models_for_qa_cut

This model is a fine-tuned version of google-bert/bert-base-chinese on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6446

Model description

使用說明

from transformers import pipeline

pipe = pipeline("question-answering", model="roberthsu2003/models_for_qa_cut")
answer = pipe(question="蔡英文何時卸任?",context="蔡英文於2024年5月卸任中華民國總統，交棒給時任副總統賴清德。卸任後較少公開露面，直至2024年10月她受邀訪問歐洲。[25]")
print(answer['answer'])
#'2024年5月'


context='台積電也承諾未來在台灣的各項投資不變，計劃未來在本國建造九座廠，包括新竹、高雄、台中、嘉義和台南等地，在2035年，台灣仍將生產高達80%的晶片。'''
answer = pipe(question='台積電未來要建立幾座廠',context=context)
print(answer['answer'])
answer = pipe(question='2035年在台灣生產的晶片比例?',context=context)
print(answer['answer'])
#九座
#80%

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
0.6584	1.0	842	0.6412
0.4002	2.0	1684	0.6446

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0