diff --git a/NEWS.md b/NEWS.md
index bafe4ad5..09962b16 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -25,6 +25,7 @@
 * feat: `PipeOpBlock` should no longer create ID clashes with other PipeOps in the graph (#260)
 * fix: `device` is no longer part of the `dataset` which allows for parallel dataloading
   on GPUs.
+* feat: The `shuffle` parameter during model training is now initialized to `TRUE`.
 
 # mlr3torch 0.1.2
 
diff --git a/R/paramset_torchlearner.R b/R/paramset_torchlearner.R
index 098743dd..0af5d74d 100644
--- a/R/paramset_torchlearner.R
+++ b/R/paramset_torchlearner.R
@@ -69,7 +69,7 @@ paramset_torchlearner = function(task_type) {
     min_delta             = p_dbl(lower = 0, tags = c("train", "required"), init = 0),
     # dataloader parameters
     batch_size            = p_int(tags = c("train", "predict", "required"), lower = 1L),
-    shuffle               = p_lgl(tags = "train", default = FALSE),
+    shuffle               = p_lgl(tags = "train", default = FALSE, init = TRUE),
     sampler               = p_uty(tags = c("train", "predict")),
     batch_sampler         = p_uty(tags = c("train", "predict")),
     num_workers           = p_int(lower = 0, default = 0, tags = c("train", "predict")),
diff --git a/man-roxygen/paramset_torchlearner.R b/man-roxygen/paramset_torchlearner.R
index 20a0ec04..46e62e5c 100644
--- a/man-roxygen/paramset_torchlearner.R
+++ b/man-roxygen/paramset_torchlearner.R
@@ -63,8 +63,8 @@
 #' * `batch_size` :: `integer(1)`\cr
 #'   The batch size (required).
 #' * `shuffle` :: `logical(1)`\cr
-#'   Whether to shuffle the instances in the dataset. Default is `FALSE`.
-#'   This does not impact validation.
+#'   Whether to shuffle the instances in the dataset. This is initialized to `TRUE`,
+#'   which differs from the default (`FALSE`).
 #' * `sampler` :: [`torch::sampler`]\cr
 #'   Object that defines how the dataloader draw samples.
 #' * `batch_sampler` :: [`torch::sampler`]\cr
diff --git a/man/mlr_learners_torch.Rd b/man/mlr_learners_torch.Rd
index 303f0760..da3f05b6 100644
--- a/man/mlr_learners_torch.Rd
+++ b/man/mlr_learners_torch.Rd
@@ -125,8 +125,8 @@ Is initialized to 0.
 \item \code{batch_size} :: \code{integer(1)}\cr
 The batch size (required).
 \item \code{shuffle} :: \code{logical(1)}\cr
-Whether to shuffle the instances in the dataset. Default is \code{FALSE}.
-This does not impact validation.
+Whether to shuffle the instances in the dataset. This is initialized to \code{TRUE},
+which differs from the default (\code{FALSE}).
 \item \code{sampler} :: \code{\link[torch:sampler]{torch::sampler}}\cr
 Object that defines how the dataloader draw samples.
 \item \code{batch_sampler} :: \code{\link[torch:sampler]{torch::sampler}}\cr
diff --git a/man/mlr_pipeops_torch_model.Rd b/man/mlr_pipeops_torch_model.Rd
index 7b7438d8..2e5875be 100644
--- a/man/mlr_pipeops_torch_model.Rd
+++ b/man/mlr_pipeops_torch_model.Rd
@@ -80,8 +80,8 @@ Is initialized to 0.
 \item \code{batch_size} :: \code{integer(1)}\cr
 The batch size (required).
 \item \code{shuffle} :: \code{logical(1)}\cr
-Whether to shuffle the instances in the dataset. Default is \code{FALSE}.
-This does not impact validation.
+Whether to shuffle the instances in the dataset. This is initialized to \code{TRUE},
+which differs from the default (\code{FALSE}).
 \item \code{sampler} :: \code{\link[torch:sampler]{torch::sampler}}\cr
 Object that defines how the dataloader draw samples.
 \item \code{batch_sampler} :: \code{\link[torch:sampler]{torch::sampler}}\cr