Model Card for Gia2

This is a multi-modal and multi-task model.

Model Details

Model Description

• Developed by: The GIA Team
• License: Apache 2.0

Model Sources

• Repository: https://github.com/huggingface/gia
• Paper: Coming soon
• Demo: Coming soon

Training

The model was trained on the following tasks:

• ALE/Freeway-v5
• ALE/Frostbite-v5
• ALE/MsPacman-v5
• ALE/Pong-v5
• ALE/Qbert-v5
• button-press-wall-v2
• drawer-close-v2
• pick-place-wall-v2
• push-back-v2
• soccer-v2
• sweep-into-v2
• window-close-v2
• Ant-v4
• InvertedDoublePendulum-v4
• HalfCheetah-v4
• Hopper-v4
• Humanoid-v4
• InvertedPendulum-v4
• Pusher-v4
• Walker2d-v4

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("gia-project/gia2-small-atari-metaworld-mujoco")