change

2. ./run_essay.sh (if you don't have slurm, just adapt accordingly)

3. evaluate predictions: python run_eval.py

That's it.

That's it. But here's also an 

###In-depth explanation of the main run-command example

```

python arg_rel_plausibility_ranker.py ESSAY bert 0 -sup_connect "Moreover+," -att_connect "However+," -maxlen=50 --use_markers

ESSAY: refers to the paths in data_config.py

bert: use bert (or "elmo")... other languade models are easy to integrate thanks to allennlp/huggingface

sup_connect: connector to signal support (you can put anything here what you want, + means whitespave)

att_connect: connector to signal attack (you can put anything here what you want)

use_markers: shall we use marker embeddings to show the model what part of text belongs to src or tgt unit?

```

## FAQ: