Openvla 7b Finetuned Libero Goal

🚀 OpenVLA 7B Fine-Tuned on LIBERO-Goal

This model is fine-tuned from the OpenVLA 7B model on the LIBERO-Goal dataset, aiming to enhance performance in image-text-to-text multimodal tasks for robotics.

This model was produced by fine-tuning the OpenVLA 7B model via LoRA (r = 32) on the LIBERO-Goal dataset from the LIBERO simulation benchmark. We made a few modifications to the training dataset to improve final performance (see the OpenVLA paper for details).

🚀 Quick Start

This section provides an overview of the model and its fine - tuning process. For detailed running and evaluation instructions, refer to the relevant sections below.

✨ Features

• Multimodal Capability: Suitable for image - text - to - text tasks, which is beneficial for robotics applications.
• Fine - Tuned on Specific Dataset: Fine - tuned on the LIBERO - Goal dataset to improve performance in related scenarios.

🔧 Technical Details

Below are the hyperparameters we used for all LIBERO experiments:

• Hardware: 8 x A100 GPUs with 80GB memory
• Fine - tuned with LoRA: use_lora == True, lora_rank == 32, lora_dropout == 0.0
• Learning rate: 5e - 4
• Batch size: 128 (8 GPUs x 16 samples each)
• Number of training gradient steps: 60K
• Quantization: No quantization at train or test time
• Gradient accumulation: No gradient accumulation (i.e., grad_accumulation_steps == 1)
• Shuffle buffer size: shuffle_buffer_size == 100_000
• Image augmentations: Random crop, color jitter (see training code for details)

💻 Usage Examples

Basic Usage

See the OpenVLA GitHub README for instructions on how to run and evaluate this model in the LIBERO simulator.

📄 License

This model is released under the MIT license.

📚 Documentation

Citation

BibTeX:

@article{kim24openvla,
    title={OpenVLA: An Open-Source Vision-Language-Action Model},
    author={{Moo Jin} Kim and Karl Pertsch and Siddharth Karamcheti and Ted Xiao and Ashwin Balakrishna and Suraj Nair and Rafael Rafailov and Ethan Foster and Grace Lam and Pannag Sanketi and Quan Vuong and Thomas Kollar and Benjamin Burchfiel and Russ Tedrake and Dorsa Sadigh and Sergey Levine and Percy Liang and Chelsea Finn},
    journal = {arXiv preprint arXiv:2406.09246},
    year={2024}
} 

Information Table