facebookresearch / DisCo
File Size

The distribution of size of files (measured in lines of code).

Intro
  • File size measurements show the distribution of size of files.
  • Files are classified in four categories based on their size (lines of code): 1-100 (very small files), 101-200 (small files), 201-500 (medium size files), 501-1000 (long files), 1001+(very long files).
  • It is a good practice to keep files small. Long files may become "bloaters", code that have increased to such gargantuan proportions that they are hard to work with.
Learn more...
File Size Overall
  • There are 153 files with 14,239 lines of code.
    • 0 very long files (0 lines of code)
    • 1 long files (566 lines of code)
    • 16 medium size files (4,874 lines of codeclsfd_ftr_w_mp_ins)
    • 35 small files (5,192 lines of code)
    • 101 very small files (3,607 lines of code)
0% | 3% | 34% | 36% | 25%
Legend:
1001+
501-1000
201-500
101-200
1-100


explore: zoomable circles | sunburst | 3D view
File Size per Extension
1001+
501-1000
201-500
101-200
1-100
py0% | 4% | 34% | 37% | 24%
cpp0% | 0% | 0% | 0% | 100%
lua0% | 0% | 0% | 0% | 100%
File Size per Logical Decomposition
primary
1001+
501-1000
201-500
101-200
1-100
fairseq/models0% | 32% | 55% | 0% | 11%
fairseq0% | 0% | 64% | 21% | 13%
fairseq/data0% | 0% | 37% | 45% | 17%
fairseq/modules0% | 0% | 20% | 51% | 28%
fairseq/optim0% | 0% | 18% | 34% | 46%
ROOT0% | 0% | 18% | 68% | 13%
fairseq_cli0% | 0% | 24% | 69% | 5%
fairseq/tasks0% | 0% | 0% | 92% | 7%
scripts0% | 0% | 0% | 25% | 75%
fairseq/criterions0% | 0% | 0% | 0% | 100%
fairseq/strategies0% | 0% | 0% | 0% | 100%
fairseq/clib0% | 0% | 0% | 0% | 100%
fairseq/fb_tbmf_wrapper0% | 0% | 0% | 0% | 100%
Longest Files (Top 50)
File# lines# units
transformer.py
in fairseq/models
 566 36
bert_seq2seq.py
in fairseq/models
 496 29
sequence_generator.py
in fairseq
 442 14
options.py
in fairseq
 423 19
utils.py
in fairseq
 409 28
indexed_dataset.py
in fairseq/data
 392 62
trainer.py
in fairseq
 373 23
checkpoint_utils.py
in fairseq
 265 11
multihead_attention.py
in fairseq/modules
 256 14
disco_transformer.py
in fairseq/models
 244 9
dictionary.py
in fairseq/data
 241 28
fp16_optimizer.py
in fairseq/optim
 233 32
fb_elmo_token_embedder.py
in fairseq/modules
 231 10
train.py
in fairseq_cli
 222 7
train.py
in root
 222 7
fairseq_model.py
in fairseq/models
 215 39
file_utils.py
in fairseq
 210 12
preprocess.py
in root
 200 5
preprocess.py
in fairseq_cli
 200 5
progress_bar.py
in fairseq
 197 36
downsampled_multihead_attention.py
in fairseq/modules
 197 8
iterators.py
in fairseq/data
 193 34
lightweight_convolution.py
in fairseq/modules
 192 13
dynamic_convolution.py
in fairseq/modules
 182 11
generate_disco.py
in root
 179 5
search.py
in fairseq
 172 13
language_pair_self_dataset_mask.py
in fairseq/data
 171 13
adam.py
in fairseq/optim
 164 9
eval_lm.py
in fairseq_cli
 163 5
eval_lm.py
in root
 163 5
data_utils.py
in fairseq/data
 155 9
masked_multihead_attention.py
in fairseq/modules
 148 1
generate.py
in root
 144 3
transformer_sentence_encoder.py
in fairseq/modules
 144 4
generate.py
in fairseq_cli
 144 3
translation_self.py
in fairseq/tasks
 142 7
interactive.py
in fairseq_cli
 139 4
interactive.py
in root
 139 4
translation.py
in fairseq/tasks
 134 10
bmuf.py
in fairseq/optim
 133 20
distributed_utils.py
in fairseq
 133 9
adaptive_softmax.py
in fairseq/modules
 133 10
adafactor.py
in fairseq/optim
 131 10
character_token_embedder.py
in fairseq/modules
 121 7
gpt2_bpe.py
in fairseq/data/encoders
 120 11
language_pair_dataset.py
in fairseq/data
 120 11
fairseq_task.py
in fairseq/tasks
 118 20
rm_pt.py
in scripts
 109 4
hub_utils.py
in fairseq
 105 3
fb_bidirectional_multihead_attention.py
in fairseq/modules
 104 8
Files With Most Units (Top 20)
File# lines# units
indexed_dataset.py
in fairseq/data
 392 62
fairseq_model.py
in fairseq/models
 215 39
transformer.py
in fairseq/models
 566 36
progress_bar.py
in fairseq
 197 36
iterators.py
in fairseq/data
 193 34
fp16_optimizer.py
in fairseq/optim
 233 32
bert_seq2seq.py
in fairseq/models
 496 29
utils.py
in fairseq
 409 28
dictionary.py
in fairseq/data
 241 28
trainer.py
in fairseq
 373 23
fairseq_task.py
in fairseq/tasks
 118 20
bmuf.py
in fairseq/optim
 133 20
options.py
in fairseq
 423 19
fairseq_optimizer.py
in fairseq/optim
 57 15
sequence_generator.py
in fairseq
 442 14
multihead_attention.py
in fairseq/modules
 256 14
meters.py
in fairseq
 47 13
search.py
in fairseq
 172 13
lightweight_convolution.py
in fairseq/modules
 192 13
language_pair_self_dataset_mask.py
in fairseq/data
 171 13
Files With Long Lines (Top 15)

There are 15 files with lines longer than 120 characters. In total, there are 28 long lines.

File# lines# units# long lines
generate_disco.py
in root
 179 5 5
average_checkpoints.py
in scripts
 102 3 4
transformer.py
in fairseq/models
 566 36 3
mask_predict.py
in fairseq/strategies
 52 4 2
strategy_utils.py
in fairseq/strategies
 27 7 2
language_pair_self_dataset_mask.py
in fairseq/data
 171 13 2
label_smoothed_cross_entropy.py
in fairseq/criterions
 62 6 2
options.py
in fairseq
 423 19 1
easy_first.py
in fairseq/strategies
 84 4 1
adam.py
in fairseq/optim
 164 9 1
disco_transformer.py
in fairseq/models
 244 9 1
bert_seq2seq.py
in fairseq/models
 496 29 1
dynamic_convolution.py
in fairseq/modules
 182 11 1
simple_sinusoidal_positional_embedding.py
in fairseq/modules
 18 2 1
average_best_checkpoints.py
in scripts
 19 2 1