faster-translate


Namefaster-translate JSON
Version 1.0.2 PyPI version JSON
download
home_pageNone
SummaryA high-performance translation library using CTTranslate2 and vLLM.
upload_time2025-03-03 23:14:12
maintainerNone
docs_urlNone
authorNone
requires_python>=3.8
licenseMIT License Copyright (c) 2024 Sawradip Saha Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
keywords translation huggingface nlp ctranslate2 vllm
VCS
bugtrack_url
requirements No requirements were recorded.
Travis-CI No Travis.
coveralls test coverage No coveralls. 
            # Faster Translate

<div align="center">

[![PyPI Downloads](https://static.pepy.tech/badge/faster-translate)](https://pepy.tech/projects/faster-translate)
[![Monthly Downloads](https://static.pepy.tech/badge/faster-translate/month)](https://pepy.tech/projects/faster-translate)
[![GitHub License](https://img.shields.io/github/license/sawradip/faster-translate)](https://github.com/sawradip/faster-translate/blob/main/LICENSE)
[![PyPI Version](https://img.shields.io/pypi/v/faster-translate)](https://pypi.org/project/faster-translate/)

</div>

A high-performance translation library powered by state-of-the-art models. Faster Translate offers optimized inference using CTranslate2 and vLLM backends, providing an easy-to-use interface for applications requiring efficient and accurate translations.

## 🚀 Features

- **High-performance inference** using CTranslate2 and vLLM backends
- **Seamless integration** with Hugging Face models
- **Flexible API** for single sentence, batch, and large-scale translation
- **Dataset translation** with direct Hugging Face integration
- **Multi-backend support** for both traditional (CTranslate2) and LLM-based (vLLM) models
- **Text normalization** for improved translation quality

## đŸ“Ļ Installation

```bash
pip install faster-translate
```

### Optional Dependencies

For specific normalizers or model backends:

```bash
# For Bengali text normalization
pip install git+https://github.com/csebuetnlp/normalizer

# For vLLM backend support (required for LLM-based models)
pip install vllm
```

## 🔍 Usage

### Basic Translation

```python
from faster_translate import TranslatorModel

# Initialize with a pre-configured model
translator = TranslatorModel.from_pretrained("banglanmt_bn2en")

# Translate a single sentence
english_text = translator.translate_single("āĻĻā§‡āĻļā§‡ āĻŦāĻŋāĻĻā§‡āĻļāĻŋ āĻ‹āĻŖ āĻ¨āĻŋāĻ¯āĻŧā§‡ āĻāĻ–āĻ¨ āĻŦā§‡āĻļ āĻ†āĻ˛ā§‹āĻšāĻ¨āĻž āĻšāĻšā§āĻ›ā§‡āĨ¤")
print(english_text)

# Translate a batch of sentences
bengali_sentences = [
    "āĻĻā§‡āĻļā§‡ āĻŦāĻŋāĻĻā§‡āĻļāĻŋ āĻ‹āĻŖ āĻ¨āĻŋāĻ¯āĻŧā§‡ āĻāĻ–āĻ¨ āĻŦā§‡āĻļ āĻ†āĻ˛ā§‹āĻšāĻ¨āĻž āĻšāĻšā§āĻ›ā§‡āĨ¤",
    "āĻ°āĻžāĻ¤ āĻ¤āĻŋāĻ¨āĻŸāĻžāĻ° āĻĻāĻŋāĻ•ā§‡ āĻ•āĻžāĻāĻšāĻžāĻŽāĻžāĻ˛ āĻ¨āĻŋāĻ¯āĻŧā§‡ āĻ—ā§āĻ˛āĻŋāĻ¸ā§āĻ¤āĻžāĻ¨ āĻĨā§‡āĻ•ā§‡ āĻĒā§āĻ°āĻžāĻ¨ āĻĸāĻžāĻ•āĻžāĻ° āĻļā§āĻ¯āĻžāĻŽāĻŦāĻžāĻœāĻžāĻ°ā§‡āĻ° āĻ†āĻĄāĻŧāĻ¤ā§‡ āĻ¯āĻžāĻšā§āĻ›āĻŋāĻ˛ā§‡āĻ¨ āĻ˛āĻŋāĻŸāĻ¨ āĻŦā§āĻ¯āĻžāĻĒāĻžāĻ°ā§€āĨ¤"
]
translations = translator.translate_batch(bengali_sentences)
```

### Using Different Model Backends

```python
# Using a CTTranslate2-based model
ct2_translator = TranslatorModel.from_pretrained("banglanmt_bn2en")

# Using a vLLM-based model
vllm_translator = TranslatorModel.from_pretrained("bangla_qwen_en2bn")
```

### Loading Models from Hugging Face

```python
# Load a specific model from Hugging Face
translator = TranslatorModel.from_pretrained(
    "sawradip/faster-translate-banglanmt-bn2en-t5",
    normalizer_func="buetnlpnormalizer"
)
```

### Translating Hugging Face Datasets

Translate an entire dataset with a single function call:

```python
translator = TranslatorModel.from_pretrained("banglanmt_en2bn")

# Translate the entire dataset
translator.translate_hf_dataset(
    "sawradip/bn-translation-mega-raw-noisy", 
    batch_size=16
)

# Translate specific subsets
translator.translate_hf_dataset(
    "sawradip/bn-translation-mega-raw-noisy",
    subset_name=["google"], 
    batch_size=16
)

# Translate a portion of the dataset
translator.translate_hf_dataset(
    "sawradip/bn-translation-mega-raw-noisy",
    subset_name="alt",
    batch_size=16, 
    translation_size=0.5  # Translate 50% of the dataset
)
```

### Publishing Translated Datasets

Push translated datasets directly to Hugging Face:

```python
translator.translate_hf_dataset(
    "sawradip/bn-translation-mega-raw-noisy",
    subset_name="alt",
    batch_size=16, 
    push_to_hub=True,
    token="your_huggingface_token",
    save_repo_name="your-username/translated-dataset"
)
```

## 🌐 Supported Models

| Model ID | Source Language | Target Language | Backend | Description |
|----------|----------------|----------------|---------|-------------|
| `banglanmt_bn2en` | Bengali | English | CTranslate2 | BanglaNMT model from BUET |
| `banglanmt_en2bn` | English | Bengali | CTranslate2 | BanglaNMT model from BUET |
| `bangla_mbartv1_en2bn` | English | Bengali | CTranslate2 | MBart-based translation model |
| `bangla_qwen_en2bn` | English | Bengali | vLLM | Qwen-based translation model |

## 🛠ī¸ Advanced Configuration

### Custom Sampling Parameters for vLLM Models

```python
from vllm import SamplingParams

# Create custom sampling parameters
sampling_params = SamplingParams(
    temperature=0.7,
    top_p=0.9,
    max_tokens=512
)

# Initialize translator with custom parameters
translator = TranslatorModel.from_pretrained(
    "bangla_qwen_en2bn", 
    sampling_params=sampling_params
)
```

## đŸ’Ē Contributors

<a href="https://github.com/sawradip/faster-translate/graphs/contributors">
  <img src="https://contributors-img.web.app/image?repo=sawradip/faster-translate" alt="List of Contributors"/>
</a>

## 📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

## 📚 Citation

If you use Faster Translate in your research, please cite:

```bibtex
@software{faster_translate,
  author = {Sawradip Saha and Contributors},
  title = {Faster Translate: High-Performance Machine Translation Library},
  url = {https://github.com/sawradip/faster-translate},
  year = {2024},
}
```

Raw data

            {
    "_id": null,
    "home_page": null,
    "name": "faster-translate",
    "maintainer": null,
    "docs_url": null,
    "requires_python": ">=3.8",
    "maintainer_email": null,
    "keywords": "translation, huggingface, nlp, ctranslate2, vllm",
    "author": null,
    "author_email": "Sawradip Saha <sawradip0@gmail.com>",
    "download_url": "https://files.pythonhosted.org/packages/73/53/9f3deb31c6b75e281dfd78d03c96eac2d8fe0b4eb9c00ff01ddca04f4ba3/faster_translate-1.0.2.tar.gz",
    "platform": null,
    "description": "# Faster Translate\n\n<div align=\"center\">\n\n[![PyPI Downloads](https://static.pepy.tech/badge/faster-translate)](https://pepy.tech/projects/faster-translate)\n[![Monthly Downloads](https://static.pepy.tech/badge/faster-translate/month)](https://pepy.tech/projects/faster-translate)\n[![GitHub License](https://img.shields.io/github/license/sawradip/faster-translate)](https://github.com/sawradip/faster-translate/blob/main/LICENSE)\n[![PyPI Version](https://img.shields.io/pypi/v/faster-translate)](https://pypi.org/project/faster-translate/)\n\n</div>\n\nA high-performance translation library powered by state-of-the-art models. Faster Translate offers optimized inference using CTranslate2 and vLLM backends, providing an easy-to-use interface for applications requiring efficient and accurate translations.\n\n## \ud83d\ude80 Features\n\n- **High-performance inference** using CTranslate2 and vLLM backends\n- **Seamless integration** with Hugging Face models\n- **Flexible API** for single sentence, batch, and large-scale translation\n- **Dataset translation** with direct Hugging Face integration\n- **Multi-backend support** for both traditional (CTranslate2) and LLM-based (vLLM) models\n- **Text normalization** for improved translation quality\n\n## \ud83d\udce6 Installation\n\n```bash\npip install faster-translate\n```\n\n### Optional Dependencies\n\nFor specific normalizers or model backends:\n\n```bash\n# For Bengali text normalization\npip install git+https://github.com/csebuetnlp/normalizer\n\n# For vLLM backend support (required for LLM-based models)\npip install vllm\n```\n\n## \ud83d\udd0d Usage\n\n### Basic Translation\n\n```python\nfrom faster_translate import TranslatorModel\n\n# Initialize with a pre-configured model\ntranslator = TranslatorModel.from_pretrained(\"banglanmt_bn2en\")\n\n# Translate a single sentence\nenglish_text = translator.translate_single(\"\u09a6\u09c7\u09b6\u09c7 \u09ac\u09bf\u09a6\u09c7\u09b6\u09bf \u098b\u09a3 \u09a8\u09bf\u09af\u09bc\u09c7 \u098f\u0996\u09a8 \u09ac\u09c7\u09b6 \u0986\u09b2\u09cb\u099a\u09a8\u09be \u09b9\u099a\u09cd\u099b\u09c7\u0964\")\nprint(english_text)\n\n# Translate a batch of sentences\nbengali_sentences = [\n    \"\u09a6\u09c7\u09b6\u09c7 \u09ac\u09bf\u09a6\u09c7\u09b6\u09bf \u098b\u09a3 \u09a8\u09bf\u09af\u09bc\u09c7 \u098f\u0996\u09a8 \u09ac\u09c7\u09b6 \u0986\u09b2\u09cb\u099a\u09a8\u09be \u09b9\u099a\u09cd\u099b\u09c7\u0964\",\n    \"\u09b0\u09be\u09a4 \u09a4\u09bf\u09a8\u099f\u09be\u09b0 \u09a6\u09bf\u0995\u09c7 \u0995\u09be\u0981\u099a\u09be\u09ae\u09be\u09b2 \u09a8\u09bf\u09af\u09bc\u09c7 \u0997\u09c1\u09b2\u09bf\u09b8\u09cd\u09a4\u09be\u09a8 \u09a5\u09c7\u0995\u09c7 \u09aa\u09c1\u09b0\u09be\u09a8 \u09a2\u09be\u0995\u09be\u09b0 \u09b6\u09cd\u09af\u09be\u09ae\u09ac\u09be\u099c\u09be\u09b0\u09c7\u09b0 \u0986\u09a1\u09bc\u09a4\u09c7 \u09af\u09be\u099a\u09cd\u099b\u09bf\u09b2\u09c7\u09a8 \u09b2\u09bf\u099f\u09a8 \u09ac\u09cd\u09af\u09be\u09aa\u09be\u09b0\u09c0\u0964\"\n]\ntranslations = translator.translate_batch(bengali_sentences)\n```\n\n### Using Different Model Backends\n\n```python\n# Using a CTTranslate2-based model\nct2_translator = TranslatorModel.from_pretrained(\"banglanmt_bn2en\")\n\n# Using a vLLM-based model\nvllm_translator = TranslatorModel.from_pretrained(\"bangla_qwen_en2bn\")\n```\n\n### Loading Models from Hugging Face\n\n```python\n# Load a specific model from Hugging Face\ntranslator = TranslatorModel.from_pretrained(\n    \"sawradip/faster-translate-banglanmt-bn2en-t5\",\n    normalizer_func=\"buetnlpnormalizer\"\n)\n```\n\n### Translating Hugging Face Datasets\n\nTranslate an entire dataset with a single function call:\n\n```python\ntranslator = TranslatorModel.from_pretrained(\"banglanmt_en2bn\")\n\n# Translate the entire dataset\ntranslator.translate_hf_dataset(\n    \"sawradip/bn-translation-mega-raw-noisy\", \n    batch_size=16\n)\n\n# Translate specific subsets\ntranslator.translate_hf_dataset(\n    \"sawradip/bn-translation-mega-raw-noisy\",\n    subset_name=[\"google\"], \n    batch_size=16\n)\n\n# Translate a portion of the dataset\ntranslator.translate_hf_dataset(\n    \"sawradip/bn-translation-mega-raw-noisy\",\n    subset_name=\"alt\",\n    batch_size=16, \n    translation_size=0.5  # Translate 50% of the dataset\n)\n```\n\n### Publishing Translated Datasets\n\nPush translated datasets directly to Hugging Face:\n\n```python\ntranslator.translate_hf_dataset(\n    \"sawradip/bn-translation-mega-raw-noisy\",\n    subset_name=\"alt\",\n    batch_size=16, \n    push_to_hub=True,\n    token=\"your_huggingface_token\",\n    save_repo_name=\"your-username/translated-dataset\"\n)\n```\n\n## \ud83c\udf10 Supported Models\n\n| Model ID | Source Language | Target Language | Backend | Description |\n|----------|----------------|----------------|---------|-------------|\n| `banglanmt_bn2en` | Bengali | English | CTranslate2 | BanglaNMT model from BUET |\n| `banglanmt_en2bn` | English | Bengali | CTranslate2 | BanglaNMT model from BUET |\n| `bangla_mbartv1_en2bn` | English | Bengali | CTranslate2 | MBart-based translation model |\n| `bangla_qwen_en2bn` | English | Bengali | vLLM | Qwen-based translation model |\n\n## \ud83d\udee0\ufe0f Advanced Configuration\n\n### Custom Sampling Parameters for vLLM Models\n\n```python\nfrom vllm import SamplingParams\n\n# Create custom sampling parameters\nsampling_params = SamplingParams(\n    temperature=0.7,\n    top_p=0.9,\n    max_tokens=512\n)\n\n# Initialize translator with custom parameters\ntranslator = TranslatorModel.from_pretrained(\n    \"bangla_qwen_en2bn\", \n    sampling_params=sampling_params\n)\n```\n\n## \ud83d\udcaa Contributors\n\n<a href=\"https://github.com/sawradip/faster-translate/graphs/contributors\">\n  <img src=\"https://contributors-img.web.app/image?repo=sawradip/faster-translate\" alt=\"List of Contributors\"/>\n</a>\n\n## \ud83d\udcc4 License\n\nThis project is licensed under the MIT License - see the LICENSE file for details.\n\n## \ud83d\udcda Citation\n\nIf you use Faster Translate in your research, please cite:\n\n```bibtex\n@software{faster_translate,\n  author = {Sawradip Saha and Contributors},\n  title = {Faster Translate: High-Performance Machine Translation Library},\n  url = {https://github.com/sawradip/faster-translate},\n  year = {2024},\n}\n```\n",
    "bugtrack_url": null,
    "license": "MIT License\n        \n        Copyright (c) 2024 Sawradip Saha\n        \n        Permission is hereby granted, free of charge, to any person obtaining a copy\n        of this software and associated documentation files (the \"Software\"), to deal\n        in the Software without restriction, including without limitation the rights\n        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell\n        copies of the Software, and to permit persons to whom the Software is\n        furnished to do so, subject to the following conditions:\n        \n        The above copyright notice and this permission notice shall be included in all\n        copies or substantial portions of the Software.\n        \n        THE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\n        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\n        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE\n        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\n        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,\n        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE\n        SOFTWARE.",
    "summary": "A high-performance translation library using CTTranslate2 and vLLM.",
    "version": "1.0.2",
    "project_urls": {
        "Bug Tracker": "https://github.com/sawradip/faster-translate/issues",
        "Documentation": "https://github.com/sawradip/faster-translate/blob/main/README.md",
        "Homepage": "https://github.com/sawradip/faster-translate"
    },
    "split_keywords": [
        "translation",
        " huggingface",
        " nlp",
        " ctranslate2",
        " vllm"
    ],
    "urls": [
        {
            "comment_text": null,
            "digests": {
                "blake2b_256": "6f3d5d4ba2ebc57c0e6d740c6afcad3772b17038517bbb5965f32838efa85f3c",
                "md5": "9400eae67e59b441853a07d9103d42cf",
                "sha256": "13e25dbef4263a47eb5e7c4c6063efce69530158dffc24d31baaad05cb4095df"
            },
            "downloads": -1,
            "filename": "faster_translate-1.0.2-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "9400eae67e59b441853a07d9103d42cf",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": ">=3.8",
            "size": 11815,
            "upload_time": "2025-03-03T23:14:10",
            "upload_time_iso_8601": "2025-03-03T23:14:10.347401Z",
            "url": "https://files.pythonhosted.org/packages/6f/3d/5d4ba2ebc57c0e6d740c6afcad3772b17038517bbb5965f32838efa85f3c/faster_translate-1.0.2-py3-none-any.whl",
            "yanked": false,
            "yanked_reason": null
        },
        {
            "comment_text": null,
            "digests": {
                "blake2b_256": "73539f3deb31c6b75e281dfd78d03c96eac2d8fe0b4eb9c00ff01ddca04f4ba3",
                "md5": "9afdbe581c0f3bd80a1534d7cd11f96b",
                "sha256": "d8810cbe18a4ab8d851f80c7991000ce5025891880a5bcf6bc87ee4016b58fc3"
            },
            "downloads": -1,
            "filename": "faster_translate-1.0.2.tar.gz",
            "has_sig": false,
            "md5_digest": "9afdbe581c0f3bd80a1534d7cd11f96b",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": ">=3.8",
            "size": 13410,
            "upload_time": "2025-03-03T23:14:12",
            "upload_time_iso_8601": "2025-03-03T23:14:12.344412Z",
            "url": "https://files.pythonhosted.org/packages/73/53/9f3deb31c6b75e281dfd78d03c96eac2d8fe0b4eb9c00ff01ddca04f4ba3/faster_translate-1.0.2.tar.gz",
            "yanked": false,
            "yanked_reason": null
        }
    ],
    "upload_time": "2025-03-03 23:14:12",
    "github": true,
    "gitlab": false,
    "bitbucket": false,
    "codeberg": false,
    "github_user": "sawradip",
    "github_project": "faster-translate",
    "travis_ci": false,
    "coveralls": false,
    "github_actions": false,
    "lcname": "faster-translate"
}
None
Elapsed time: 1.57978s