SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Fix pad to multiple of (#25732)

Browse Source 
* nits

* update the test

* nits

* update

* fix bark

* fix bark tests and allow padding to multiple of without new tokens
This commit is contained in:
Arthur
2023-09-15 11:53:39 -04:00
committed by GitHub
parent ebd21e904f
commit eb644980eb
3 changed files with 51 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
src/transformers/modeling_utils.py
View File

@@ -1436,7 +1436,8 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
vectors at the end. Reducing the size will remove vectors from the end. If not provided or `None`, just
returns a pointer to the input tokens `torch.nn.Embedding` module of the model without doing anything.
pad_to_multiple_of (`int`, *optional*):
If set will pad the embedding matrix to a multiple of the provided value.
If set will pad the embedding matrix to a multiple of the provided value.If `new_num_tokens` is set to
`None` will just pad the embedding to a multiple of `pad_to_multiple_of`.
This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability
`>= 7.5` (Volta), or on TPUs which benefit from having sequence lengths be a multiple of 128. For more
@@ -1447,12 +1448,12 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
`torch.nn.Embedding`: Pointer to the input tokens Embeddings Module of the model.
"""
model_embeds = self._resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
if new_num_tokens is None:
if new_num_tokens is None and pad_to_multiple_of is None:
return model_embeds
# Update base model and current model config
self.config.vocab_size = new_num_tokens
self.vocab_size = new_num_tokens
self.config.vocab_size = model_embeds.weight.shape[0]
self.vocab_size = model_embeds.weight.shape[0]
# Tie weights again if needed
self.tie_weights()
@@ -1508,7 +1509,8 @@ class PreTrainedModel(nn.Module, ModuleUtilsMixin, GenerationMixin, PushToHubMix
vectors from the end. If not provided or `None`, just returns a pointer to the input tokens
`torch.nn.Embedding` module of the model without doing anything.
pad_to_multiple_of (`int`, *optional*):
If set will pad the embedding matrix to a multiple of the provided value.
If set will pad the embedding matrix to a multiple of the provided value. If `new_num_tokens` is set to
`None` will just pad the embedding to a multiple of `pad_to_multiple_of`.
This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability
`>= 7.5` (Volta), or on TPUs which benefit from having sequence lengths be a multiple of 128. For more

46
src/transformers/models/bark/modeling_bark.py
View File

@@ -1086,21 +1086,57 @@ class BarkFineModel(BarkPreTrainedModel):
]
)
self.set_input_embeddings(new_embeddings_list)
new_num_tokens = [embed.weight.shape[0] for embed in new_embeddings_list]
new_num_tokens = new_embeddings_list[0].weight.shape[0]
# if word embeddings are not tied, make sure that lm head is resized as well
if self.get_output_embeddings() is not None and not self.config.tie_word_embeddings:
old_lm_head_list = self.get_output_embeddings()
new_lm_head_list = nn.ModuleList(
[
self._get_resized_lm_head(old_lm_head, new_num_token)
for old_lm_head, new_num_token in zip(old_lm_head_list, new_num_tokens)
]
[self._get_resized_lm_head(old_lm_head, new_num_tokens) for old_lm_head in old_lm_head_list]
)
self.set_output_embeddings(new_lm_head_list)
return self.get_input_embeddings()
def resize_token_embeddings(
self, new_num_tokens: Optional[int] = None, pad_to_multiple_of: Optional[int] = None
) -> nn.Embedding:
"""
Resizes input token embeddings matrix of the model if `new_num_tokens != config.vocab_size`.
Takes care of tying weights embeddings afterwards if the model class has a `tie_weights()` method.
Arguments:
new_num_tokens (`int`, *optional*):
The number of new tokens in the embedding matrix. Increasing the size will add newly initialized
vectors at the end. Reducing the size will remove vectors from the end. If not provided or `None`, just
returns a pointer to the input tokens `torch.nn.Embedding` module of the model without doing anything.
pad_to_multiple_of (`int`, *optional*):
If set will pad the embedding matrix to a multiple of the provided value.
This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability
`>= 7.5` (Volta), or on TPUs which benefit from having sequence lengths be a multiple of 128. For more
details about this, or help on choosing the correct value for resizing, refer to this guide:
https://docs.nvidia.com/deeplearning/performance/dl-performance-matrix-multiplication/index.html#requirements-tc
Return:
`torch.nn.Embedding`: Pointer to the input tokens Embeddings Module of the model.
"""
model_embeds = self._resize_token_embeddings(new_num_tokens, pad_to_multiple_of)
if new_num_tokens is None and pad_to_multiple_of is None:
return model_embeds
# Update base model and current model config
self.config.output_vocab_size = model_embeds[0].weight.shape[0]
self.config.vocab_size = model_embeds[0].weight.shape[0]
self.output_vocab_size = model_embeds[0].weight.shape[0]
self.vocab_size = model_embeds[0].weight.shape[0]
# Tie weights again if needed
self.tie_weights()
return model_embeds
def tie_weights(self):
"""
Tie the weights between the input embeddings list and the output embeddings list.

3
tests/test_modeling_common.py
View File

@@ -1431,6 +1431,9 @@ class ModelTesterMixin:
model_embed = model.resize_token_embeddings(model_vocab_size, pad_to_multiple_of=64)
self.assertTrue(model_embed.weight.shape[0] // 64, 0)
self.assertTrue(model_embed.weight.shape[0], model.config.vocab_size)
self.assertTrue(model.config.vocab_size, model.vocab_size)
model_embed = model.resize_token_embeddings(model_vocab_size + 13, pad_to_multiple_of=64)
self.assertTrue(model_embed.weight.shape[0] // 64, 0)