:

• Implement UI modelling of the model graph (model rendering, layers and neurons actions (freeze, reset, add, prune)
• Neurons evolution follow-up in the graph
• WeightsLab modelling code review