rnncodeclones's Introduction
rnncodeclones's People
rnncodeclones's Issues
[TreeMutator] Clone and train word2vec on intellij Idea Community repo
Clone repo with java, not manually clonning.
After training - delete repo
[RNN] Data sets
Create data sets for Rnn training
[RNN] Create RNN model with tensorflow
Create RNN model with python and tensorflow. Train it with seq2seq outputs
[Project] Think about code refactoring
It's separated now. There is word2vec in .jar and seq2seq in .py. Will it be better to combine word2vec and seq2seq in .jar and serialize vector after seq2seq? RNN will be in .py
Sort word2vec weights
Sort word2vec weights according to tokens indicies
[seq2seq] Loss more than 2
Loss of seq2seq model is more than 2 when sequence has big length. Have to understand how to fix it
[TreeMutator] Rename TreeMutator
It's not only mutates tree. It also trains word2vec model and creates list of used tokens indices
[RNNCodeClones] Answer representation
Have to make user-friendly answer representation. Current answer has form like: clones are vectors [indicies]
[seq2seq] Display loss during training seq2seq
For easy understanding what is going on, it's need to display loss near epochs amount
[TreeMutator] Figure out where clone and where not
Have to figure out, where mutated code is clone and where - not
[Project] Merge
Merge branches
Deal with seq2seq model
Do I need seq2seq using if I send vector with tokens indicies instead of token weights
[seq2seq] Python seq2seq model with pretrained word2vec embeddings
Create seq2seq model from pretrained word2vec embedding and list of token indices
[TreeMutator] Paths to python scripts from .jar
Figure out how to specify path to python scripts. Tried to do it with resources in maven. During execution of this file itβs an error. File from .jar is not an ordinary file. Could be get with Stream. But this solution is not good. This means that for execution file have to be get as InputStream and saved to tmp file or executed like a string
[RNN] Contrasitive loss function fixing
Loss function has to be fixed. It remains 0 after every step
[TreeMutator] Create vector for results
Besides creation origin and mutated indicies, create vector with prediction results. For example, 1 for clones and 0 for non-clones
[seq2seq] Get encoder_output vector from seq2seq
[RNNCodeClones] Add arguments like `--train` or `--eval`
For using this util it'll be better to add some args
[seq2seq] Make code more convenient
Instead of solid code create methods and classes
[RNN] GRU or LSTM?
Select one RNN: GRU or LSTM
Serialize word2vec weights
Serialize in java list of weights from word2vec model.
[TreeMutator] Change mutation logic
Instead of copying/deleting one line, generate rnd number for amount of lines. After that work with them
[RNN] Difference between CPU & GPU
Measure evaluation time with GPU using and CPU
[seq2seq] Serialization seq2seq training
Have to serialize trained seq2seq with tensorflow
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. πππ
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.