Back to all models
Model card Files and versions Use in transformers
text2text-generation mask_token:
Query this model
πŸ”₯ This model is currently loaded and running on the Inference API. ⚠️ This model could not be loaded by the inference API. ⚠️ This model can be loaded on the Inference API on-demand.
JSON Output
API endpoint  

⚑️ Upgrade your account to access the Inference API

Share Copied link to clipboard

Contributed by

Microsoft company
17 team members Β· 33 models

prophetnet-large-uncased-squad-qg Fine-tuned weights(converted from original fairseq version repo) for ProphetNet on question generation SQuAD 1.1.
ProphetNet is a new pre-trained language model for sequence-to-sequence learning with a novel self-supervised objective called future n-gram prediction.
ProphetNet is able to predict more future tokens with a n-stream decoder. The original implementation is Fairseq version at github repo.


from transformers import ProphetNetTokenizer, ProphetNetForConditionalGeneration, ProphetNetConfig

model = ProphetNetForConditionalGeneration.from_pretrained('microsoft/prophetnet-large-uncased-squad-qg')
tokenizer = ProphetNetTokenizer.from_pretrained('microsoft/prophetnet-large-uncased-squad-qg')

FACT_TO_GENERATE_QUESTION_FROM = ""Bill Gates [SEP] Microsoft was founded by Bill Gates and Paul Allen on April 4, 1975."

inputs = tokenizer([FACT_TO_GENERATE_QUESTION_FROM], return_tensors='pt')

# Generate Summary
question_ids = model.generate(inputs['input_ids'], num_beams=5, early_stopping=True)
tokenizer.batch_decode(question_ids, skip_special_tokens=True)

# should give: 'along with paul allen, who founded microsoft?'


  title={Prophetnet: Predicting future n-gram for sequence-to-sequence pre-training},
  author={Yan, Yu and Qi, Weizhen and Gong, Yeyun and Liu, Dayiheng and Duan, Nan and Chen, Jiusheng and Zhang, Ruofei and Zhou, Ming},
  journal={arXiv preprint arXiv:2001.04063},