Sign upLogin

tensorflow/models

View on GitHub
Star
official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py

Summary

Maintainability
F
1 wk
Test Coverage

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  @parameterized.named_parameters(
      ("4d_inputs_1freebatch_mask2", [3, 4], [3, 2], [4, 2],
       (2,)), ("4d_inputs_1freebatch_mask3", [3, 4], [3, 2], [3, 4, 2], (2,)),
      ("4d_inputs_1freebatch_mask4", [3, 4], [3, 2], [3, 2, 4, 2],
       (2,)), ("4D_inputs_2D_attention", [3, 4], [3, 2], [3, 4, 3, 2], (1, 2)),
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 4 days to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 206..242

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 428.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_masked_attention_with_scores(self):
    """Test with a mask tensor."""
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=2, key_dim=2)
    # Create a 3-dimensional input (the first dimension is implicit).
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 3 days to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 161..204

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 351.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 3 locations. Consider refactoring.
Open

  @parameterized.named_parameters(
      ("key_value_same_proj", None, None, [40, 80]),
      ("key_value_different_proj", 32, 60, [40, 60]),
  )
  def test_non_masked_attention(self, value_dim, output_shape, output_dims):
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 2 other locations - About 1 day to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 27..42
official/nlp/modeling/layers/talking_heads_attention_test.py on lines 27..42

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 129.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_dropout(self):
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=2, key_dim=2, dropout=0.5)

    # Generate data for the input (non-mask) tensors.
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 7 hrs to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 244..258

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 117.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_attention_scores_with_values(self):
    """Test attention outputs with coefficients."""
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=12, key_dim=64)
    # Create a 3-dimensional input (the first dimension is implicit).
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 7 hrs to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 63..72

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 115.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_attention_scores(self):
    """Test attention outputs with coefficients."""
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=12, key_dim=64)
    # Create a 3-dimensional input (the first dimension is implicit).
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 6 hrs to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 53..61

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 99.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_initializer(self):
    """Test with a specified initializer."""
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=12,
        key_dim=64,
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 4 hrs to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 150..159

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 84.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Similar blocks of code found in 2 locations. Consider refactoring.
Open

  def test_non_masked_self_attention(self):
    """Test with one input (self-attenntion) and no mask tensor."""
    test_layer = edgetpu_layers.EdgeTPUMultiHeadAttention(
        num_heads=12, key_dim=64)
    # Create a 3-dimensional input (the first dimension is implicit).
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 1 other location - About 3 hrs to fix
official/nlp/modeling/layers/reuse_attention_test.py on lines 44..51

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 71.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

Identical blocks of code found in 4 locations. Consider refactoring.
Open

    if use_bias:
      self.assertLen(test_layer._query_dense.trainable_variables, 2)
      self.assertLen(test_layer._output_dense.trainable_variables, 2)
    else:
      self.assertLen(test_layer._query_dense.trainable_variables, 1)
Severity: Major
Found in official/projects/edgetpu/nlp/modeling/edgetpu_layers_test.py and 3 other locations - About 2 hrs to fix
official/nlp/modeling/layers/block_sparse_attention_test.py on lines 127..132
official/nlp/modeling/layers/multi_query_attention_test.py on lines 150..155
official/nlp/modeling/layers/talking_heads_attention_test.py on lines 109..114

Duplicated Code

Duplicated code can lead to software that is hard to understand and difficult to change. The Don't Repeat Yourself (DRY) principle states:

Every piece of knowledge must have a single, unambiguous, authoritative representation within a system.

When you violate DRY, bugs and maintenance problems are sure to follow. Duplicated code has a tendency to both continue to replicate and also to diverge (leaving bugs as two similar implementations differ in subtle ways).

Tuning

This issue has a mass of 61.

We set useful threshold defaults for the languages we support but you may want to adjust these settings based on your project guidelines.

The threshold configuration represents the minimum mass a code block must have to be analyzed for duplication. The lower the threshold, the more fine-grained the comparison.

If the engine is too easily reporting duplication, try raising the threshold. If you suspect that the engine isn't catching enough duplication, try lowering the threshold. The best setting tends to differ from language to language.

See codeclimate-duplication's documentation for more information about tuning the mass threshold in your .codeclimate.yml.

Refactorings

Further Reading

There are no issues that match your filters.

Category
Status