Knowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912461452476416 |
|---|---|
| author | Kuzmenko, Dmytro Shvai, Nadiya |
| author_facet | Kuzmenko, Dmytro Shvai, Nadiya |
| contents | We propose an efficient knowledge transfer approach for model-based reinforcement learning, addressing the challenge of deploying large world models in resource-constrained environments. Our method distills a high-capacity multi-task agent (317M parameters) into a compact 1M parameter model, achieving state-of-the-art performance on the MT30 benchmark with a normalized score of 28.45, a substantial improvement over the original 1M parameter model's score of 18.93. This demonstrates the ability of our distillation technique to consolidate complex multi-task knowledge effectively. Additionally, we apply FP16 post-training quantization, reducing the model size by 50% while maintaining performance. Our work bridges the gap between the power of large models and practical deployment constraints, offering a scalable solution for efficient and accessible multi-task reinforcement learning in robotics and other resource-limited domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_05329 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Knowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning Kuzmenko, Dmytro Shvai, Nadiya Machine Learning Robotics I.2.9; I.2.11 We propose an efficient knowledge transfer approach for model-based reinforcement learning, addressing the challenge of deploying large world models in resource-constrained environments. Our method distills a high-capacity multi-task agent (317M parameters) into a compact 1M parameter model, achieving state-of-the-art performance on the MT30 benchmark with a normalized score of 28.45, a substantial improvement over the original 1M parameter model's score of 18.93. This demonstrates the ability of our distillation technique to consolidate complex multi-task knowledge effectively. Additionally, we apply FP16 post-training quantization, reducing the model size by 50% while maintaining performance. Our work bridges the gap between the power of large models and practical deployment constraints, offering a scalable solution for efficient and accessible multi-task reinforcement learning in robotics and other resource-limited domains. |
| title | Knowledge Transfer in Model-Based Reinforcement Learning Agents for Efficient Multi-Task Learning |
| topic | Machine Learning Robotics I.2.9; I.2.11 |
| url | https://arxiv.org/abs/2501.05329 |