|
--- |
|
language: en |
|
datasets: |
|
- squad_v2 |
|
license: cc-by-4.0 |
|
model-index: |
|
- name: plm_qa |
|
results: |
|
- task: |
|
type: question-answering |
|
name: Question Answering |
|
dataset: |
|
name: squad_v2 |
|
type: squad_v2 |
|
config: squad_v2 |
|
split: validation |
|
metrics: |
|
- name: Exact Match |
|
type: exact_match |
|
value: 0 |
|
verified: false |
|
- name: F1 |
|
type: f1 |
|
value: 0 |
|
verified: false |
|
- name: total |
|
type: total |
|
value: 11869 |
|
verified: false |
|
--- |
|
|
|
# roberta-base for QA finetuned over community safety domain data |
|
|
|
We fine-tuned the roBERTa-based model (https://huggingface.co/deepset/roberta-base-squad2) over LiveSafe community safety dialogue data for event argument extraction with the objective of question-answering. |
|
|
|
|
|
### Using model in Transformers |
|
```python |
|
from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline |
|
model_name = "yirenl2/plm_qa" |
|
# a) Get predictions |
|
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name) |
|
QA_input = { |
|
'question': 'What is the location of the incident?', |
|
'context': 'I was attacked by someone in front of the bus station.' |
|
} |
|
res = nlp(QA_input) |
|
# b) Load model & tokenizer |
|
model = AutoModelForQuestionAnswering.from_pretrained(model_name) |
|
tokenizer = AutoTokenizer.from_pretrained(model_name) |
|
``` |
