wuzhiye7 / induction-network-on-fewrel Goto Github PK
View Code? Open in Web Editor NEWAn attempt at replicating the Induction Network for FewRel data in Tensorflow
An attempt at replicating the Induction Network for FewRel data in Tensorflow
hi 试着运行了一下,发现loss有下降但acc一直保持20%不变,并且几个batch预测的概率都一样
self.probs =
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.39375782],
[0.6477719 , 0.4859865 , 0.60721827, 0.38423938, 0.3937578 ]],
dtype=float32)
You can add support 2 support pair and yield a better perfomance
with c_i_hat and euclidian, the result can be better
but still bad then proto net
I'm not sure why.......
From the intuition, can't say attention based dynamic routing is better than average.
But the self attention way might works
请问这个数据集那个网站上怎么使用呢,谢谢了
ow.python.framework.errors_impl.InvalidArgumentError: indices[37,20] = 400001 is not in [0, 400000)
InvalidArgumentError (see above for traceback): indices[0,36] = 400001 is not in [0, 400000)
[[Node: EncoderModule/embedding_lookup = GatherV2[Taxis=DT_INT32, Tindices=DT_INT32, Tparams=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"](embedding/Variable/read, _arg_input_words_0_2, EncoderModule/embedding_lookup/axis)]]
实验结果显示一直是20% 这是什么原因呢? 另外,有点不太明白这个MASK在代码中的作用。
可否加一下微信,咱们交流一下。18002594663
模型训练结束后,验证时有点小问题,加载的模型不是之前保存的最好的模型。
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.