{"library":"torchrl","type":"library","category":null,"description":"TorchRL is an open-source, PyTorch-native library for Reinforcement Learning (RL). It provides modular, primitive-first abstractions for building efficient and flexible RL solutions, focusing on research and rapid prototyping. The library offers components for environments, data collection, replay buffers, policy and value networks, and loss functions, all designed to integrate seamlessly with the PyTorch ecosystem. It is currently at version 0.11.1 and follows a regular release cadence, often synced with PyTorch releases.","language":"python","status":"active","version":"0.11.1","tags":["reinforcement-learning","pytorch","machine-learning","deep-learning","rlhf"],"last_verified":"Sat May 23","install":[{"cmd":"pip install torchrl","imports":["from tensordict import TensorDict","from torchrl.envs import GymEnv","from torchrl.modules import MLP","from torchrl.modules import QValueActor","from torchrl.objectives import PPOLoss","from torchrl.collectors import SyncDataCollector"]},{"cmd":"pip install tensordict-nightly torchrl-nightly","imports":[]}],"homepage":null,"github":"https://github.com/pytorch/rl","docs":"https://pytorch.org/rl","changelog":null,"pypi":"https://pypi.org/project/torchrl/","npm":null,"openapi_spec":null,"status_page":null,"smithery":null,"compatibility":{"summary":{"python_range":"3.10–3.9","success_rate":40,"avg_install_s":68,"avg_import_s":11.66,"wheel_type":"wheel"},"url":"https://checklist.day/v1/registry/torchrl/compatibility"}}