{"id":13689263,"url":"https://github.com/pranz24/pytorch-soft-actor-critic","last_synced_at":"2025-05-01T23:33:18.158Z","repository":{"id":37734608,"uuid":"146885631","full_name":"pranz24/pytorch-soft-actor-critic","owner":"pranz24","description":"PyTorch implementation of soft actor critic","archived":false,"fork":false,"pushed_at":"2021-11-09T05:30:52.000Z","size":1009,"stargazers_count":781,"open_issues_count":13,"forks_count":180,"subscribers_count":9,"default_branch":"master","last_synced_at":"2024-08-03T15:16:43.331Z","etag":null,"topics":["deep-reinforcement-learning","pytorch","pytorch-implmention","reinforcement-learning","soft-actor-critic"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/pranz24.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2018-08-31T11:53:00.000Z","updated_at":"2024-07-28T07:11:44.000Z","dependencies_parsed_at":"2022-07-14T09:22:19.173Z","dependency_job_id":null,"html_url":"https://github.com/pranz24/pytorch-soft-actor-critic","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pranz24%2Fpytorch-soft-actor-critic","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pranz24%2Fpytorch-soft-actor-critic/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pranz24%2Fpytorch-soft-actor-critic/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pranz24%2Fpytorch-soft-actor-critic/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/pranz24","download_url":"https://codeload.github.com/pranz24/pytorch-soft-actor-critic/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":224282196,"owners_count":17285786,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["deep-reinforcement-learning","pytorch","pytorch-implmention","reinforcement-learning","soft-actor-critic"],"created_at":"2024-08-02T15:01:40.592Z","updated_at":"2024-11-12T13:31:11.731Z","avatar_url":"https://github.com/pranz24.png","language":"Python","funding_links":[],"categories":["Python"],"sub_categories":[],"readme":"### Description\n------------\nReimplementation of [Soft Actor-Critic Algorithms and Applications](https://arxiv.org/pdf/1812.05905.pdf) and a deterministic variant of SAC from [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement\nLearning with a Stochastic Actor](https://arxiv.org/pdf/1801.01290.pdf).\n\nAdded another branch for [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement\nLearning with a Stochastic Actor](https://arxiv.org/pdf/1801.01290.pdf) -\u003e [SAC_V](https://github.com/pranz24/pytorch-soft-actor-critic/tree/SAC_V).\n\n### Requirements\n------------\n*   [mujoco-py](https://github.com/openai/mujoco-py)\n*   [PyTorch](http://pytorch.org/)\n\n### Default Arguments and Usage\n------------\n### Usage\n\n```\nusage: main.py [-h] [--env-name ENV_NAME] [--policy POLICY] [--eval EVAL]\n               [--gamma G] [--tau G] [--lr G] [--alpha G]\n               [--automatic_entropy_tuning G] [--seed N] [--batch_size N]\n               [--num_steps N] [--hidden_size N] [--updates_per_step N]\n               [--start_steps N] [--target_update_interval N]\n               [--replay_size N] [--cuda]\n```\n\n(Note: There is no need for setting Temperature(`--alpha`) if `--automatic_entropy_tuning` is True.)\n\n#### For SAC\n\n```\npython main.py --env-name Humanoid-v2 --alpha 0.05\n```\n\n#### For SAC (Hard Update)\n\n```\npython main.py --env-name Humanoid-v2 --alpha 0.05 --tau 1 --target_update_interval 1000\n```\n\n#### For SAC (Deterministic, Hard Update)\n\n```\npython main.py --env-name Humanoid-v2 --policy Deterministic --tau 1 --target_update_interval 1000\n```\n\n### Arguments\n------------\n```\nPyTorch Soft Actor-Critic Args\n\noptional arguments:\n  -h, --help            show this help message and exit\n  --env-name ENV_NAME   Mujoco Gym environment (default: HalfCheetah-v2)\n  --policy POLICY       Policy Type: Gaussian | Deterministic (default:\n                        Gaussian)\n  --eval EVAL           Evaluates a policy a policy every 10 episode (default:\n                        True)\n  --gamma G             discount factor for reward (default: 0.99)\n  --tau G               target smoothing coefficient(τ) (default: 5e-3)\n  --lr G                learning rate (default: 3e-4)\n  --alpha G             Temperature parameter α determines the relative\n                        importance of the entropy term against the reward\n                        (default: 0.2)\n  --automatic_entropy_tuning G\n                        Automaically adjust α (default: False)\n  --seed N              random seed (default: 123456)\n  --batch_size N        batch size (default: 256)\n  --num_steps N         maximum number of steps (default: 1e6)\n  --hidden_size N       hidden size (default: 256)\n  --updates_per_step N  model updates per simulator step (default: 1)\n  --start_steps N       Steps sampling random actions (default: 1e4)\n  --target_update_interval N\n                        Value target update per no. of updates per step\n                        (default: 1)\n  --replay_size N       size of replay buffer (default: 1e6)\n  --cuda                run on CUDA (default: False)\n```\n\n| Environment **(`--env-name`)**| Temperature **(`--alpha`)**|\n| ---------------| -------------|\n| HalfCheetah-v2| 0.2|\n| Hopper-v2| 0.2|\n| Walker2d-v2| 0.2|\n| Ant-v2| 0.2|\n| Humanoid-v2| 0.05|\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpranz24%2Fpytorch-soft-actor-critic","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpranz24%2Fpytorch-soft-actor-critic","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpranz24%2Fpytorch-soft-actor-critic/lists"}