{"id":17353524,"url":"https://github.com/jeffersonlicet/santander-questions-classification","last_synced_at":"2025-07-18T03:40:07.157Z","repository":{"id":122992892,"uuid":"279933117","full_name":"jeffersonlicet/santander-questions-classification","owner":"jeffersonlicet","description":"🥈 Second Place Solution - Public Leaderboard Top 3 - 0.86639 | Clasificación de preguntas de clientes | Escuela de Ciencias Informáticas 2020. Universidad de Buenos Aires y Banco Santander.","archived":false,"fork":false,"pushed_at":"2020-11-26T21:33:51.000Z","size":230,"stargazers_count":6,"open_issues_count":0,"forks_count":1,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-03-28T09:51:30.278Z","etag":null,"topics":["fastext","keras","machine-learning","nlp","tensorflow","word2vec"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/jeffersonlicet.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"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":"2020-07-15T17:24:45.000Z","updated_at":"2023-01-10T21:37:09.000Z","dependencies_parsed_at":null,"dependency_job_id":"7512fd26-54a2-44a4-9fec-06adf8dfa6e5","html_url":"https://github.com/jeffersonlicet/santander-questions-classification","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/jeffersonlicet%2Fsantander-questions-classification","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jeffersonlicet%2Fsantander-questions-classification/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jeffersonlicet%2Fsantander-questions-classification/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/jeffersonlicet%2Fsantander-questions-classification/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/jeffersonlicet","download_url":"https://codeload.github.com/jeffersonlicet/santander-questions-classification/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248966928,"owners_count":21190870,"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":["fastext","keras","machine-learning","nlp","tensorflow","word2vec"],"created_at":"2024-10-15T17:17:10.039Z","updated_at":"2025-04-14T21:45:55.997Z","avatar_url":"https://github.com/jeffersonlicet.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Santander Questions Classification 2020 - Second Place\n## 🏅 TOP 3 - 0.86639\n\n \u003ca href=\"https://badges.pufler.dev/visits/jeffersonlicet/santander-questions-classification\"\u003e\u003cimg src=\"https://badges.pufler.dev/visits/jeffersonlicet/santander-questions-classification\"\u003e\u003c/a\u003e\n\n## Exploración de datos y explicación de la solución:\nhttps://github.com/jeffersonlicet/santander-questions-classification/blob/master/Informe.ipynb\n\n## Solución:\n\n![](https://camo.githubusercontent.com/c814494ff12e6337a262bf025f01d3d8eefb3725/68747470733a2f2f692e696d6775722e636f6d2f56744b6437704b2e706e67)\n\nIndicaciones para entrenar los modelos:\n\nSe adjuntan 4 archivos:\n* LSTM_GRU_0_86.ipynb Notebook con todo el código para entrenar los modelos basados en LSTM y GRU\n* BERT_0_86.ipynb Notebook con todo el código para entrenar el modelo basado en BERT\n* Informe.ipynb Notebook con el informe y el análisis de datos.\n* assemble.py Script que ensambla las predicciones de los 3 modelo y genera un archivo listo para ser enviado a la competencia.\n\n# 1: Entrenar los modelos:\n\nCorrer los notebooks preferiblemente en paralelo y utilizando Google Colaboratory.\n  * LSTM_GRU_0_86.ipynb con GPU Activado\n  * BERT_0_86.ipynb con TPU Activado\n\n# 2: Descargar los archivos\nUna vez finalizado el entrenamiento, que toma algo más de una hora, ambos notebooks van a intentar\ndescargar archivos, en caso de que no posean permisos para descargarlos por parte del browser puede\nintentar descargarlos manualmente.\n\n# 3: Ensamblar la solución\nArchivos que deben estar descargados y en el mismo directorio:\n* test_ids.npy Contiene los ids de los casos de testing\n* labels.npy Contiene los nombres de las clases mapeadas a indices\n* bert.npy Contiene la distribución de probabilidad calculada usando BERT\n* lstm.npy Contiene la distribución de probabilidad calculada usando LSTM\n* gru.npy Contiene la distribución de probabilidad calculada usando GRU\n\nCon los archivos en el mismo directorio que el archivo assemble.py que se encuentra\nadjuntado realizar lo siguiete:\n\n\u003e pip install numpy\n\ny luego \n\u003e python assemble.py\n\nFinalmente se va a generar un archivo llamado submission.csv con las predicciones correspondientes.\n\n\nGracias.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjeffersonlicet%2Fsantander-questions-classification","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjeffersonlicet%2Fsantander-questions-classification","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjeffersonlicet%2Fsantander-questions-classification/lists"}