all warnings gone

diffusion_models/__init__.py

@@ -1 +0,0 @@
-version = "0.0.1"

diffusion_models/losses/kl_divergence.py

@@ -10,7 +10,7 @@ def gaussian_kl(
     ) -> Float[Tensor, "1"]:
     """Calculate KL Divergence of 2 Gaussian distributions.
 
-    KL divergence between two univariate Gaussians, as derived in [1], with k=1 (dimensionality).
+    KL divergence between two univariate Gaussians, as derived in [1]_, with k=1 (dimensionality).
 
     Parameters
     ----------
@@ -42,7 +42,7 @@ def log_gaussian_kl(
     ) -> Float[Tensor, "1"]:
     """Calculate KL Divergence of 2 Gaussian distributions.
 
-    KL divergence between two univariate Gaussians, as derived in [1], with k=1 (dimensionality) and log variances.
+    KL divergence between two univariate Gaussians, as derived in [1]_, with k=1 (dimensionality) and log variances.
 
     Parameters
     ----------

diffusion_models/models/diffusion.py

@@ -2,9 +2,9 @@
 from torch import nn, Tensor
 from jaxtyping import Float, Int64, Int
 from typing import Literal, Tuple, Union, List
-from models.positional_encoding import PositionalEncoding
+from diffusion_models.models.positional_encoding import PositionalEncoding
 import math
-from models.unet import UNet
+from diffusion_models.models.unet import UNet
 
 class ForwardDiffusion(nn.Module):
     """Class for forward diffusion process in DDPMs (denoising diffusion probabilistic models).

diffusion_models/models/multicoil.py

@@ -21,7 +21,7 @@ def __init__(self, channel_factors: List[int]=(4, 8, 16, 32), kernel_size: int=3
         This class takes every coil independently (treats them like a sub-fraction of a batch), increases the channel size
         massively (from 2 initial channels for complex k-space data) via several convolutional layers and then averages
         those channels over the coil dimension. Averaging is invariant to permutations of the input order, so the coil order
-        or the number of coils will not matter anymore. Inspiration was drawn from point cloud processing, see below.
+        or the number of coils will not matter anymore. Inspiration was drawn from point cloud processing [1]_, see below.
 
         .. [1] Qi et al., PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation, 2017