translate-package 0.0.8__tar.gz → 0.0.9__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (32) hide show
  1. {translate_package-0.0.8 → translate_package-0.0.9}/PKG-INFO +1 -1
  2. {translate_package-0.0.8 → translate_package-0.0.9}/setup.py +1 -1
  3. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/__init__.py +1 -1
  4. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/machine_translation.py +2 -1
  5. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/load_tokenizer.py +7 -2
  6. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/PKG-INFO +1 -1
  7. {translate_package-0.0.8 → translate_package-0.0.9}/setup.cfg +0 -0
  8. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/data/__init__.py +0 -0
  9. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/data/__pycache__/__init__.cpython-310.pyc +0 -0
  10. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/data/__pycache__/data_preparation.cpython-310.pyc +0 -0
  11. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/data/data_preparation.py +0 -0
  12. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/errors/__init__.py +0 -0
  13. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/errors/__pycache__/__init__.cpython-310.pyc +0 -0
  14. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/__init__.py +0 -0
  15. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/__pycache__/__init__.cpython-310.pyc +0 -0
  16. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/__pycache__/code_generation.cpython-310.pyc +0 -0
  17. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/__pycache__/machine_translation.cpython-310.pyc +0 -0
  18. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/gradient_observation.py +0 -0
  19. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/lstm.py +0 -0
  20. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/__init__.py +0 -0
  21. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/__pycache__/__init__.cpython-310.pyc +0 -0
  22. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/__pycache__/load_tokenizer.cpython-310.pyc +0 -0
  23. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/__pycache__/train_tokenizer.cpython-310.pyc +0 -0
  24. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/train_tokenizer.py +0 -0
  25. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/utils/__init__.py +0 -0
  26. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/utils/__pycache__/__init__.cpython-310.pyc +0 -0
  27. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/utils/__pycache__/checkpoint.cpython-310.pyc +0 -0
  28. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package/utils/checkpoint.py +0 -0
  29. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/SOURCES.txt +0 -0
  30. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/dependency_links.txt +0 -0
  31. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/requires.txt +0 -0
  32. {translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/top_level.txt +0 -0
{translate_package-0.0.8 → translate_package-0.0.9}/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: translate_package
3
- Version: 0.0.8
3
+ Version: 0.0.9
4
4
  Summary: Contain functions and classes to efficiently train a sequence to sequence to translate between two languages.
5
5
  Author: Oumar Kane
6
6
  Author-email: oumar.kane@univ-thies.sn
{translate_package-0.0.8 → translate_package-0.0.9}/setup.py RENAMED
@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
2
2
 
3
3
  setup(
4
4
  name="translate_package",
5
- version="0.0.8",
5
+ version="0.0.9",
6
6
  author="Oumar Kane",
7
7
  packages=find_packages(),
8
8
  author_email="oumar.kane@univ-thies.sn",
{translate_package-0.0.8 → translate_package-0.0.9}/translate_package/__init__.py RENAMED
@@ -14,7 +14,7 @@ from transformers import (
14
14
  Trainer, AutoModelForSeq2SeqLM,
15
15
  get_linear_schedule_with_warmup,
16
16
  T5ForConditionalGeneration, Adafactor, BartForConditionalGeneration,
17
- MT5ForConditionalGeneration, AdamWeightDecay
17
+ MT5ForConditionalGeneration, AdamWeightDecay, AutoTokenizer
18
18
  )
19
19
  from wolof_translate.utils.bucket_iterator import SequenceLengthBatchSampler, BucketSampler
20
20
  from wolof_translate.utils.sent_transformers import TransformerSequences
{translate_package-0.0.8 → translate_package-0.0.9}/translate_package/models/machine_translation.py RENAMED
@@ -11,6 +11,7 @@ from translate_package import (
11
11
  MT5ForConditionalGeneration,
12
12
  BartForConditionalGeneration,
13
13
  AutoModelForSeq2SeqLM,
14
+ AutoTokenizer,
14
15
  Adafactor,
15
16
  AdamWeightDecay
16
17
  )
@@ -57,7 +58,7 @@ class MachineTranslationTransformer(pl.LightningModule):
57
58
  hidden_size=128,
58
59
  dropout=0.1,
59
60
  bidirectional=False,
60
- length_penalty=1.2
61
+ length_penalty=1.2,
61
62
  ):
62
63
 
63
64
  super().__init__()
{translate_package-0.0.8 → translate_package-0.0.9}/translate_package/tokenization/load_tokenizer.py RENAMED
@@ -3,13 +3,18 @@ from transformers import T5TokenizerFast
3
3
  from transformers import AutoTokenizer
4
4
  import os
5
5
 
6
- def load_tokenizer(tokenizer_name, model, dir_path, file_name, model_name = None):
6
+ BCP_47_languages = {
7
+ 'french': 'fra_Latn',
8
+ 'wolof': 'wol_Latn',
9
+ }
10
+
11
+ def load_tokenizer(tokenizer_name, model, dir_path, file_name, model_name = None, src_lang = "french", tgt_lang = "wolof"):
7
12
 
8
13
  if model == "nllb":
9
14
 
10
15
  if not model_name is None:
11
16
 
12
- tokenizer = AutoTokenizer.from_pretrained(model_name)
17
+ tokenizer = AutoTokenizer.from_pretrained(model_name, src_lang = BCP_47_languages[src_lang], tgt_lang = BCP_47_languages[tgt_lang])
13
18
 
14
19
  print(f"The {model}'s tokenizer was successfully loaded")
15
20
 
{translate_package-0.0.8 → translate_package-0.0.9}/translate_package.egg-info/PKG-INFO RENAMED
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.1
2
2
  Name: translate-package
3
- Version: 0.0.8
3
+ Version: 0.0.9
4
4
  Summary: Contain functions and classes to efficiently train a sequence to sequence to translate between two languages.
5
5
  Author: Oumar Kane
6
6
  Author-email: oumar.kane@univ-thies.sn