diff --git a/docs/usage.md b/docs/usage.md
index c073e3a..1bc5066 100644
--- a/docs/usage.md
+++ b/docs/usage.md
@@ -9,7 +9,7 @@ This document describes the parameter options used by the pipeline.
     * [`--bed`](#--bed-arg-required)
     * [`--output`](#--output-arg-required)
 * [Workflow modifying parameters](#workflow-modifying-parameters)
-    * [`--subsampling`](#--subsampling)
+    * [`--subsampling`](#--subsampling)https://github.com/genepi/umi-pipeline-nf/blob/medaka_GPU/docs/usage.md
     * [`--call_variants`](#--call_variants)
 * [Read filtering parameters](#read-filtering-parameters)
     * [`--min_read_length`](#--min_read_length-arg)
@@ -158,14 +158,14 @@ Specify if secondary mappings should be included in the analysis. [default: fals
 ### `--balance_strands <ARG>`
 Specify if the number of forward and reverse reads per cluster should be equalized. [default: true]
 
-### `--use_gpu`
+### `--use_gpu <ARG>`
 Specify if GPU acceleration for Medaka should be enabled. [default: false]  
 Note: GPU acceleration is only available with the `docker` profile and has been tested with an RTX 4080 SUPER GPU (16 GB).
-
-### `--gpu_driver`
-Specify the path to the GPU driver.  
 The NVIDIA toolkit (https://developer.nvidia.com/cuda-toolkit) and NVIDIA container toolkit (https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html) must be installed for GPU usage.
 
+### `--gpu_memory <ARG>`
+Specify the available GPU memory in GB. [default: 16]  
+
 ### `--medaka_model <ARG>`
 Specify the medaka model that is used for cluster polishing. [default: "r1041_e82_400bps_hac_g615"]  
 Note: The models are specific for Chemistry, basecalling algorithm and sequencing speed.