DM05-libero

Introduction

DM05-libero is the LIBERO fine-tuned checkpoint of DM0.5, Dexmal's open-world Vision-Language-Action foundation model for embodied intelligence. DM0.5 uses a Gemma3 4B vision-language backbone with a 680M Action Expert to generate continuous robot actions, and is designed for natural-language manipulation, zero-shot generalization, efficient downstream fine-tuning, long-horizon historical context, robust policy behavior, and transfer across robot embodiments.

LIBERO Results

Method	Spatial	Object	Goal	Long	Average
DM0.5	99.0	99.8	99.6	97.4	99.0

Quick Start

We recommend using Docker to set up the runtime environment first, which helps avoid version mismatches across CUDA, PyTorch, flash-attn, and other dependencies on the host machine.

Requirements

System requirements:
Ubuntu 20.04 / 22.04
NVIDIA GPU
NVIDIA Driver
Docker
NVIDIA Container Toolkit
Conda (optional, only required for local pip installation)

Recommended GPUs:
RTX 4090, A100, H100, H20
8 GPUs are recommended for training, and 1 GPU is sufficient for deployment inference.

Docker Installation

git clone https://github.com/dexmal/opendm.git
cd opendm

docker run -it --rm --gpus all --network host \
  --name opendm \
  --shm-size=16g \
  -v "$PWD":/app/opendm \
  -w /app/opendm \
  dexmal/opendm:latest /bin/bash

# Run from the OpenDM repository root inside the container.
conda activate opendm
pip install -e .

Local Installation

conda create -n opendm python=3.10 -y
conda activate opendm

pip install torch torchvision \
  --index-url https://download.pytorch.org/whl/cu128

pip install ninja packaging
MAX_JOBS=2 pip install flash-attn --no-build-isolation

# Enter the OpenDM repository root.
cd opendm
pip install -e .

LIBERO Testing

Use the LIBERO-specific experiment configuration when evaluating this checkpoint. The inference service and benchmark client run separately and communicate over HTTP. When possible, use one GPU for the inference service and another for benchmark evaluation.

For the complete training and evaluation workflow, see the DM05 LIBERO Training and Evaluation Guide.

Start the LIBERO Inference Service

Run this command from the OpenDM repository root. The checkpoint must contain the matching norm_stats.json.

script/dm05_launcher.sh \
  --exp playground/dm05_libero.py \
  --task inference \
  --nproc_per_node 1 \
  --model-config.model-name-or-path ./checkpoints/DM05-libero-bf16 \
  --model-config.chunk-size 10 \
  --inference-config.output-action-dim 7 \
  --inference-config.image-keys images_1 images_2 \
  --inference-config.port 7891

Keep the service running while executing the benchmark.

Prepare the Benchmark Client

git clone https://github.com/dexmal/dexbotic-benchmark.git
cd dexbotic-benchmark
git submodule update --init --recursive libero
docker pull dexmal/dexbotic_benchmark

Edit evaluation/configs/libero/example_dm05_libero.yaml so that base_url points to the inference service. Select one of libero_spatial, libero_goal, libero_object, or libero_10 as the benchmark:

benchmark: libero_spatial
num_trails_per_task: 50
num_steps_wait: 10
seed: 7

base_url: http://localhost:7891
replan_step: 10

send_state: true
send_image:
  - image
  - wrist_image
discrete_gripper: false
use_text_template: false

output_dir: "results/example_dm05_libero_spatial"

Run the LIBERO Benchmark

From the dexbotic-benchmark repository root, run:

docker run --rm --gpus all --network host \
  -v "$(pwd)":/workspace \
  -w /workspace \
  dexmal/dexbotic_benchmark \
  bash /workspace/scripts/env_sh/libero.sh \
  /workspace/evaluation/configs/libero/example_dm05_libero.yaml

Evaluation artifacts are written under the configured output_dir, including results.json, config.yaml, logs/evaluation.log, and rollout videos under videos/*.mp4.

Community and Support

• Learn more about Dexmal products and model updates on the Dexmal website.
• If you encounter issues, please report them through GitHub Issues.
• For further discussion, scan the WeChat QR code to contact us.

We will continue to release more model weights, technical documentation, and examples. If this project is helpful to you, please consider giving us a star on GitHub . Your support helps us move forward.

Citation

@misc{dm05,
    title  = {{DM0.5}: An Open-World Foundation Model for General-Purpose Embodied Intelligence},
    author = {{Dexmal Team}},
    month  = {July},
    year   = {2026},
    url    = {https://www.dexmal.com/blog/dm0.5/index_en.html}
}