-
在dureader上提交榜单两天一直在calculating,请问是什么问题?
-
### Describe the issue
We are interested in your longllmlingua results on longbench. We referred to these two parts of your code[https://github.com/microsoft/LLMLingua/blob/main/experiments/llmli…
-
I have three questions
First: Can i directly use SQuAD for chinese (Close-Domain)QA task?
Second: Is it the best solution to use run_squda.py to fine tune bert model with chinese dataset which for…
-
python3 index.py zh ../data/dureader.para test_index 出现这个错误
![image](https://user-images.githubusercontent.com/44161003/233961386-e5eb9b32-1154-4b5c-ad03-e12cd0d2bf08.png)
-
models/model.py line23, 24需要给squeeze函数指定参数dim=-1,否则在设置batch_size不当的情况下,可能产生单样本batch,进而导致此函数产生的输出由应有的2维tensor变成1维tensor,引起崩溃。
-
hi, 我看本项目不支持高版本的cuda, 就在conda上装了个cuda10.2, 发现还是会有同样的问题, 如下:
```
import rocketqa
import paddle
print(paddle.version.cuda())
cross_encoder = rocketqa.load_model(model="zh_dureader_ce", use_cuda=T…
-
zh_dureader_de_v2 文件757M,自己微调后的保存的文件2.22G,为什么会差别这么多呢
-
您好!感谢rocketQA的工作,我已经成功将rocketQA应用到我们的系统中。
为了能提高rocketQA的召回率,我尝试对zh_dureader_de_v2进行finetune,使用的是repo中自带的训练集,下面是训练代码:
```
import rocketqa
dual_encoder = rocketqa.load_model(model="zh_dureader_de_v2…
-
感谢大佬的分享,不过大佬我想问一下,为什么的把模型下载到本地来调用与直接用的huggingface的API结果不一样呢?是huggingface的API部署的模型跟公开的出来的这个模型不是同一个嘛?他们好像实验的时候,得到的答案跟huggingface上的答案也是不一样的https://github.com/basketballandlearn/MRC_Competition_Dureader/i…
-
你好,感谢开源。
我这边有一些QA数据集,用户来了新问题,想先和数据集里的Q做匹配,如果匹配到了,就返回A。
我看本项目的工作,更多的是QA直接匹配,所以想问下QQ匹配的效果,是否可以用QQ数据集,在以下两个模型基础上做一些微调。
效果会不会不好,如果不可以,是否可以推荐一些目前比较好的方法呢
```
de_model = 'zh_dureader_de_v2'
…