diff --git a/R/models-resnet.R b/R/models-resnet.R
index 14b161a..11c7863 100644
--- a/R/models-resnet.R
+++ b/R/models-resnet.R
@@ -252,7 +252,7 @@ resnet <- torch::nn_module(
 #' ResNet implementation
 #'
 #' ResNet models implementation from
-#'   [Deep Residual Learning for Image Recognition](https://arxiv.org/pdf/1512.03385.pdf) and later
+#'   [Deep Residual Learning for Image Recognition](https://arxiv.org/pdf/1512.03385) and later
 #'   related papers (see Functions)
 #'
 #' @param pretrained (bool): If TRUE, returns a model pre-trained on ImageNet.
@@ -295,28 +295,28 @@ model_resnet152 <- function(pretrained = FALSE, progress = TRUE, ...) {
   .resnet("resnet152", bottleneck, c(3,8,36,3), pretrained, progress, ...)
 }
 
-#' @describeIn model_resnet ResNeXt-50 32x4d model from ["Aggregated Residual Transformation for Deep Neural Networks"](https://arxiv.org/pdf/1611.05431.pdf)
+#' @describeIn model_resnet ResNeXt-50 32x4d model from ["Aggregated Residual Transformation for Deep Neural Networks"](https://arxiv.org/pdf/1611.05431)
 #' with 32 groups having each a width of 4.
 #' @export
 model_resnext50_32x4d <- function(pretrained = FALSE, progress = TRUE, ...) {
   .resnet("resnext50_32x4d", bottleneck, c(3,4,6,3), pretrained, progress, groups=32, width_per_group=4,...)
 }
 
-#' @describeIn model_resnet ResNeXt-101 32x8d model from ["Aggregated Residual Transformation for Deep Neural Networks"](https://arxiv.org/pdf/1611.05431.pdf)
+#' @describeIn model_resnet ResNeXt-101 32x8d model from ["Aggregated Residual Transformation for Deep Neural Networks"](https://arxiv.org/pdf/1611.05431)
 #' with 32 groups having each a width of 8.
 #' @export
 model_resnext101_32x8d <- function(pretrained = FALSE, progress = TRUE, ...) {
   .resnet("resnext101_32x8d", bottleneck, c(3,4,23,3), pretrained, progress, groups=32, width_per_group=8,...)
 }
 
-#' @describeIn model_resnet Wide ResNet-50-2 model from ["Wide Residual Networks"](https://arxiv.org/pdf/1605.07146.pdf)
+#' @describeIn model_resnet Wide ResNet-50-2 model from ["Wide Residual Networks"](https://arxiv.org/pdf/1605.07146)
 #' with width per group of 128.
 #' @export
 model_wide_resnet50_2 <- function(pretrained = FALSE, progress = TRUE, ...) {
   .resnet("wide_resnet50_2", bottleneck, c(3,4,6,3), pretrained, progress, width_per_group=64*2,...)
 }
 
-#' @describeIn model_resnet Wide ResNet-101-2 model from ["Wide Residual Networks"](https://arxiv.org/pdf/1605.07146.pdf)
+#' @describeIn model_resnet Wide ResNet-101-2 model from ["Wide Residual Networks"](https://arxiv.org/pdf/1605.07146)
 #' with width per group of 128.
 #' @export
 model_wide_resnet101_2 <- function(pretrained = FALSE, progress = TRUE, ...) {
diff --git a/R/models-vgg.R b/R/models-vgg.R
index e9f81ff..2bac8bc 100644
--- a/R/models-vgg.R
+++ b/R/models-vgg.R
@@ -121,7 +121,7 @@ vgg <- function(arch, cfg, batch_norm, pretrained, progress, ...) {
 #'
 #'
 #' VGG models implementations based on
-#' [Very Deep Convolutional Networks For Large-Scale Image Recognition](https://arxiv.org/pdf/1409.1556.pdf)
+#' [Very Deep Convolutional Networks For Large-Scale Image Recognition](https://arxiv.org/pdf/1409.1556)
 #'
 #' @param pretrained (bool): If TRUE, returns a model pre-trained on ImageNet
 #' @param progress (bool): If TRUE, displays a progress bar of the download
diff --git a/R/transforms-generics.R b/R/transforms-generics.R
index 803504c..9a472e3 100644
--- a/R/transforms-generics.R
+++ b/R/transforms-generics.R
@@ -431,7 +431,7 @@ transform_random_perspective <- function(img, distortion_scale=0.5, p=0.5,
 #' Randomly selects a rectangular region in an image and erases its pixel values
 #'
 #' 'Random Erasing Data Augmentation' by Zhong _et al._
-#'   See <https://arxiv.org/pdf/1708.04896.pdf>
+#'   See <https://arxiv.org/pdf/1708.04896>
 #'
 #' @inheritParams transform_to_tensor
 #' @param p probability that the random erasing operation will be performed.
diff --git a/man/model_resnet.Rd b/man/model_resnet.Rd
index deff51d..0dac9e4 100644
--- a/man/model_resnet.Rd
+++ b/man/model_resnet.Rd
@@ -41,7 +41,7 @@ stderr.}
 }
 \description{
 ResNet models implementation from
-\href{https://arxiv.org/pdf/1512.03385.pdf}{Deep Residual Learning for Image Recognition} and later
+\href{https://arxiv.org/pdf/1512.03385}{Deep Residual Learning for Image Recognition} and later
 related papers (see Functions)
 }
 \section{Functions}{
@@ -56,16 +56,16 @@ related papers (see Functions)
 
 \item \code{model_resnet152()}: ResNet 152-layer model
 
-\item \code{model_resnext50_32x4d()}: ResNeXt-50 32x4d model from \href{https://arxiv.org/pdf/1611.05431.pdf}{"Aggregated Residual Transformation for Deep Neural Networks"}
+\item \code{model_resnext50_32x4d()}: ResNeXt-50 32x4d model from \href{https://arxiv.org/pdf/1611.05431}{"Aggregated Residual Transformation for Deep Neural Networks"}
 with 32 groups having each a width of 4.
 
-\item \code{model_resnext101_32x8d()}: ResNeXt-101 32x8d model from \href{https://arxiv.org/pdf/1611.05431.pdf}{"Aggregated Residual Transformation for Deep Neural Networks"}
+\item \code{model_resnext101_32x8d()}: ResNeXt-101 32x8d model from \href{https://arxiv.org/pdf/1611.05431}{"Aggregated Residual Transformation for Deep Neural Networks"}
 with 32 groups having each a width of 8.
 
-\item \code{model_wide_resnet50_2()}: Wide ResNet-50-2 model from \href{https://arxiv.org/pdf/1605.07146.pdf}{"Wide Residual Networks"}
+\item \code{model_wide_resnet50_2()}: Wide ResNet-50-2 model from \href{https://arxiv.org/pdf/1605.07146}{"Wide Residual Networks"}
 with width per group of 128.
 
-\item \code{model_wide_resnet101_2()}: Wide ResNet-101-2 model from \href{https://arxiv.org/pdf/1605.07146.pdf}{"Wide Residual Networks"}
+\item \code{model_wide_resnet101_2()}: Wide ResNet-101-2 model from \href{https://arxiv.org/pdf/1605.07146}{"Wide Residual Networks"}
 with width per group of 128.
 
 }}
diff --git a/man/model_vgg.Rd b/man/model_vgg.Rd
index 163bee5..77900bf 100644
--- a/man/model_vgg.Rd
+++ b/man/model_vgg.Rd
@@ -38,7 +38,7 @@ to stderr}
 }
 \description{
 VGG models implementations based on
-\href{https://arxiv.org/pdf/1409.1556.pdf}{Very Deep Convolutional Networks For Large-Scale Image Recognition}
+\href{https://arxiv.org/pdf/1409.1556}{Very Deep Convolutional Networks For Large-Scale Image Recognition}
 }
 \section{Functions}{
 \itemize{
diff --git a/man/transform_random_erasing.Rd b/man/transform_random_erasing.Rd
index 20f63ae..74e6dff 100644
--- a/man/transform_random_erasing.Rd
+++ b/man/transform_random_erasing.Rd
@@ -31,7 +31,7 @@ If a str of 'random', erasing each pixel with random values.}
 }
 \description{
 'Random Erasing Data Augmentation' by Zhong \emph{et al.}
-See \url{https://arxiv.org/pdf/1708.04896.pdf}
+See \url{https://arxiv.org/pdf/1708.04896}
 }
 \seealso{
 Other transforms: