Skip to content

Tk-Instruct is a Transformer model that is tuned to solve many NLP tasks by following instructions.

License

Notifications You must be signed in to change notification settings

jiashenggu/Tk-Instruct

 
 

Repository files navigation

There are three tasks in total

Task 1

The initial task is to modify the perturb_method parameter in the scripts eval_tk_instruct_english_perturb.sh and eval_tk_instruct_xlingual_perturb.sh. After changing this parameter, execute the scripts to run experiments. You can find the specific perturbation methods in src/ni_dataset_perturb.py. To ensure reproducible results, each perturbation method should be run three times with the seed set to 1, 2, and 3, respectively.

Task 2

The second task is to conduct experiments on induction_data. This will involve modifying the run files in the scripts eval_tk_instruct_english_orignal.sh and eval_tk_instruct_xlingual_orignal.sh. You will need to change these scripts to point to Tk-Instruct/src/run_s2s_induction.py, then run them.

Task 3

We constructed a dataset named Para-Instructions that contains multiple human-oriented instructions for each task.

Para-Instructions are files tmp/new_instruction_*.csv.

The third task is to test Para-Instructions. You'll need to replace the output in tmp/new_instruction.ipynb with data/splits/default/dev_tasks.txt. Use the suffix for the new definition and no suffix for the original one. Then, execute scripts/eval_tk_instruct_english_orignal.sh and conduct an experiment to compare the results between the new and original definitions.

Citation

@misc{gu2023robustness,
      title={Robustness of Learning from Task Instructions}, 
      author={Jiasheng Gu and Hongyu Zhao and Hanzi Xu and Liangyu Nie and Hongyuan Mei and Wenpeng Yin},
      year={2023},
      eprint={2212.03813},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

Forked from GitHub - yizhongw/Tk-Instruct: Tk-Instruct is a Transformer model that is tuned to solve many NLP tasks by following instructions.

About

Tk-Instruct is a Transformer model that is tuned to solve many NLP tasks by following instructions.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 64.7%
  • Jupyter Notebook 31.5%
  • Shell 3.6%
  • Dockerfile 0.2%