{"id":28461035,"url":"https://github.com/asrot0/polyglottransformer","last_synced_at":"2025-07-03T06:31:20.941Z","repository":{"id":288622888,"uuid":"963911696","full_name":"asRot0/PolyglotTransformer","owner":"asRot0","description":"A fully modular, notebook-powered Transformer model for translating across languages—including English, Bengali, Japanese, German, and more. Built for flexibility and global communication.","archived":false,"fork":false,"pushed_at":"2025-05-18T12:04:20.000Z","size":14801,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-06-07T03:07:26.075Z","etag":null,"topics":["ai","deep-learning","multilingual","transformer"],"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/asRot0.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,"zenodo":null}},"created_at":"2025-04-10T12:00:21.000Z","updated_at":"2025-05-23T14:05:33.000Z","dependencies_parsed_at":"2025-04-20T11:48:44.138Z","dependency_job_id":null,"html_url":"https://github.com/asRot0/PolyglotTransformer","commit_stats":null,"previous_names":["asrot0/polyglottransformer"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/asRot0/PolyglotTransformer","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/asRot0%2FPolyglotTransformer","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/asRot0%2FPolyglotTransformer/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/asRot0%2FPolyglotTransformer/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/asRot0%2FPolyglotTransformer/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/asRot0","download_url":"https://codeload.github.com/asRot0/PolyglotTransformer/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/asRot0%2FPolyglotTransformer/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":263275102,"owners_count":23441143,"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":["ai","deep-learning","multilingual","transformer"],"created_at":"2025-06-07T03:07:25.928Z","updated_at":"2025-07-03T06:31:20.929Z","avatar_url":"https://github.com/asRot0.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# PolyglotTransformer\n\n\u003e A multilingual Transformer-based Neural Machine Translation (NMT) system built with TensorFlow/Keras — modular, extendable, and production-ready.\n\n**PolyglotTransformer** is a fully modular implementation of the Transformer architecture for sequence-to-sequence tasks, such as language translation. It supports language-specific notebooks and is designed for easy extensibility and experimentation.\n\n- Built on TensorFlow 2 \u0026 Keras  \n- Clean modular codebase under `models/`  \n- Notebook-ready workflows for multilingual translation  \n- Supports English to Bengali, German, and more  \n\n```bash\nPolyglotTransformer/\n│\n├── data/\n├── models/\n│   ├── __init__.py\n│   ├── transformer_encoder.py       # Transformer Encoder Block\n│   ├── transformer_decoder.py       # Transformer Decoder Block\n│   ├── multi_head_attention.py      # Multi-Head Self-Attention\n│   ├── feed_forward.py              # Position-wise Feed Forward Layer\n│   ├── positional_encoding.py       # Sinusoidal Positional Encoding\n│   ├── positional_embedding.py      # Token + Positional Embedding\n│   └── transformer.py               # Full Encoder-Decoder Transformer\n│\n├── eng_to_ban.ipynb                 # English ➜ Bengali notebook\n├── eng_to_japanese.ipynb            # English ➜ Japanese notebook\n├── README.md                        \n```\n\n## Model Architecture\n\n- Multi-head attention\n- Position-wise feed-forward networks\n- Layer normalization \u0026 residual connections\n- Sinusoidal or learned positional embeddings\n- Encoder–Decoder structure (as proposed in \"Attention Is All You Need\")\n\n\n### 🗾 Translation Demo (English ➜ Japanese)\n\n| 🌐 English Sentence                         | 🇯🇵 Japanese Translation                        |\n|--------------------------------------------|-------------------------------------------------|\n| Do you have any plans for today?           | 今日の予定はありますか？                       |\n| I forgot to bring my umbrella.             | 傘を持ってくるのを忘れました。                  |\n| Let's meet at the cafe at 3 p.m.           | 午後3時にカフェで会いましょう。                 |\n| This movie was really interesting.         | この映画は本当に面白かったです。                |\n| I want to visit Japan someday.             | いつか日本を訪れたいです。                      |\n\n\n### 🌍 Translation Demo (English ➜ Bengali)\n\n| 🌐 English Sentence                         | 🇧🇩 Bengali Translation                      |\n|--------------------------------------------|---------------------------------------------|\n| What are you doing?                        | তুমি কী করছো?                               |\n| I love learning new languages.             | আমি নতুন ভাষা শিখতে ভালোবাসি।              |\n| Can you help me, please?                   | তুমি কি আমাকে সাহায্য করতে পারো?           |\n| This is my favorite book.                  | এটি আমার প্রিয় বই।                         |\n| We need to leave early tomorrow morning.   | আমাদের আগামীকাল সকালেই রওনা দিতে হবে।     |\n\n\u003e 📝 *Translations generated using the trained Transformer model.*\n\n## Extend to More Languages\n\nJust copy an existing notebook like `eng_to_ban.ipynb` and change:\n- The parallel corpus (`.txt`) file\n- `source_vectorization` / `target_vectorization` setup\n- Save as `eng_to_german.ipynb` or similar!\n\n#### Recommended Dataset Source\n\nYou can download free English-to-X language sentence pairs from:\n🔗 [https://www.manythings.org/anki/](https://www.manythings.org/anki/)\n\nExample files:\n- `eng-ger.txt` → English to German\n- `eng-jpn.txt` → English to Japanese\n- `eng-ita.txt` → English to Italian\n- `eng-spa.txt` → English to Spanish\n\nJust place the downloaded `.txt` file into your project and update the `load_data()` function accordingly!\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fasrot0%2Fpolyglottransformer","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fasrot0%2Fpolyglottransformer","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fasrot0%2Fpolyglottransformer/lists"}