README.md

metadata

task_categories:
  - translation
language:
  - pl
  - de
size_categories:
  - 1K<n<10K
configs:
  - config_name: v0
    data_files:
      - split: test
        path: data/v0_pl-de-outputs_modified.csv
  - config_name: v0-blank-discard
    data_files:
      - split: test
        path: data/v0_pl-de-outputs.csv

Dataset structure

polish column is the Polish name of a place. It's the input for both the models, which is translated to produce finetuned_output and basemodel_output.

german column is the actual German name of a place.

Outputs of Polish→German Toponym Translation model.

V0 subset, "finetuned_output" column is outputs of this model → https://huggingface.co/DebasishDhal99/polish-to-german-toponym-model-opus-mt-pl-de (Fine-tuned specifically for Toponym translation)

V0 subset, "basemodel_output" column is outputs of this model → https://huggingface.co/Helsinki-NLP/opus-mt-pl-de