search

CoderPat / structured-neural-summarization

A repository with the code for the paper with the same title
MIT License
74 stars 26 forks source link

about Java-small preprocessing #25

Open hukuda222 opened 4 years ago

hukuda222 commented 4 years ago

I am trying to validate your model with a method-naming task in Java-small. Therefore, I converted the source code into a graph representation using Java2Graph code. However, I didn't find any source code to generate files that are necessary to train the model. So, I would like to ask you to publish the source code or the preprocessed dataset. Thank you.

shengqiangzhang commented 4 years ago

@hukuda222

Have you solved the problem? I have the same problem.

hukuda222 commented 4 years ago

@shengqiangzhang

It has not been solved yet. For the time being, it was possible to change from graph to corpus shape using this code(https://github.com/CoderPat/structured-neural-summarization/blob/master/parsers/sourcecode/csharp/programgraphs2opengnn.py), but my score was very low. So I think the author uses a different code for preprocessing.

Please let me know if you understand anything.

shengqiangzhang commented 4 years ago

@hukuda222

When i use programgraphs2opengnn.py to transfer programgraphs to opengnn. I got the very low score. And then i use graph2subtokengraph.py to transfer opengnn to subtokengraph. But score is very low too.

I have sent a email to author and if the author reply me, I will let you know as soon as possible.

villmow commented 4 years ago

Hi, I want to replicate the results as well. Could you provide instructions or a trained checkpoint of the java-small model? Thanks!

BUCTwangkun commented 4 years ago

@hukuda222

When i use programgraphs2opengnn.py to transfer programgraphs to opengnn. I got the very low score. And then i use graph2subtokengraph.py to transfer opengnn to subtokengraph. But score is very low too.

I have sent a email to author and if the author reply me, I will let you know as soon as possible.

Hi，Have you solved the problem? I have the same problem. And I find the rouge score is always 0. And the predictions are always same invariant ，just like this: prediction: test performed target: set prediction: test performed target: set prediction: test performed target: cpy prediction: test performed target: len prediction: test performed target: len prediction: test performed target: to string prediction: test performed target: set euler angles prediction: test performed target: set euler angles rad prediction: test performed target: get gimbal pole prediction: test performed target: get roll rad prediction: test performed target: get roll prediction: test performed target: get pitch rad prediction: test performed target: get pitch prediction: test performed target: get yaw rad prediction: test performed target: get yaw

BUCTwangkun commented 4 years ago

@hukuda222

When i use programgraphs2opengnn.py to transfer programgraphs to opengnn. I got the very low score. And then i use graph2subtokengraph.py to transfer opengnn to subtokengraph. But score is very low too.

I have sent a email to author and if the author reply me, I will let you know as soon as possible.

Hi, have you solved the problem?

hukuda222 commented 4 years ago

@BUCTwangkun Sorry, I'm late. Not resolved yet.