Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in
Toggle navigation
Menu
Open sidebar
chenpangpang
transformers
Commits
78a57c5e
Unverified
Commit
78a57c5e
authored
Apr 30, 2024
by
Joao Gante
Committed by
GitHub
Apr 30, 2024
Browse files
DBRX: make fixup (#30578)
parent
1bff6a0b
Changes
1
Show whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
5 additions
and
1 deletion
+5
-1
src/transformers/models/dbrx/modeling_dbrx.py
src/transformers/models/dbrx/modeling_dbrx.py
+5
-1
No files found.
src/transformers/models/dbrx/modeling_dbrx.py
View file @
78a57c5e
...
...
@@ -1215,6 +1215,7 @@ class DbrxModel(DbrxPreTrainedModel):
# KV cache is used. This is an issue for torch.compile which then recaptures cudagraphs at each decode steps due to the dynamic shapes.
# (`recording cudagraph tree for symint key 13`, etc.), which is VERY slow. A workaround is `@torch.compiler.disable`, but this prevents using
# `fullgraph=True`. See more context in https://github.com/huggingface/transformers/pull/29114
if
self
.
config
.
_attn_implementation
==
"flash_attention_2"
:
if
attention_mask
is
not
None
and
0.0
in
attention_mask
:
return
attention_mask
...
...
@@ -1227,7 +1228,10 @@ class DbrxModel(DbrxPreTrainedModel):
using_static_cache
=
isinstance
(
past_key_values
,
StaticCache
)
if
self
.
config
.
_attn_implementation
==
"sdpa"
and
not
using_static_cache
:
if
AttentionMaskConverter
.
_ignore_causal_mask_sdpa
(
attention_mask
,
inputs_embeds
=
input_tensor
,
past_key_values_length
=
past_seen_tokens
attention_mask
,
inputs_embeds
=
input_tensor
,
past_key_values_length
=
past_seen_tokens
,
is_training
=
self
.
training
,
):
return
None
...
...
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment