Sign upLogin

tensorflow/models

View on GitHub
Star
official/nlp/modeling/layers/transformer_encoder_block_test.py

Summary

Maintainability
F
2 wks
Test Coverage

File transformer_encoder_block_test.py has 582 lines of code (exceeds 250 allowed). Consider refactoring.
Open

# Copyright 2024 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
Severity: Major
Found in official/nlp/modeling/layers/transformer_encoder_block_test.py - About 1 day to fix

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

      def test_layer_invocation_with_float16_dtype(self, transformer_cls):
    tf_keras.mixed_precision.set_global_policy('mixed_float16')
    test_layer = transformer_cls(
        num_attention_heads=10, inner_dim=2048, inner_activation='relu')
    sequence_length = 21
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 1 day to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 149..175
    official/projects/roformer/roformer_encoder_block_test.py on lines 184..208

    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 167.

    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

      @parameterized.named_parameters(('value_dim_is_none', None, 128, 2, 128 // 2),
                                  ('value_dim_is_not_none', 30, 128, 2, 30))
  def test_value_dim(self, value_dim, q_tensor_last_dim,
                     some_num_attention_heads, expected):
    some_inner_dim = 32
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 1 day to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 341..359

    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 163.

    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

      @parameterized.named_parameters(('key_dim_is_none', None, 128, 2, 128 // 2),
                                  ('key_dim_is_not_none', 30, 128, 2, 30))
  def test_key_dim(self, key_dim, q_tensor_last_dim, some_num_attention_heads,
                   expected):
    some_inner_dim = 32
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 1 day to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 387..405

    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 163.

    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 2 locations. Consider refactoring.
    Open

      def test_layer_output_range_without_mask(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10,
        inner_dim=2048,
        inner_activation='relu',
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 1 day to fix
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 126..149

    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 162.

    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 4 locations. Consider refactoring.
    Open

      def test_layer_invocation_with_mask(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10, inner_dim=2048, inner_activation='relu')
    sequence_length = 21
    width = 80
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 1 day to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 94..119
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 76..99
    official/projects/roformer/roformer_encoder_block_test.py on lines 75..98

    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 156.

    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_use_rms_norm(self):
    num_attention_heads = 2
    hidden_size = 16
    encoder_block = TransformerEncoderBlock(
        num_attention_heads=num_attention_heads,
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 1 day to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 410..430

    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 142.

    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_use_bias_norm_first(self):
    num_attention_heads = 2
    hidden_size = 16
    encoder_block = TransformerEncoderBlock(
        num_attention_heads=num_attention_heads,
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 1 day to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 432..452

    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 142.

    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 3 locations. Consider refactoring.
    Open

      def test_separate_qkv(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=2,
        inner_dim=128,
        inner_activation='relu',
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 1 day to fix
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 210..222
    official/projects/roformer/roformer_encoder_block_test.py on lines 224..236

    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 132.

    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

      def test_dynamic_layer_sequence(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10,
        inner_dim=2048,
        inner_activation='relu',
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 1 day to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 191..207
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 192..208

    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 126.

    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 4 locations. Consider refactoring.
    Open

      def test_layer_invocation(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10, inner_dim=2048, inner_activation='relu')
    sequence_length = 21
    width = 80
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 7 hrs to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 73..92
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 57..74
    official/projects/roformer/roformer_encoder_block_test.py on lines 56..73

    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 110.

    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 4 locations. Consider refactoring.
    Open

      def test_layer_creation_with_mask(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10, inner_dim=2048, inner_activation='relu')
    sequence_length = 21
    width = 80
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 6 hrs to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 44..57
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 44..55
    official/projects/roformer/roformer_encoder_block_test.py on lines 43..54

    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 100.

    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 4 locations. Consider refactoring.
    Open

      def test_transform_with_initializer(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10,
        inner_dim=2048,
        inner_activation='relu',
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 5 hrs to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 177..189
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 178..190
    official/projects/roformer/roformer_encoder_block_test.py on lines 210..222

    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 94.

    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 4 locations. Consider refactoring.
    Open

      def test_layer_creation(self, transformer_cls):
    test_layer = transformer_cls(
        num_attention_heads=10, inner_dim=2048, inner_activation='relu')
    sequence_length = 21
    width = 80
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 4 hrs to fix
    official/nlp/modeling/layers/tn_transformer_test.py on lines 31..42
    official/projects/qat/nlp/modeling/layers/transformer_encoder_block_test.py on lines 33..42
    official/projects/roformer/roformer_encoder_block_test.py on lines 32..41

    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 81.

    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

      def test_value_dim_is_part_of_config_2(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    value_dim = 10
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 4 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 525..536
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 573..584

    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 80.

    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

      def test_key_dim_is_part_of_config_2(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    key_dim = 10
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 4 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 549..560
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 573..584

    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 80.

    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

      def test_output_last_dim_is_part_of_config_2(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    output_last_dim = 10
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 4 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 525..536
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 549..560

    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 80.

    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_diff_q_kv_att_layer_norm_is_part_of_config_1(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 4 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 502..512

    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 74.

    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_use_query_residual_is_part_of_config_2(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 1 other location - About 4 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 466..476

    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 74.

    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 4 locations. Consider refactoring.
    Open

      def test_output_last_dim_is_part_of_config_1(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 3 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 491..500
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 514..523
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 538..547

    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

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

      def test_use_query_residual_is_part_of_config_1(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 3 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 514..523
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 538..547
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 562..571

    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

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

      def test_value_dim_is_part_of_config_1(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 3 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 491..500
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 514..523
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 562..571

    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

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

      def test_key_dim_is_part_of_config_1(self):
    some_num_attention_heads = 2
    some_inner_dim = 32
    some_inner_activation = 'relu'
    encoder = TransformerEncoderBlock(
    Severity: Major
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 3 other locations - About 3 hrs to fix
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 491..500
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 538..547
    official/nlp/modeling/layers/transformer_encoder_block_test.py on lines 562..571

    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 3 locations. Consider refactoring.
    Open

      @parameterized.parameters({'attention_axes': None}, {'attention_axes': [1]},
                            {'attention_axes': [2]}, {'attention_axes': [1, 2]})
    Severity: Minor
    Found in official/nlp/modeling/layers/transformer_encoder_block_test.py and 2 other locations - About 35 mins to fix
    official/nlp/modeling/layers/reuse_transformer_test.py on lines 310..311
    official/projects/roformer/roformer_encoder_block_test.py on lines 298..299

    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 33.

    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