GithubHelp home page GithubHelp logo

This code is for paper "Explainable Natural Language Inference via Identifying Important Rationales".

  1. To run this code, you should obtain the data from https://github.com/OanaMariaCamburu/e-SNLI, and put the datas in the 'datas/raw_datas/' directory. Then process these data by run the 'process_data.py';

  2. go to the file of 'rationale_extra' and run the train.py and then the predict.py to predict the rationales;

  3. go to the file of 'rationales_sel' and run the main_selector.py to selector the rationales;

  4. go to the file of 'generator' and

    4.1 finetune the GPT2 model: run the 'prepare_data_for_finetune.py' and 'GPT2_finetune_lm.py';

    4.2 generate the NLEs by the finetuned model: run the 'prepare_data_for_generation.py' and 'GPT2_generate.py';

  5. go to the file of 'classify', and run the 'prepare_data_task_pred.py' to prepare data for train. Then, run the 'main_classify.py' to obtain the prediction results.

Note: All parameters use default values, and you can also set them at runtime.

After running the above code correctly, you will get the following files in the 'datas/snli_data_dir/' directory

 --train.json
 --dev.json
 --test.json
 --train-rationales.json
 --dev-rationales.json
 --test-rationales.json
 --train-with-hints.json
 --dev-with-hints.json
 --test-with-hints.json
 --train-prompts.txt  
 --train-prompts.txt
 --train-prompts.txt
 --train-finetune.txt
 --dev-finetune.txt
 --train-gen.txt
 --dev-gen.txt
 --test-gen.txt
 --train-r-with-e.txt
 --dev-r-with-e.txt
 --test-r-with-e.txt

boazyang1's Projects

exp-nli icon exp-nli

Config files for my GitHub profile.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.