README.md

---
datasets:
- squad_v2
language: en
license: mit
pipeline_tag: question-answering
tags:
- deberta
- deberta-v3
model-index:
- name: navteca/deberta-v3-base-squad2
  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: 88.0876
      verified: true
    - name: F1
      type: f1
      value: 91.1623
      verified: true
  - task:
      type: question-answering
      name: Question Answering
    dataset:
      name: squad
      type: squad
      config: plain_text
      split: validation
    metrics:
    - name: Exact Match
      type: exact_match
      value: 89.2366
      verified: true
    - name: F1
      type: f1
      value: 95.0569
      verified: true
---

# Deberta v3 large model for QA (SQuAD 2.0)

This is the [deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) model, fine-tuned using the [SQuAD2.0](https://huggingface.co/datasets/squad_v2) dataset. It's been trained on question-answer pairs, including unanswerable questions, for the task of Question Answering.

## Training Data
The models have been trained on the [SQuAD 2.0](https://rajpurkar.github.io/SQuAD-explorer/) dataset.

It can be used for question answering task.

## Usage and Performance
The trained model can be used like this:
```python
from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline

# Load model & tokenizer
deberta_model = AutoModelForQuestionAnswering.from_pretrained('navteca/deberta-v3-large-squad2')
deberta_tokenizer = AutoTokenizer.from_pretrained('navteca/deberta-v3-large-squad2')

# Get predictions
nlp = pipeline('question-answering', model=deberta_model, tokenizer=deberta_tokenizer)

result = nlp({
    'question': 'How many people live in Berlin?',
    'context': 'Berlin had a population of 3,520,031 registered inhabitants in an area of 891.82 square kilometers.'
})

print(result)

#{
#  "answer": "3,520,031"
#  "end": 36,
#  "score": 0.96186668,
#  "start": 27,
#}
```

## Author
[deepset](http://deepset.ai/)