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

Not use -1e4 as attn mask (#17306)

Browse Source 
* Use torch.finfo(self.dtype).min

* for GPTNeoX

* for Albert

* For Splinter

* Update src/transformers/models/data2vec/modeling_data2vec_audio.py

Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>

* fix -inf used in Bart-like models

* Fix a few remaining -inf

* more fix

* clean up

* For CLIP

* For FSMT

* clean up

* fix test

* Add dtype argument and use it for LayoutLMv3

* update FlaxLongT5Attention

Co-authored-by: ydshieh <ydshieh@users.noreply.github.com>
Co-authored-by: Patrick von Platen <patrick.v.platen@gmail.com>
This commit is contained in:
Yih-Dar
2022-06-20 16:16:16 +02:00
committed by GitHub
parent fdb120805c
commit d3cb28886a
54 changed files with 172 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
templates/adding_a_new_model/cookiecutter-template-{{cookiecutter.modelname}}/modeling_{{cookiecutter.lowercase_modelname}}.py
View File

@@ -1632,7 +1632,7 @@ def _make_causal_mask(input_ids_shape: torch.Size, dtype: torch.dtype, past_key_
Make causal mask used for bi-directional self-attention.
"""
bsz, tgt_len = input_ids_shape
mask = torch.full((tgt_len, tgt_len), float("-inf"))
mask = torch.full((tgt_len, tgt_len), torch.tensor(torch.finfo(dtype).min))
mask_cond = torch.arange(mask.size(-1))
mask.masked_fill_(mask_cond < (mask_cond + 1).view(mask.size(-1), 1), 0)
mask = mask.to(dtype)