add README

README.md

@@ -0,0 +1,61 @@
+## Usage
+
+Training a discriminator:
+```
+python -m src params/prec_inn.yaml
+```
+
+Loading a trained model:
+```
+python -m src --load_model 20230303_100000_run_name
+```
+
+Loading a trained model, but compute weights again:
+```
+python -m src --load_model --load_weights 20230303_100000_run_name
+```
+
+## Parameters
+
+### Data loader
+
+Parameter      | Description
+---------------|-------------------------------------------------
+loader\_module | Name of the data loader module in src.loader
+loader\_params | Data-loader specific parameters
+
+### Architecture
+
+Parameter       | Description
+----------------|-----------------------------------------------------------------------------
+layers          | Number of layers
+hidden\_size    | Number of nodes of the hidden layers
+activation      | Activation function. `relu`, `leaky_relu` or `elu`
+dropout         | Dropout fraction
+prior\_prec     | Gaussian prior standard deviation of the Bayesian network
+std\_init       | Logarithm of the initial standard deviation of the Bayesian network weights
+negative\_slope | Negative slope of the leaky ReLU activation
+
+### Training
+
+Parameter         | Description
+------------------|--------------------------------------------------------
+bayesian          | Train as a Bayesian network
+batch\_size       | Batch size
+lr                | Initial learning rate
+betas             | Adam optimizer betas
+eps               | Adam optimizer eps
+weight\_decay     | L2 weight decay
+lr\_scheduler     | Type of LR scheduler: `one_cycle` or `step`
+max\_lr           | One Cycle scheduler: maximum LR
+lr\_decay\_epochs | Step scheduler: Epochs after which to reduce the LR
+lr\_decay\_factor | Step scheduler: Decay factor
+epochs            | Number of epochs
+
+### Evaluation
+
+Parameter         | Description
+------------------|-------------------------------------------------------------
+bayesian\_samples | Number of samples to draw from the network weight posterior
+lower\_thresholds | List of lower weight thresholds for the clustering plots
+upper\_thresholds | List of upper weight thresholds for the clustering plots