Downloads · 30 days
33
5% of all-time downloads
MacSQuAD
Nadav/MacSQuAD
MacSQuAD is a question answering model from Nadav. Use it when the input is a question plus a passage. It is set up for transformers. The card lists the license as afl-3.0.
A MacBERTh model fine-tuned on SQuADv2. Hopefully, this will allow the model to perform well on QA tasks on historical texts. Finetune parameters:
All-time downloads
704
Public
Repo size
868 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin434 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- License
- afl-3.0
- Model type
- bert
- Access
- Public
- Created
- Mar 16, 2022
- Updated
- Mar 17, 2022
- SHA
- 0ef05211
Try a prompt
Model card
From the Hugging Face model README
A MacBERTh model fine-tuned on SQuAD_v2. Hopefully, this will allow the model to perform well on QA tasks on historical texts. Finetune parameters:
training_args = TrainingArguments(
output_dir="./results",
evaluation_strategy="epoch",
learning_rate=3e-5,
per_device_train_batch_size=64,
per_device_eval_batch_size=64,
num_train_epochs=2,
weight_decay=0.01,
lr_scheduler_type=SchedulerType.LINEAR,
warmup_ratio=0.2
)
Evaluation metrics on the validation set of SQuAD_v2:
{'exact': 49.49886296639434, 'f1': 53.9199170778635, 'total': 11873, 'HasAns_exact': 60.08771929824562, 'HasAns_f1': 68.94250598270429, 'HasAns_total': 5928, 'NoAns_exact': 38.940285954583686, 'NoAns_f1': 38.940285954583686, 'NoAns_total': 5945, 'best_exact': 50.5095595047587, 'best_exact_thresh': 0.0, 'best_f1': 51.75825524534494, 'best_f1_thresh': 0.0}
- Task
- Question Answering
- Library
- transformers
- Type
- bert
- License
- afl-3.0
- Created
- Mar 16, 2022
- Updated
- Mar 17, 2022