# %% Avoid issuing memory warning due to number of plots
plt.rcParams.update({'figure.max_open_warning': 0})
# %% Load Iris data
iris = XYDataset.from_eeyore('iris',
yndmin=1,
dtype=torch.float32,
yonehot=True)
dataloader = DataLoader(iris, batch_size=len(iris), shuffle=True)
# %% Setup MLP model
hparams = mlp.Hyperparameters(dims=[4, 3, 3],
activations=[torch.sigmoid, None])
model = mlp.MLP(loss=loss_functions['multiclass_classification'],
hparams=hparams,
dtype=torch.float32)
model.prior = Normal(
torch.zeros(model.num_params(), dtype=model.dtype),
(3 * torch.ones(model.num_params(), dtype=model.dtype)).sqrt())
# %% Setup MALA sampler
sampler = MALA(model,
theta0=model.prior.sample(),
dataloader=dataloader,
step=0.003)
# %% Run MALA sampler
# %% Import packages
import numpy as np
import torch
from torch.distributions import Normal
from eeyore.constants import loss_functions
from eeyore.models import mlp
from bnn_mcmc_examples.examples.mlp.pima.setting2.constants import dtype, mlp_dims, mlp_bias, mlp_activations
# %% Setup MLP model
hparams = mlp.Hyperparameters(dims=mlp_dims,
bias=mlp_bias,
activations=mlp_activations)
model = mlp.MLP(loss=loss_functions['binary_classification'],
hparams=hparams,
dtype=dtype)
prior_scale = np.sqrt(10.)
model.prior = Normal(
torch.zeros(model.num_params(), dtype=model.dtype),
torch.full([model.num_params()], prior_scale, dtype=model.dtype))
# %% Import packages
import numpy as np
import torch
from torch.distributions import Normal
from eeyore.constants import loss_functions
from eeyore.models import mlp
from bnn_mcmc_examples.examples.mlp.exact_xor.constants import dtype, mlp_dims
# %% Setup MLP model
hparams = mlp.Hyperparameters(dims=mlp_dims)
model = mlp.MLP(loss=loss_functions['binary_classification'],
hparams=hparams,
dtype=dtype)
prior_scale = np.sqrt(10.)
model.prior = Normal(
torch.zeros(model.num_params(), dtype=model.dtype),
torch.full([model.num_params()], prior_scale, dtype=model.dtype))