nb-nn-translation / README.md
pere's picture
Update README.md
fe6a973
|
raw
history blame
1.26 kB
metadata
language: 'no'
license: CC-BY 4.0
tags:
  - translation
datasets:
  - oscar
widget:
  - text: Dette er en test!

Norwegian mT5 - Translation Bokmål Nynorsk

Description

This is a sample reference model.

Here is an example of how to use the model from Python

# Import libraries
from transformers import T5ForConditionalGeneration, AutoTokenizer
model = T5ForConditionalGeneration.from_pretrained('andrek/nb2nn',from_flax=True)
tokenizer = AutoTokenizer.from_pretrained('andrek/nb2nn')
#Encode the text
text = "Hun vil ikke gi bort sine personlige data."
inputs = tokenizer.encode(text, return_tensors="pt")
outputs = model.generate(inputs, max_length=255, num_beams=4, early_stopping=True)

#Decode and print the result
print(tokenizer.decode(outputs[0]))

Or if you like to use the pipeline instead

# Set up the pipeline
from transformers import pipeline, T5ForConditionalGeneration, AutoTokenizer
model = T5ForConditionalGeneration.from_pretrained('andrek/nb2nn',from_flax=True)
tokenizer = AutoTokenizer.from_pretrained('andrek/nb2nn')
translator = pipeline("translation", model=model, tokenizer=tokenizer)

# Do the translation
text = "Hun vil ikke gi bort sine personlige data."
print(translator(text, max_length=255))