where may I find code for ML?
@AUdaltsova
@peterdudfield
Dear Team,
I hope you're doing well!
I’ve been reviewing your Python script from the project "Forecasting Regional PV Power in Great Britain With a Multi-Modal Late Fusion Network" as I explore similar studies and methods.
Regarding the deep learning architecture, I’m particularly interested in replicating the approach described in your paper.
However, while browsing through your code in the PVNet repository, I wasn’t able to locate the specific parts where the architectures are defined, such as:
x = layers.Flatten()(x)
x = layers.Dense(256, activation='elu')(x)
x = layers.Dense(256, activation='elu')(x)
Could you please point me to where these architectures are implemented or compiled? I’ve been looking through the PVNet repository, particularly here:
https://github.com/openclimatefix/PVNet/tree/main/pvnet/models/multimodal, but haven’t been able to find the exact code.
Thank you in advance for your help!
Hi @kwon-encored ,
So a good way to find where these components of the models are is to look at the model config files on HF model repos such as this: https://huggingface.co/openclimatefix/pvnet_uk_region/blob/ae0b8006841ac6227db873a1fc7f7331dc7dadb5/config.json which outline the config of the models themselves, the"_target_": parameters in those fields have values which are essentially import paths in the PVNet repository, so for example you can find one of the convolutional encoders for the NWPS here and one of the linear networks for the output network here
Hope that answers your question!