# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("WindstormLabs/translate-pt-en", dtype="auto")Quick Links
WindyWord.ai Translation — Portuguese → English
Translates Portuguese (pt) text into English (en).
Quality & Model Lineage
- Tier: Standard ⭐⭐⭐⭐
- Base model: Helsinki-NLP/opus-mt-ROMANCE-en — 68k+ downloads, multilingual Romance-source model
- Architecture: MarianMT
- Note: Helsinki-NLP does not publish a dedicated Portuguese-only
opus-mt-pt-enmodel. Their canonical path for Portuguese-source translation is the multilingualopus-mt-ROMANCE-en, which handles Portuguese alongside Italian, Spanish, French, Catalan, and Romanian as source languages. This WindyWord repo wraps the same underlying weights with a Portuguese-specific repo name to satisfy installer / consumer code that expects atranslate-pt-enartifact.
For sibling Romance-source pairs use WindyWord/translate-ROMANCE-en directly (same underlying model).
Available Variants
| Variant | Description |
|---|---|
lora/ |
WindyStandard — production baseline. Helsinki-NLP/opus-mt-ROMANCE-en weights wrapped under the WindyWord pt-en namespace. |
Quick Usage
from transformers import MarianMTModel, MarianTokenizer
tokenizer = MarianTokenizer.from_pretrained("WindyWord/translate-pt-en", subfolder="lora")
model = MarianMTModel.from_pretrained("WindyWord/translate-pt-en", subfolder="lora")
Commercial Use
Visit windyword.ai for apps and API access.
Provenance & License
Weights derived from Helsinki-NLP/opus-mt-ROMANCE-en under CC-BY-4.0. Wrapped as a Portuguese-specific entry by WindyWord 2026-04-28 to support installer expectations for a translate-pt-en repo. Underlying model is identical to WindyWord/translate-ROMANCE-en.
Certified by Opus 4.6 Opus-Claw (Dr. C) on Veron-1 (RTX 5090, Mt Pleasant SC).
# 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="WindstormLabs/translate-pt-en")