diff --git a/README.md b/README.md
index c363856..0d13a55 100644
--- a/README.md
+++ b/README.md
@@ -52,18 +52,32 @@ ray job submit --address=http://localhost:8265 --working-dir . -- python hello_r
 source ~/pw/activate-ray/ray_venv/bin/activate
 
 # Submit job through the session tunnel (uploads code via --working-dir)
-ray job submit --address=http://localhost:8888 --working-dir . -- python your_script.py
+ray job submit --address=http://localhost:8888 --working-dir . -- python hello_ray.py
 ```
 
 **From your local machine** (via SSH tunnel):
 ```bash
 # Set up tunnel first
-ssh -L 8888:localhost:8888 -o ProxyCommand="pw ssh --proxy-command %h" $USER@workspace
+ssh -i ~/.ssh/pwcli -L 8888:localhost:8888 -o ProxyCommand="pw ssh --proxy-command %h" $USER@workspace
 
 # Then submit jobs (uploads code via --working-dir)
-ray job submit --address=http://localhost:8888 --working-dir . -- python your_script.py
+ray job submit --address=http://localhost:8888 --working-dir . -- python hello_ray.py
 ```
 
+### Installing Additional Packages
+
+The default Ray installation includes only `ray[default]`. For GPU workloads with PyTorch or TensorFlow:
+
+```bash
+# Install PyTorch with CUDA support (uv is pre-installed by the workflow)
+uv pip install --python ~/pw/activate-ray/ray_venv/bin/python torch
+
+# Or install TensorFlow
+uv pip install --python ~/pw/activate-ray/ray_venv/bin/python tensorflow
+```
+
+> **Note:** The `examples/gpu_workload.py` script works without PyTorch but will show "PyTorch not available" messages. Install PyTorch to enable GPU matrix computations in the example.
+
 ### Connect Programmatically
 
 ```python
diff --git a/workflow.yaml b/workflow.yaml
index fbb86b2..9594154 100644
--- a/workflow.yaml
+++ b/workflow.yaml
@@ -188,7 +188,7 @@ jobs:
           echo "Ray Dashboard Session is ready!"
           echo "=========================================="
           echo "  Proxy URL: https://${PW_PLATFORM_HOST}${basepath}/"
-          echo "  Local tunnel: ssh -L ${port}:localhost:${port} -o ProxyCommand=\"pw ssh --proxy-command %h\" ${PW_USER}@workspace"
+          echo "  Local tunnel: ssh -i ~/.ssh/pwcli -L ${port}:localhost:${port} -o ProxyCommand=\"pw ssh --proxy-command %h\" ${PW_USER}@workspace"
           echo "  Then access: http://localhost:${port}/"
           echo "=========================================="