Instructions to use procedure2012/Orion-Translate-XL with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use procedure2012/Orion-Translate-XL with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="procedure2012/Orion-Translate-XL")# Load model directly from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("procedure2012/Orion-Translate-XL") model = AutoModel.from_pretrained("procedure2012/Orion-Translate-XL", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| license: mit | |
| library_name: transformers | |
| # Orion-Translate-XL | |
| <!-- markdownlint-disable first-line-h1 --> | |
| <!-- markdownlint-disable html --> | |
| <!-- markdownlint-disable no-duplicate-header --> | |
| <div align="center"> | |
| <img src="figures/fig1.png" width="60%" alt="Orion-Translate-XL" /> | |
| </div> | |
| <hr> | |
| <div align="center" style="line-height: 1;"> | |
| <a href="LICENSE" style="margin: 2px;"> | |
| <img alt="License" src="figures/fig2.png" style="display: inline-block; vertical-align: middle;"/> | |
| </a> | |
| </div> | |
| ## 1. Introduction | |
| Orion-Translate-XL focuses on high-fidelity multilingual translation while retaining general reasoning ability. | |
| ## 2. Evaluation Results | |
| ### Comprehensive Benchmark Results | |
| <div align="center"> | |
| | | Benchmark | LinguaX | Orion-base | TransPro | Orion-Translate-XL | | |
| |---|---|---|---|---|---| | |
| | **Core Reasoning Tasks** | Math Reasoning | 0.558 | 0.541 | 0.572 | 0.595 | | |
| | | Logical Reasoning | 0.805 | 0.809 | 0.808 | 0.846 | | |
| | | Common Sense | 0.727 | 0.745 | 0.737 | 0.763 | | |
| | **Language Understanding** | Reading Comprehension | 0.688 | 0.727 | 0.685 | 0.734 | | |
| | | Question Answering | 0.616 | 0.593 | 0.590 | 0.629 | | |
| | | Text Classification | 0.795 | 0.787 | 0.798 | 0.844 | | |
| | | Sentiment Analysis | 0.767 | 0.752 | 0.791 | 0.807 | | |
| | **Generation Tasks** | Code Generation | 0.636 | 0.642 | 0.650 | 0.695 | | |
| | | Creative Writing | 0.617 | 0.604 | 0.636 | 0.660 | | |
| | | Dialogue Generation | 0.646 | 0.649 | 0.636 | 0.676 | | |
| | | Summarization | 0.777 | 0.767 | 0.780 | 0.789 | | |
| | **Specialized Capabilities** | Translation | 0.804 | 0.800 | 0.783 | 0.817 | | |
| | | Knowledge Retrieval | 0.683 | 0.649 | 0.692 | 0.699 | | |
| | | Instruction Following | 0.771 | 0.725 | 0.766 | 0.781 | | |
| | | Safety Evaluation | 0.718 | 0.753 | 0.714 | 0.761 | | |
| </div> | |
| ### Overall Performance Summary | |
| The Orion-Translate-XL demonstrates strong performance across all evaluated benchmark categories, with particularly notable results in reasoning and generation tasks. | |
| ## 3. Chat Website & API Platform | |
| We offer a chat interface and API for you to interact with Orion-Translate-XL. Please check our official website for more details. | |
| ## 4. How to Run Locally | |
| Please refer to our code repository for more information about running Orion-Translate-XL locally. | |
| ### Temperature | |
| We recommend setting the temperature parameter to 0.6. | |
| ## 5. License | |
| This repository is released under the mit license. The model supports commercial use. | |
| ## 6. Contact | |
| If you have any questions, please contact us at support@orion-nlp.org. | |