{"id":23666523,"url":"https://github.com/hugcis/benchmark_learning_efficiency","last_synced_at":"2025-07-19T07:05:22.664Z","repository":{"id":45337918,"uuid":"429761336","full_name":"hugcis/benchmark_learning_efficiency","owner":"hugcis","description":"Code to reproduce results from \"Benchmarking learning efficiency in deep reservoir computing\"","archived":false,"fork":false,"pushed_at":"2022-10-27T07:22:57.000Z","size":2132,"stargazers_count":2,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2024-12-29T07:32:55.011Z","etag":null,"topics":["learning","research-paper","reservoir-computing"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","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/hugcis.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":"2021-11-19T10:44:43.000Z","updated_at":"2023-07-19T12:53:23.000Z","dependencies_parsed_at":"2023-01-20T16:15:44.100Z","dependency_job_id":null,"html_url":"https://github.com/hugcis/benchmark_learning_efficiency","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/hugcis%2Fbenchmark_learning_efficiency","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hugcis%2Fbenchmark_learning_efficiency/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hugcis%2Fbenchmark_learning_efficiency/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/hugcis%2Fbenchmark_learning_efficiency/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/hugcis","download_url":"https://codeload.github.com/hugcis/benchmark_learning_efficiency/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":239674771,"owners_count":19678468,"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":["learning","research-paper","reservoir-computing"],"created_at":"2024-12-29T07:32:40.954Z","updated_at":"2025-02-19T14:27:33.952Z","avatar_url":"https://github.com/hugcis.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Benchmarking learning efficiency in deep reservoir computing\n\nThis is the code to reproduce results from the paper \n\n*Benchmarking Learning Efficiency in Deep Reservoir Computing. Cisneros, H.,\nMikolov, T., \u0026 Sivic, J. (2022). 1st Conference on Lifelong Learning Agents,\nMontreal, Canada*.\n\n## Re-run experiments\n\n**WARNING**: Re-running all experiments might take a significant amount of time.\nExperiments in the paper were done on a cluster using GPUs and a lot of\nparallelism. The docker solution is particularly sub-optimal and will take a\nlong time to run experiments.\n\nAn alternative to running all the experiments is to download the data directly: \n``` sh\nwget https://data.ciirc.cvut.cz/public/projects/2022BenchmarkingLearningEfficiency/experiment_2022-07-13T15:32:50.tar\n\ntar -xvf \"experiment_2022-07-13T15:32:50.tar\"\n```\n\n### Running with poetry\n\nThe easiest way to run the experiments is to use\n[poetry](https://python-poetry.org/). First, clone the repo \n \n```sh\ngit clone https://github.com/hugcis/benchmark_learning_efficiency.git\n```\n\nThen, run `poetry install` to create a virtual environment and install all\nthe dependencies.\n\nThen run: \n```sh\n./run_experiments.sh\n```\n\n### Running in Docker\n\nIf you don't have or don't want to install poetry, you can build and install\neverything within a docker container. Just run the following from inside the\nrepo:\n\n``` sh\ndocker build -t pypoetry_bledrc .\ndocker run -it --entrypoint=/bin/bash pypoetry_bledrc -i\n```\n\nThis will open a bash tty within the docker container where you can run \n```sh\n./run_experiments.sh\n```\n\n## Generate figures and tables\n\nOnce the data is generated or downloaded (make sure that you have the\nexperiment_gpu and experiment_sgd folders), you can run jupyter notebooks in\norder to re-generate the figures and tables from the paper.\n\nJust run\n``` sh\npoetry run jupyter notebook\n```\nand open the two jupyter notebooks in the folder `notebooks`.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhugcis%2Fbenchmark_learning_efficiency","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhugcis%2Fbenchmark_learning_efficiency","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhugcis%2Fbenchmark_learning_efficiency/lists"}