riaz.somc/transformers
1
0
Fork 0
mirror of https://github.com/huggingface/transformers.git synced 2025-07-31 10:12:23 +06:00
Code Issues Actions 23 Packages Projects Releases Wiki Activity

[UdopTokenizer] Fix post merge imports (#29451)

Browse Source 
* update

* ...

* nits

* arf

* 🧼

* beat the last guy

* style everyone
This commit is contained in:
Arthur 2024-03-05 09:42:52 +01:00 committed by GitHub
parent fa7f3cf336
commit 132852203a
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 16 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/transformers/models/udop/tokenization_udop.py
View File

@ -157,12 +157,6 @@ PRETRAINED_VOCAB_FILES_MAP = {
}
# TODO(PVP) - this should be removed in Transformers v5
PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
"microsoft/udop-large": 512,
}
class UdopTokenizer(PreTrainedTokenizer):
"""
Adapted from [`LayoutXLMTokenizer`] and [`T5Tokenizer`]. Based on
@ -256,7 +250,6 @@ class UdopTokenizer(PreTrainedTokenizer):
vocab_files_names = VOCAB_FILES_NAMES
pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
model_input_names = ["input_ids", "attention_mask"]
def __init__(

17
src/transformers/models/udop/tokenization_udop_fast.py
View File

@ -29,11 +29,6 @@ from ...tokenization_utils_base import (
)
from ...tokenization_utils_fast import PreTrainedTokenizerFast
from ...utils import PaddingStrategy, TensorType, add_end_docstrings, is_sentencepiece_available, logging
from ..udop.tokenization_udop import (
PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES,
PRETRAINED_VOCAB_FILES_MAP,
VOCAB_FILES_NAMES,
)
if is_sentencepiece_available():
@ -42,6 +37,17 @@ else:
UdopTokenizer = None
VOCAB_FILES_NAMES = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.json"}
PRETRAINED_VOCAB_FILES_MAP = {
"vocab_file": {
"microsoft/udop-large": "https://huggingface.co/microsoft/udop-large/resolve/main/spiece.model",
},
"tokenizer_file": {
"microsoft/udop-large": "https://huggingface.co/microsoft/udop-large/resolve/main/tokenizer.json",
},
}
logger = logging.get_logger(__name__)
UDOP_ENCODE_KWARGS_DOCSTRING = r"""
@ -197,7 +203,6 @@ class UdopTokenizerFast(PreTrainedTokenizerFast):
vocab_files_names = VOCAB_FILES_NAMES
pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
model_input_names = ["input_ids", "attention_mask"]
slow_tokenizer_class = UdopTokenizer

6
tests/models/udop/test_tokenization_udop.py
View File

@ -22,12 +22,12 @@ from typing import List
from transformers import (
AddedToken,
SpecialTokensMixin,
UdopTokenizer,
UdopTokenizerFast,
is_tf_available,
is_torch_available,
logging,
)
from transformers.models.udop.tokenization_udop import UdopTokenizer
from transformers.testing_utils import (
get_tests_dir,
is_pt_tf_cross_test,
@ -1717,6 +1717,10 @@ class UdopTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
def test_alignement_methods(self):
pass
@unittest.skip("#TODO will be removed in main")
def test_pretrained_model_lists(self):
pass
@unittest.skip("UDOP tokenizer requires boxes besides sequences.")
def test_maximum_encoding_length_pair_input(self):
pass