Acme is a framework from DeepMind for building scalable and reproducible reinforcement learning agents. It emphasizes modular components, distributed training, and ease of experimentation.
Features
- Highly modular RL agent design for flexible research
- Supports distributed training with multiple backends
- Integrates with TensorFlow and JAX for hardware acceleration
- Provides benchmarks and ready-to-use algorithms
- Focuses on reproducibility and scalability
Categories
Reinforcement Learning FrameworksLicense
Apache License V2.0Follow Acme
Other Useful Business Software
Cut Data Warehouse Costs up to 54% with BigQuery
BigQuery delivers up to 54% lower TCO than cloud alternatives. Migrate from legacy or competing warehouses using free BigQuery Migration Service with automated SQL translation. Get serverless scale with no infrastructure to manage, compressed storage, and flexible pricing—pay per query or commit for deeper discounts. New customers get $300 in free credit.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of Acme!