{"id":16472516,"url":"https://github.com/nhynes/abc","last_synced_at":"2025-03-23T11:32:47.875Z","repository":{"id":71680018,"uuid":"110458641","full_name":"nhynes/abc","owner":"nhynes","description":"SeqGAN but with more bells and whistles","archived":false,"fork":false,"pushed_at":"2018-02-15T02:57:00.000Z","size":133,"stargazers_count":24,"open_issues_count":0,"forks_count":7,"subscribers_count":4,"default_branch":"master","last_synced_at":"2025-03-18T20:07:22.703Z","etag":null,"topics":["deep-reinforcement-learning","gan","pytorch","seqgan"],"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/nhynes.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE.txt","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2017-11-12T18:39:13.000Z","updated_at":"2024-04-25T01:51:37.000Z","dependencies_parsed_at":"2023-06-01T19:45:56.276Z","dependency_job_id":null,"html_url":"https://github.com/nhynes/abc","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/nhynes%2Fabc","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nhynes%2Fabc/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nhynes%2Fabc/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/nhynes%2Fabc/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/nhynes","download_url":"https://codeload.github.com/nhynes/abc/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":245097158,"owners_count":20560311,"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","gan","pytorch","seqgan"],"created_at":"2024-10-11T12:17:25.211Z","updated_at":"2025-03-23T11:32:47.863Z","avatar_url":"https://github.com/nhynes.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Adversarial Behavioral Cloning\n\nImproves on the [SeqGAN](https://aaai.org/ocs/index.php/AAAI/AAAI17/paper/view/14344)\nidea by adding more reinfocement learning and GAN techniques like:\n* a replay buffer\n* [Consensus Optimization](https://arxiv.org/abs/1705.10461)\n* [count-based exploration bonus](https://arxiv.org/abs/1611.04717)\n* Proximal Policy Optimization (was not found to help, but can be found in\n  [this commit](https://github.com/nhynes/abc/commit/af7d921ab96e20ba75a60558e1e293b8667b4480))\n* advantage normalization\n\n\n## How to run\n\nIf you wish to enable Consensus Optimization (via the `--grad-reg` option), you'll need\nto [patch](scripts/dx2_rnn.patch) PyTorch to allow forcing the use a\ntwice-differentiable RNN.\n\n`python3 main.py` will run the project with the default options.\nOutput will be written to the `run/` directory.\n\n## Shameless plug\n\nThe [em](https://github.com/nhynes/em) tool makes it really easy to twiddle\nhyperparameters by tracking changes to code (no need to make everything an option!).\n\nJust run `em run -g 0 exp_name` with your desired options and you'll find a reproducable\nsnapshot in `experiments/\u003cexp_name\u003e`!\n\nIf you want to resume from a snapshot (perhaps with different options),\nuse `em resume -g 0 exp_name ...`\n\nYou can also fork an experiment and its changes using `em fork`, but the quick and dirty\nsolution is to run `bash scripts/make_links.sh` :)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnhynes%2Fabc","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnhynes%2Fabc","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnhynes%2Fabc/lists"}