MIOpen kdb installation instructions for PyTorch warmup performance improvement (#2248)

.github/workflows/linting.yml

@@ -5,10 +5,12 @@ on:
     branches: 
     - develop
     - main
+    - 'roc**'    
   pull_request:
     branches: 
     - develop
     - main
+    - 'roc**'
 
 concurrency:
   group: ${{ github.ref }}-${{ github.workflow }}

.wordlist.txt

@@ -27,3 +27,6 @@ DGEMM
 HPCG
 HPL
 IOPM
+# pytorch_install
+kdb
+precompiled

docs/how_to/pytorch_install/pytorch_install.md

@@ -405,6 +405,22 @@ Follow these steps:
    python3 main.py
    ```
 
+## Using MIOpen kdb files with ROCm PyTorch wheels
+
+PyTorch uses MIOpen for machine learning primitives. These primitives are compiled into kernels at runtime. Runtime compilation causes a small warm-up phase when starting PyTorch. MIOpen kdb files contain precompiled kernels that can speed up the warm-up phase of an application. More information is available in the {doc}`MIOpeninstallation page <miopen:install>`.
+
+MIOpen kdb files can be used with ROCm PyTorch wheels. However, the kdb files need to be placed in a specific location with respect to the PyTorch installation path. A helper script simplifies this task for the user. The script takes in the ROCm version and user's GPU architecture as inputs, and works for Ubuntu and CentOS.
+
+Helper script: [install_kdb_files_for_pytorch_wheels.sh](https://raw.githubusercontent.com/wiki/ROCmSoftwarePlatform/pytorch/files/install_kdb_files_for_pytorch_wheels.sh)
+
+Usage:
+
+After installing ROCm PyTorch wheels:
+
+1. [Optional] `export GFX_ARCH=gfx90a`
+2. [Optional] `export ROCM_VERSION=5.5`
+3. `./install_kdb_files_for_pytorch_wheels.sh`
+
 ## References
 
 C. Szegedy, V. Vanhoucke, S. Ioffe, J. Shlens and Z. Wojna, "Rethinking the Inception Architecture for Computer Vision," CoRR, p. abs/1512.00567, 2015