{"id":34066804,"url":"https://github.com/puffinsoft/benchstreet","last_synced_at":"2026-04-07T06:32:58.246Z","repository":{"id":305221100,"uuid":"1021870502","full_name":"puffinsoft/benchstreet","owner":"puffinsoft","description":"The benchmark for financial time series forecasting.","archived":false,"fork":false,"pushed_at":"2025-07-19T00:23:27.000Z","size":3686,"stargazers_count":14,"open_issues_count":0,"forks_count":1,"subscribers_count":0,"default_branch":"master","last_synced_at":"2026-01-03T06:48:39.455Z","etag":null,"topics":["machine-learning","time-series"],"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/puffinsoft.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,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2025-07-18T04:48:11.000Z","updated_at":"2025-10-23T07:15:58.000Z","dependencies_parsed_at":"2025-07-19T00:40:52.359Z","dependency_job_id":"abe1d861-cf93-4892-b159-b5be244f713a","html_url":"https://github.com/puffinsoft/benchstreet","commit_stats":null,"previous_names":["puffinsoft/benchstreet"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/puffinsoft/benchstreet","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/puffinsoft%2Fbenchstreet","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/puffinsoft%2Fbenchstreet/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/puffinsoft%2Fbenchstreet/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/puffinsoft%2Fbenchstreet/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/puffinsoft","download_url":"https://codeload.github.com/puffinsoft/benchstreet/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/puffinsoft%2Fbenchstreet/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31503382,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-07T03:10:19.677Z","status":"ssl_error","status_checked_at":"2026-04-07T03:10:13.982Z","response_time":105,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.5:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"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":["machine-learning","time-series"],"created_at":"2025-12-14T06:50:17.376Z","updated_at":"2026-04-07T06:32:58.225Z","avatar_url":"https://github.com/puffinsoft.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cimg src=\"docs/images/logo.png\" width=\"600\"\u003e\n\n\u003chr/\u003e\n\n![](https://img.shields.io/badge/Keras-FF0000?style=for-the-badge\u0026logo=keras\u0026logoColor=white) ![](https://img.shields.io/badge/PyTorch-EE4C2C?style=for-the-badge\u0026logo=pytorch\u0026logoColor=white) ![](https://img.shields.io/badge/TensorFlow-FF6F00?style=for-the-badge\u0026logo=tensorflow\u0026logoColor=white) ![](https://img.shields.io/badge/Weights_\u0026_Biases-FFBE00?style=for-the-badge\u0026logo=WeightsAndBiases\u0026logoColor=white) ![](https://img.shields.io/badge/-HuggingFace-FDEE21?style=for-the-badge\u0026logo=HuggingFace\u0026logoColor=black)\n\n**Benchstreet** is a curated collection of time series prediction models designed to help developers evaluate and compare the performance of different approaches in **one-shot**, **long-term** financial data forecasting.\n\nThe models are trained on 20 years of S\u0026P 500 daily closing prices provided by [Investing.com](https://investing.com/indices/us-spx-500-historical-data).\n\n\u003e [!IMPORTANT]\n\u003e This is not an objective benchmark! It's intended as a qualitative guide and a reference on how to implement these models.\n\n![](docs/images/diagram.png)\n\n\u003chr /\u003e\n\n### Table of Contents\n\n| Model Type                           |                                                       |\n|--------------------------------------|-------------------------------------------------------|\n| Transformer/Foundation Models        | TimesFM ([baseline](#timesfm-baseline) • [**fine-tuned**](#timesfm-fine-tuned)), Chronos ([baseline](#chronos-baseline) • [**fine-tuned**](#chronos-fine-tuned)) |\n| Feedforward Neural Networks (FNNs)   | MLP ([recursive](#mlp-recursive) • [vector](#mlp-vector)), N-BEATS ([direct](#n-beats)) |\n| Convolutional Neural Networks (CNNs) | 1D-CNN ([recursive](#1d-cnn-recursive) • [vector](#1d-cnn-vector)), TemporalCN ([vector](#temporalcn)) |\n| Recurrent Neural Networks (RNNs)     | LSTM ([recursive](#lstm-recursive) • [vector](#lstm-vector) • [encoder-decoder](#lstm-encoder-decoder)), GRU ([recursive](#gru-recursive) • [vector](#gru-vector)) |\n| Statistical Models                   | ARIMA ([recursive](#arima)), SARIMAX ([vector](#sarimax)), FBProphet ([direct](#fbprophet)) |\n\nWant a model added to this list? Raise an issue [here](https://github.com/puffinsoft/benchstreet/issues) or [make a PR](https://github.com/puffinsoft/benchstreet/pulls)!\n\n\u003e [!TIP]\n\u003e **The winner**: [N-BEATS](#n-beats). High accuracy with extremely low training time.\n\n\u003chr /\u003e\n\n### TimesFM\n\n#### timesfm-baseline\n\n\u003cimg src=\"docs/images/timesfm_baseline.png\"/\u003e\n\n[`timesfm/baseline.py`](benchstreet/timesfm/baseline.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/timesfm_baseline_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### timesfm-fine-tuned\n\n\u003cimg src=\"docs/images/timesfm_finetuned.png\"/\u003e\n\n[`timesfm/fine_tune.py`](benchstreet/timesfm/fine_tune.py) • [download on huggingface 🤗](https://huggingface.co/ColonelParrot/benchstreet-timesfm-2.0-500m-torch-sp500)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/timesfm_finetuned_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### Chronos\n\n#### chronos-baseline\n\n\u003cimg src=\"docs/images/chronos_baseline.png\"/\u003e\n\n[`chronos/baseline.py`](benchstreet/chronos/baseline.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/chronos_baseline_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### chronos-fine-tuned\n\n\u003cimg src=\"docs/images/chronos_finetuned.png\"/\u003e\n\n[`chronos/fine_tune.ipynb`](benchstreet/chronos/fine_tune.ipynb) • [download on huggingface 🤗](https://huggingface.co/ColonelParrot/benchstreet-chronos-t5-small-sp500)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/chronos_finetuned_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### MLP\n\n#### mlp-recursive\n\n\u003cimg src=\"docs/images/mlp_recursive.png\"/\u003e\n\n[`mlp/recursive.py`](benchstreet/mlp/recursive.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/mlp_recursive_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### mlp-vector\n\n\u003cimg src=\"docs/images/mlp_vector_multistep.png\"/\u003e\n\n[`mlp/vector.py`](benchstreet/mlp/vector.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/mlp_vector_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### N-BEATS\n\n\u003cimg src=\"docs/images/nbeats.png\"/\u003e\n\n[`n_beats/direct.py`](benchstreet/n_beats/direct.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/nbeats_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### 1D-CNN\n\n#### 1d-cnn-recursive\n\n\u003cimg src=\"docs/images/cnn_1d_recursive.png\"/\u003e\n\n[`cnn/recursive.py`](benchstreet/cnn/recursive.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/cnn_1d_recursive_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### 1d-cnn-vector\n\n\u003cimg src=\"docs/images/cnn_1d_vector_multistep.png\"/\u003e\n\n[`cnn/vector.py`](benchstreet/cnn/vector.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/cnn_1d_vector_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### TemporalCN\n\n\u003cimg src=\"docs/images/tcn_vector_multistep.png\"/\u003e\n\n[`tcn/vector.py`](benchstreet/tcn/vector.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/tcn_vector_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### LSTM\n\n#### lstm-recursive\n\n\u003cimg src=\"docs/images/lstm_recursive.png\"/\u003e\n\n[`lstm/recursive.py`](benchstreet/lstm/recursive.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/lstm_recursive_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### lstm-vector\n\n\u003cimg src=\"docs/images/lstm_vector_multistep.png\"/\u003e\n\n[`lstm/vector.py`](benchstreet/lstm/vector.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/lstm_vector_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### lstm-encoder-decoder\n\n\u003cimg src=\"docs/images/lstm_encdec_multistep.png\"/\u003e\n\n[`lstm/encdec.py`](benchstreet/lstm/encdec.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/lstm_encdec_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### GRU\n\n#### gru-recursive\n\n\u003cimg src=\"docs/images/gru_recursive.png\"/\u003e\n\n[`gru/recursive.py`](benchstreet/gru/recursive.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/gru_recursive_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n#### gru-vector\n\n\u003cimg src=\"docs/images/gru_vector_multistep.png\"/\u003e\n\n[`gru/vector.py`](benchstreet/gru/vector.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/gru_vector_multistep_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### ARIMA\n\n\u003cimg src=\"docs/images/arima_recursive.png\"/\u003e\n\n[`arima/recursive.py`](benchstreet/arima/recursive.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/arima_recursive_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### SARIMAX\n\n\u003cimg src=\"docs/images/sarima.png\"/\u003e\n\n[`sarimax/direct.py`](benchstreet/sarimax/direct.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/sarima_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n### FBProphet\n\n\u003cimg src=\"docs/images/prophet.png\"/\u003e\n\n[`fbprophet/direct.py`](benchstreet/fbprophet/direct.py)\n\n\u003cdetails\u003e\n\u003csummary\u003eview magnified graph\u003c/summary\u003e\n\u003cimg src=\"docs/images/prophet_ZOOMED.png\" width=\"80%\"\u003e\n\u003c/details\u003e\n\n\u003chr /\u003e\n\nWant a model added to this list? Raise an issue [here](https://github.com/puffinsoft/benchstreet/issues) or [make a PR](https://github.com/puffinsoft/benchstreet/pulls)!","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpuffinsoft%2Fbenchstreet","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpuffinsoft%2Fbenchstreet","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpuffinsoft%2Fbenchstreet/lists"}