OpenGVLab / InternVL

[CVPR 2024 Oral] InternVL Family: A Pioneering Open-Source Alternative to GPT-4o. 接近GPT-4o表现的开源多模态对话模型
https://internvl.readthedocs.io/en/latest/
MIT License
5.63k stars 439 forks source link

检索微调样本 #607

Open wonder-dan opened 4 days ago

wonder-dan commented 4 days ago

Checklist

Describe the bug

请问 对比损失中,负文本添加'summarize:'字符串目的是什么

summarize_model_inputs = self.tokenizer( 'summarize:' + caption, max_length=self.data_args.max_seq_length, padding='max_length' if self.data_args.pad_to_max_length else False, truncation=True, return_tensors='pt', )

Reproduction

summarize_model_inputs = self.tokenizer( 'summarize:' + caption, max_length=self.data_args.max_seq_length, padding='max_length' if self.data_args.pad_to_max_length else False, truncation=True, return_tensors='pt', )

Environment

Error traceback

No response