ROLL x Ascend

Last updated: 11/25/2025.

We have added support for Huawei Ascend devices in ROLL.

Hardware Support

Atlas 900 A2 PODc

Installation

Basic Environment Setup

Software	Version
Python	3.11
CANN	8.3.RC1

Create Conda Environment

Use the following commands to create a new conda environment in Miniconda:

conda create --name roll python=3.11
conda activate roll

Install torch & torch_npu

To use torch and torch_npu in ROLL, install them using the commands below:

# Use CPU only torch
pip install torch==2.7.1 torchvision==0.22.1 torchaudio==2.7.1 --index-url https://download.pytorch.org/whl/cpu

# Install torch_npu 2.7.1
pip install torch_npu==2.7.1

Install vllm & vllm-ascend

To use vllm in ROLL, compile and install vllm and vllm-ascend as follows:

# vllm
git clone -b v0.11.0 --depth 1 https://github.com/vllm-project/vllm.git
cd vllm
pip install -r requirements/build.txt

VLLM_TARGET_DEVICE=empty pip install -v -e .
cd ..

# vllm-ascend
git clone -b v0.11.0rc1 --depth 1 https://github.com/vllm-project/vllm-ascend.git
cd vllm-ascend

pip install -e .
cd ..

Or you could install vllm and vllm-ascend from pre-built wheel:

# Install vllm-project/vllm. The newest supported version is v0.11.0.
pip install vllm==0.11.0

# Install vllm-project/vllm-ascend from pypi.
pip install vllm-ascend==0.11.0rc1

Install ROLL

git clone https://github.com/alibaba/ROLL.git
cd ROLL
pip install -r requirements_common.txt
pip install deepspeed==0.16.4
cd ..

Additional Third-Party Libraries

Software	Description
transformers	>= v4.57.1
flash_attn	not supported
transformer-engine[pytorch]	not supported

1. transformers v4.57.1 supports enabling --flash_attention_2.
2. flash_attn acceleration is not supported currently.
3. transformer-engine[pytorch] is currently not supported.

pip install transformers==4.57.1

Quick Start: Single-Node Deployment

Before full usage, we recommend testing the single-node pipeline to verify your environment and installation. Since Megatron-LM training is not yet supported, first change strategy_args in the relevant files to use the deepspeed option.

1. Run the single-node pipeline via shell:

bash examples/agentic_demo/run_agentic_pipeline_frozen_lake_single_node_demo.sh  

2. Run the agentic pipeline using a config file:

# Make sure you are in the root directory of the ROLL project

python examples/start_agentic_pipeline.py \
        --config_path qwen2.5-0.5B-agentic \
        --config_name agentic_val_sokoban

• --config_path – Directory containing your YAML configuration files.
• --config_name – Filename (without the .yaml extension).

Current Support Status

Feature	Example	Training Backend	Inference Backend	Hardware
Agentic	examples/qwen2.5-0.5B-agentic/run_agentic_pipeline_sokoban.sh	DeepSpeed	vLLM	Atlas 900 A2 PODc
Agentic-Rollout	examples/qwen2.5-0.5B-agentic/run_agentic_rollout_sokoban.sh	DeepSpeed	vLLM	Atlas 900 A2 PODc
DPO	examples/qwen2.5-3B-dpo_megatron/run_dpo_pipeline.sh	DeepSpeed	vLLM	Atlas 900 A2 PODc
RLVR	examples/qwen2.5-7B-rlvr_megatron/run_rlvr_pipeline.sh	DeepSpeed	vLLM	Atlas 900 A2 PODc

Disclaimer

The Ascend support provided in ROLL is intended as a reference example. For production use, please consult official channels.