Quick Start: Alibaba Cloud Function Compute DevPod for Rapid Development

DevPod is a rapid development environment provided by Alibaba Cloud Function Compute that can help you quickly deploy and run the ROLL project.

Prerequisites

Before you begin, ensure you have completed the following preparations:

• You have an Alibaba Cloud account.
• You are logged into the Function Compute FunModel Console.
• Complete the required RAM role authorization and configuration as guided by the console.

Create a Training DevPod Environment

1. Click Create Model - Custom Development.
2. Select Custom Environment and configure as follows:
   • Container Image: Choose a non-public custom image - Container Image Address - roll-registry.cn-hangzhou.cr.aliyuncs.com/roll/pytorch:nvcr-24.05-py3-torch260-vllm084
   • Model Name: Enter a name, e.g., roll-dev
   • Model Source: Select No Model
   • Startup Command: Keep the default, no modification needed
   • Instance Specification: Select GPU Performance Type
   • Click the DevPod Development & Debugging button (Note: Do not click "Create Model Service")
3. Wait for deployment to complete (typically 1–2 minutes).

Configure and Test

Download the ROLL Project and Install Dependencies

# /mnt/ModelName is the default NAS mount point; replace 'roll-dev' with your actual model name
cd /mnt/roll-dev

# 1. Clone the project code
git clone https://github.com/alibaba/ROLL.git  

# 2. Install project dependencies
cd ROLL
pip install -r requirements_torch260_vllm.txt -i https://mirrors.aliyun.com/pypi/simple/  

Run a Pipeline Example

bash examples/agentic_demo/run_agentic_pipeline_frozen_lake_single_node_demo.sh

For more DevPod usage guidelines, see https://fun-model-docs.devsapp.net/user-guide/devpod/