example.sh 1.4 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051
  1. #!/bin/usr/env sh
  2. # Copyright (c) 2018-present, Facebook, Inc.
  3. # All rights reserved.
  4. #
  5. # This source code is licensed under the license found in the
  6. # LICENSE file in the root directory of this source tree.
  7. set -e
  8. s=${1:-en}
  9. t=${2:-es}
  10. echo "Example based on the ${s}->${t} alignment"
  11. if [ ! -d data/ ]; then
  12. mkdir -p data;
  13. fi
  14. if [ ! -d res/ ]; then
  15. mkdir -p res;
  16. fi
  17. dico_train=data/${s}-${t}.0-5000.txt
  18. if [ ! -f "${dico_train}" ]; then
  19. DICO=$(basename -- "${dico_train}")
  20. wget -c "https://dl.fbaipublicfiles.com/arrival/dictionaries/${DICO}" -P data/
  21. fi
  22. dico_test=data/${s}-${t}.5000-6500.txt
  23. if [ ! -f "${dico_test}" ]; then
  24. DICO=$(basename -- "${dico_test}")
  25. wget -c "https://dl.fbaipublicfiles.com/arrival/dictionaries/${DICO}" -P data/
  26. fi
  27. src_emb=data/wiki.${s}.vec
  28. if [ ! -f "${src_emb}" ]; then
  29. EMB=$(basename -- "${src_emb}")
  30. wget -c "https://dl.fbaipublicfiles.com/fasttext/vectors-wiki/${EMB}" -P data/
  31. fi
  32. tgt_emb=data/wiki.${t}.vec
  33. if [ ! -f "${tgt_emb}" ]; then
  34. EMB=$(basename -- "${tgt_emb}")
  35. wget -c "https://dl.fbaipublicfiles.com/fasttext/vectors-wiki/${EMB}" -P data/
  36. fi
  37. output=res/wiki.${s}-${t}.vec
  38. python3 align.py --src_emb "${src_emb}" --tgt_emb "${tgt_emb}" \
  39. --dico_train "${dico_train}" --dico_test "${dico_test}" --output "${output}" \
  40. --lr 25 --niter 10
  41. python3 eval.py --src_emb "${output}" --tgt_emb "${tgt_emb}" \
  42. --dico_test "${dico_test}"