Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[ProphetNet] Bart-like Refactor #10501

Merged
merged 8 commits into from
Mar 4, 2021
Merged

[ProphetNet] Bart-like Refactor #10501

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
52e76af
first step to refactor
patrickvonplaten Mar 3, 2021
d4d25ab
make all fast tests pass
patrickvonplaten Mar 3, 2021
622a55e
make all slow tests pass
patrickvonplaten Mar 3, 2021
0203a86
save intermediate
patrickvonplaten Mar 4, 2021
4708f05
Merge branch 'refactor_prophetnet' of https://github.com/patrickvonpl…
patrickvonplaten Mar 4, 2021
3d48a3f
correct cache
patrickvonplaten Mar 4, 2021
54fde0d
finish PR
patrickvonplaten Mar 4, 2021
5271222
make fp16 work
patrickvonplaten Mar 4, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
6 changes: 6 additions & 0 deletions src/transformers/models/prophetnet/configuration_prophetnet.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,6 +92,8 @@ class ProphetNetConfig(PretrainedConfig):
smoothing is performed.
use_cache (:obj:`bool`, `optional`, defaults to :obj:`True`):
Whether or not the model should return the last key/values attentions (not used by all models).
gradient_checkpointing (:obj:`bool`, `optional`, defaults to :obj:`False`):
If True, use gradient checkpointing to save memory at the expense of slower backward pass.
"""
model_type = "prophetnet"
keys_to_ignore_at_inference = ["past_key_values"]
Expand Down Expand Up @@ -119,6 +121,7 @@ def __init__(
num_buckets=32,
relative_max_distance=128,
disable_ngram_loss=False,
gradient_checkpointing=False,
eps=0.0,
use_cache=True,
pad_token_id=0,
Expand Down Expand Up @@ -161,6 +164,9 @@ def __init__(

self.use_cache = use_cache

# 4 Training Args (should be removed soon)
self.gradient_checkpointing = gradient_checkpointing

@property
def num_attention_heads(self) -> int:
return self.num_encoder_attention_heads
Expand Down