asp-coref-t0-3b / README.md
tliu's picture
Update README.md
7d8baf5 verified
metadata
license: mit
datasets:
  - conll2012_ontonotesv5
language:
  - en
metrics:
  - f1

Model Card for asp-coref-flan-t5-large

model image

Intro

This model is initialized from flan-t5-base and finetuned for coreference resolution task. The model structure is described in the paper Autoregressive Structured Prediction with Language Models, Github repo.

Model Description

  • Task: Coreference Resolution
  • Dataset: CoNLL 2012 OntoNotes
  • Base Model: bigscience/T0_3B

Command

CUDA_VISIBLE_DEVICES=0 python evaluate_coref.py t0_3b tliu/asp-coref-t0-3b 0