fix: misc docs issues

LuxDL · Sep 5, 2024 · 8c964f4 · 8c964f4
1 parent 7b96d5d
commit 8c964f4
Show file tree

Hide file tree

Showing 5 changed files with 14 additions and 8 deletions.
diff --git a/docs/src/api/Lux/utilities.md b/docs/src/api/Lux/utilities.md
@@ -119,3 +119,9 @@ StatefulLuxLayer
 @init_fn
 @non_trainable
 ```
+
+## Miscellaneous
+
+```@docs
+Lux.set_dispatch_doctor_preferences!
+```
diff --git a/docs/src/introduction/updating_to_v1.md b/docs/src/introduction/updating_to_v1.md
@@ -108,7 +108,7 @@ abstraction.
 - `Experimental.StatefulLuxLayer` has been moved to [`Lux.StatefulLuxLayer`](@ref).
 - `st_fixed_path` kwarg has been removed from [`Lux.StatefulLuxLayer`](@ref), instead use it
  as `StatefulLuxLayer{st_fixed_path}(...)`.
-- Strings as inputs to [`Experimental.layer_map`](@ref) and
- [`Experimental.@debug_mode`](@ref) are removed, use `Functors.KeyPath` instead.
+- Strings as inputs to [`Lux.Experimental.layer_map`](@ref) and
+ [`Lux.Experimental.@debug_mode`](@ref) are removed, use `Functors.KeyPath` instead.
 
 ### Breaking Changes (Changes in Defaults)
diff --git a/docs/src/manual/distributed_utils.md b/docs/src/manual/distributed_utils.md
@@ -87,10 +87,10 @@ And that's pretty much it!
  as input.
 3. We don't automatically determine if the MPI Implementation is CUDA or ROCM aware. See
  [GPU-aware MPI](@ref gpu-aware-mpi-preferences) for more information.
-4. Older [`Lux.gpu`](@ref) implementations used to "just work" with `FluxMPI.jl`. We expect
- [`gpu_device`](@ref) to continue working as expected, however, we recommend using
- [`gpu_device`](@ref) after calling [`DistributedUtils.initialize`](@ref) to avoid any
- mismatch between the device set via `DistributedUtils` and the device stores in
+4. Older (now non-existent) `Lux.gpu` implementations used to "just work" with `FluxMPI.jl`.
+ We expect [`gpu_device`](@ref) to continue working as expected, however, we recommend
+ using [`gpu_device`](@ref) after calling [`DistributedUtils.initialize`](@ref) to avoid
+ any mismatch between the device set via `DistributedUtils` and the device stores in
  `CUDADevice` or `AMDGPUDevice`.
 
 ## Known Shortcomings

diff --git a/docs/src/manual/freezing_model_parameters.md b/docs/src/manual/freezing_model_parameters.md
@@ -13,7 +13,7 @@ We can use [`Lux.Experimental.layer_map`](@ref) and freeze layers if they are of
 `Dense`.
 
 ```@example freezing_model_parameters
-using Lux, Functors, Random
+using Lux, Random
 
 rng = Xoshiro(0)
 

diff --git a/test/runtests.jl b/test/runtests.jl
@@ -93,7 +93,7 @@ const RETESTITEMS_NWORKERS = parse(
  @info "Running tests for group: [$(i)/$(length(LUX_TEST_GROUP))] $tag"
 
  ReTestItems.runtests(Lux; tags=(tag == "all" ? nothing : [Symbol(tag)]),
- nworkers=RETESTITEMS_NWORKERS, testitem_timeout=1800, retries=1)
+ nworkers=RETESTITEMS_NWORKERS, testitem_timeout=2400, retries=1)
  end
 end