{"id":15656799,"url":"https://github.com/scitator/yats2s","last_synced_at":"2026-03-11T13:39:39.954Z","repository":{"id":133293950,"uuid":"88272857","full_name":"Scitator/YATS2S","owner":"Scitator","description":null,"archived":false,"fork":false,"pushed_at":"2017-08-15T14:38:00.000Z","size":242,"stargazers_count":24,"open_issues_count":1,"forks_count":6,"subscribers_count":2,"default_branch":"versions/tf_1.2","last_synced_at":"2025-10-03T18:59:44.551Z","etag":null,"topics":["deep-learning","deep-neural-networks","natural-language-processing","neural-machine-translation","nlp","nmt","seq2seq","sequence-to-sequence","tensorflow"],"latest_commit_sha":null,"homepage":null,"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/Scitator.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}},"created_at":"2017-04-14T14:08:55.000Z","updated_at":"2024-01-04T16:13:12.000Z","dependencies_parsed_at":null,"dependency_job_id":"ec7949ba-b933-4fd8-a351-5ee10dd5b847","html_url":"https://github.com/Scitator/YATS2S","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/Scitator/YATS2S","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Scitator%2FYATS2S","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Scitator%2FYATS2S/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Scitator%2FYATS2S/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Scitator%2FYATS2S/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Scitator","download_url":"https://codeload.github.com/Scitator/YATS2S/tar.gz/refs/heads/versions/tf_1.2","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Scitator%2FYATS2S/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":30382673,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-11T12:49:11.341Z","status":"ssl_error","status_checked_at":"2026-03-11T12:46:41.342Z","response_time":84,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 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":["deep-learning","deep-neural-networks","natural-language-processing","neural-machine-translation","nlp","nmt","seq2seq","sequence-to-sequence","tensorflow"],"created_at":"2024-10-03T13:04:23.812Z","updated_at":"2026-03-11T13:39:39.940Z","avatar_url":"https://github.com/Scitator.png","language":"Python","readme":"# YATS2S: Yet Another Tensorflow Seq2seq\n\nSo, here you are stranger. Finally, you found it!\n\n### Overview\nAfter some time of looking around for user-friendly and configurable seq2seq TF implementation, I decided to make my own one.\n\nSo, here it is:\n* Pure TF\n* any cell you want - just say it name\n* multi-layer\n* bidirectional\n* attention\n* residual connections, residual dense\n* and other seq2seq cells tricks available!\n* vocabulary trick: joint or different for source and target?\n* scheduled_sampling\n* in-graph beam search\n* TF.Estimators\n* tensorboard integration\n* and finally: best-practices for data input pipelines, let's make it quick!\n\n### Inspired by:\n* [google/seq2seq](https://github.com/google/seq2seq)\n\n### Requirements:\nTensorflow 1.2\n\n#### Example usage\n* [notebook](https://github.com/Scitator/TF-seq2seq/blob/versions/tf_1.2/seq2seq_example.ipynb)\n\n##### Step-by-step guide:\n\n1. find some parallel corpora\n    * for example, let's take en-ru pair from [here](http://www.manythings.org/anki/)\n2. prepare it for training (preprocessing and vocabulary extraction)\n    * quite simple with [this repo](https://github.com/Scitator/subword-nmt)\n    ```bash\n       sh prepare_parallel_data.sh --data ./data/tatoeba_en_ru/en_ru.txt --clear_punctuation --lowercase \\\n           --level bpe --bpe_symbols 10000 --bpe_min_freq 5 \\\n           --vocab_min_freq 5 --vocab_max_size 10000 \\ \n           --merge_sequences --test_ratio 0.1 --clear_tmp\n    ```\n3. run training process\n    * around 100 epochs for this example\n    ```bash\n       rm -r ./logs_170620_tatoeba_en_ru; python train_parallel_corpora.py \\\n           --train_corpora_path ./data/tatoeba_en_ru/train.txt --test_corpora_path ./data/tatoeba_en_ru/test.txt \\\n           --vocab_path ./data/tatoeba_en_ru/vocab.txt \\\n           --embedding_size 128 --num_units 128 --cell_num 1 \\\n           --attention bahdanau --residual_connections --residual_dense \\\n           --training_mode scheduled_sampling_embedding --scheduled_sampling_probability 0.2 \\\n           --batch_size 64 --queue_capacity 1024 --num_threads 1 \\\n           --log_dir ./logs_170620_tatoeba_en_ru \\\n           --train_steps 758200 --eval_steps 842 --min_eval_frequency 7582 \\\n           --gpu_option 0.8\n    ```\n4. run tensorboard (really helpful tool: figures, embeddings, graph structure)\n    ```bash\n       tensorboard --logdir=./logs_170620_tatoeba_en_ru\n    ```\n5. look at the results\n * [data](https://drive.google.com/file/d/0ByLYAV32riyVa3JxXzZEMXBOWVU/view?usp=sharing)\n * [model](https://drive.google.com/file/d/0ByLYAV32riyVcG1aaEVBUmpnNkE/view?usp=sharing)\n * [notebook](https://github.com/Scitator/TF-seq2seq/blob/versions/tf_1.2/seq2seq_example_tatoeba_inference.ipynb)\n\n#### Contributing\n\nIf you find an issue - you know what to do.\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fscitator%2Fyats2s","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fscitator%2Fyats2s","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fscitator%2Fyats2s/lists"}