Downloads · 30 days
70
1% of all-time downloads
xlnet-base-squadv2
ggoggam/xlnet-base-squadv2
xlnet-base-squadv2 is a question answering model from ggoggam. Use it when the input is a question plus a passage. It is set up for transformers.
XLNet jointly developed by Google and CMU and fine-tuned on SQuAD 2.0 for question answering down-stream task.
All-time downloads
8.5K
Public
Repo size
953 MB
Likes
2
Public
Repo makeup
Click a slice to open those files.
.bin476 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- Model type
- xlnet
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Jan 17, 2021
- SHA
- 36d0bd03
Try a prompt
Model card
From the Hugging Face model README
XLNet Fine-tuned on SQuAD 2.0 Dataset
XLNet jointly developed by Google and CMU and fine-tuned on SQuAD 2.0 for question answering down-stream task.
Training Results (Metrics)
{
"HasAns_exact": 74.7132253711201
"HasAns_f1": 82.11971607032643
"HasAns_total": 5928
"NoAns_exact": 73.38940285954584
"NoAns_f1": 73.38940285954584
"NoAns_total": 5945
"best_exact": 75.67590331003116
"best_exact_thresh": -19.554906845092773
"best_f1": 79.16215426779269
"best_f1_thresh": -19.554906845092773
"epoch": 4.0
"exact": 74.05036637749515
"f1": 77.74830934598614
"total": 11873
}
Results Comparison
| Metric | Paper | Model |
|---|---|---|
| EM | 78.46 | 75.68 (-2.78) |
| F1 | 81.33 | 79.16 (-2.17) |
Better fine-tuned models coming soon.
How to Use
from transformers import XLNetForQuestionAnswering, XLNetTokenizerFast
model = XLNetForQuestionAnswering.from_pretrained('jkgrad/xlnet-base-squadv2)
tokenizer = XLNetTokenizerFast.from_pretrained('jkgrad/xlnet-base-squadv2')
- Task
- Question Answering
- Library
- transformers
- Type
- xlnet
- Created
- Mar 2, 2022
- Updated
- Jan 17, 2021