Downloads · 30 days
23
0% of all-time downloads
albert-chinese-large-qa
wptoux/albert-chinese-large-qa
albert-chinese-large-qa is a question answering model from wptoux. Use it when the input is a question plus a passage. It is set up for transformers. The card lists the license as apache-2.0.
Albert large QA model pretrained from baidu webqa and baidu dureader datasets.
All-time downloads
42K
Public
Repo size
190 MB
Likes
12
Public
Repo makeup
Click a slice to open those files.
.bin62 MB · 100%
At a glance
- Task
- Question Answering
- Library
- transformers
- License
- apache-2.0
- Model type
- albert
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Mar 9, 2021
- SHA
- 02a1762f
Try a prompt
Model card
From the Hugging Face model README
albert-chinese-large-qa
Albert large QA model pretrained from baidu webqa and baidu dureader datasets.
Data source
- baidu webqa 1.0
- baidu dureader
Traing Method
We combined the two datasets together and created a new dataset in squad format, including 705139 samples for training and 69638 samples for validation. We finetune the model based on the albert chinese large model.
Hyperparams
- learning_rate 1e-5
- max_seq_length 512
- max_query_length 50
- max_answer_length 300
- doc_stride 256
- num_train_epochs 2
- warmup_steps 1000
- per_gpu_train_batch_size 8
- gradient_accumulation_steps 3
- n_gpu 2 (Nvidia Tesla P100)
Usage
from transformers import AutoModelForQuestionAnswering, BertTokenizer
model = AutoModelForQuestionAnswering.from_pretrained('wptoux/albert-chinese-large-qa')
tokenizer = BertTokenizer.from_pretrained('wptoux/albert-chinese-large-qa')
Important: use BertTokenizer
MoreInfo
Please visit https://github.com/wptoux/albert-chinese-large-webqa for details.
- Task
- Question Answering
- Library
- transformers
- Type
- albert
- License
- apache-2.0
- Languages
- zh
- Created
- Mar 2, 2022
- Updated
- Mar 9, 2021