Skip to content

[ConvBert] Fix #21523 #21849

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

Merged
ArthurZucker merged 6 commits into from
Mar 1, 2023
Merged

[ConvBert] Fix #21523 #21849

Show file tree
Hide file tree
Changes from 5 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
7f36153
fix reshaping
ArthurZucker Feb 28, 2023
b8f1132
add test
ArthurZucker Feb 28, 2023
85847cf
styling
ArthurZucker Feb 28, 2023
3d3da4c
last fixes
ArthurZucker Feb 28, 2023
b455124
Update src/transformers/models/convbert/modeling_convbert.py
ArthurZucker Feb 28, 2023
916bf96
code quallity
ArthurZucker Mar 1, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions src/transformers/models/convbert/modeling_convbert.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -316,7 +316,7 @@ def __init__(self, config):
if config.hidden_size % self.num_attention_heads != 0:
raise ValueError("hidden_size should be divisible by num_attention_heads")

self.attention_head_size = config.hidden_size // config.num_attention_heads
self.attention_head_size = (config.hidden_size // self.num_attention_heads) // 2
self.all_head_size = self.num_attention_heads * self.attention_head_size

self.query = nn.Linear(config.hidden_size, self.all_head_size)
Expand Down Expand Up @@ -413,7 +413,8 @@ def forward(
conv_out = torch.reshape(conv_out_layer, [batch_size, -1, self.num_attention_heads, self.attention_head_size])
context_layer = torch.cat([context_layer, conv_out], 2)

new_context_layer_shape = context_layer.size()[:-2] + (self.head_ratio * self.all_head_size,)
# conv and context
new_context_layer_shape = context_layer.size()[:-2] + (self.num_attention_heads * self.attention_head_size * 2,)
context_layer = context_layer.view(*new_context_layer_shape)

outputs = (context_layer, attention_probs) if output_attentions else (context_layer,)
Expand Down
5 changes: 5 additions & 0 deletions tests/models/convbert/test_modeling_convbert.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -446,6 +446,11 @@ def test_model_for_input_embeds(self):
result = model(inputs_embeds=inputs_embeds)
self.assertEqual(result.last_hidden_state.shape, (batch_size, seq_length, config.hidden_size))

def test_reducing_attention_heads(self):
config, *inputs_dict = self.model_tester.prepare_config_and_inputs()
config.head_ratio = 4
self.model_tester.create_and_check_for_masked_lm(config, *inputs_dict)


@require_torch
class ConvBertModelIntegrationTest(unittest.TestCase):
Expand Down