Instructions to use Helsinki-NLP/opus-mt-es-en with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Helsinki-NLP/opus-mt-es-en with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-es-en")# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("Helsinki-NLP/opus-mt-es-en") model = AutoModelForSeq2SeqLM.from_pretrained("Helsinki-NLP/opus-mt-es-en", device_map="auto") - Inference
- Notebooks
- Google Colab
- Kaggle
Commercial redistribution question: CTranslate2 conversion of opus-mt-en-es / opus-mt-es-en
Hi, I'm developing a commercial desktop application that uses Helsinki-NLP/opus-mt-en-es and Helsinki-NLP/opus-mt-es-en.
The original model weights are converted from Marian/Transformers format to CTranslate2 format for local inference, without retraining the models or modifying their linguistic content.
I would like to redistribute these converted model files as downloadable assets for users of the commercial application.
Both Hugging Face model pages currently list the license as Apache-2.0. Before distributing them, I would like to confirm:
Is commercial redistribution of the CTranslate2-converted weights permitted?
Which license should accompany the converted weights?
What copyright, attribution or NOTICE information should be included?
Do licenses associated with any of the training datasets impose additional redistribution restrictions on these specific model weights?
Thank you for your clarification.