pytorch / FBGEMM
File Change Frequency

File change frequency (churn) shows the distribution of file updates (days with at least one commit).

File Change Frequency Overall
File Change Frequency Overall
The number of recorded file updates
  • There are 233 files with 70,700 lines of code.
    • 0 files changed more than 100 times (0 lines of code)
    • 4 files changed 51-100 times (6,438 lines of code)
    • 46 files changed 21-50 times (24,829 lines of code)
    • 126 files changed 6-20 times (34,194 lines of code)
    • 57 files changed 1-5 times (5,239 lines of code)
0% | 9% | 35% | 48% | 7%
Legend:
101+
51-100
21-50
6-20
1-5
explore: zoomable circles | data
File Change Frequency per File Extension
cc, h, py, cpp, cu, cuh, txt, yml, md, cmake, in, gitignore, toml, rst, bzl, json, gitmodules
File Change Frequency per Extension
The number of recorded file updates
101+
51-100
21-50
6-20
1-5
py0% | 58% | 10% | 18% | 12%
cc0% | 4% | 30% | 56% | 8%
h0% | 10% | 22% | 57% | 9%
cu0% | 0% | 58% | 38% | 2%
cpp0% | 0% | 47% | 50% | 1%
cuh0% | 0% | 78% | 10% | 11%
bzl0% | 0% | 100% | 0% | 0%
toml0% | 0% | 0% | 0% | 100%
cmake0% | 0% | 0% | 0% | 100%
File Change Frequency per Logical Decomposition
primary
primary (file change frequency)
The number of recorded file updates
101+
51-100
21-50
6-20
1-5
fbgemm_gpu/fbgemm_gpu0% | 83% | 0% | 14% | 1%
fbgemm_gpu/bench0% | 50% | 0% | 30% | 19%
src0% | 5% | 35% | 51% | 7%
include/fbgemm0% | 23% | 23% | 44% | 8%
fbgemm_gpu/src0% | 0% | 50% | 45% | 4%
fbgemm_gpu/codegen0% | 0% | 63% | 36% | 0%
fbgemm_gpu/include/fbgemm_gpu0% | 0% | 66% | 19% | 13%
fbgemm_gpu0% | 0% | 88% | 0% | 11%
ROOT0% | 0% | 100% | 0% | 0%
bench0% | 0% | 0% | 91% | 8%
fbgemm_gpu/bench/scripts0% | 0% | 0% | 0% | 100%
cmake/modules0% | 0% | 0% | 0% | 100%
Most Frequently Changed Files (Top 50)

See data for all files...

File# lines# unitslast modified
(days ago)		created
(days ago)		# changes# contributors
1963 48 15 410 78 28
Fbgemm.h
in include/fbgemm
 804 46 66 1259 59 10
1755 31 94 1260 55 12
1916 21 18 471 54 14
Fbgemm.cc
in src
 510 3 60 1258 48 7
sparse_ops_cpu.cpp
in fbgemm_gpu/src
 1683 33 17 340 45 12
sparse_ops.cu
in fbgemm_gpu/src
 1953 - 15 353 44 14
Utils.h
in include/fbgemm
 214 3 60 1255 41 11
Utils.cc
in src
 397 19 60 1255 40 10
1761 15 84 1228 39 10
204 2 94 1188 38 8
259 - 94 1258 38 6
sparse_ops.h
in fbgemm_gpu/include/fbgemm_gpu
 317 - 17 353 38 10
FbgemmEmbedding.h
in include/fbgemm
 234 - 42 859 37 7
486 - 60 1258 34 9
sparse_ops_gpu.cpp
in fbgemm_gpu/src
 157 6 17 340 33 10
1251 1 70 859 33 8
setup.py
in fbgemm_gpu
 144 6 67 471 32 10
308 2 86 1259 32 12
embedding_forward_split_cpu.cpp
in fbgemm_gpu/codegen
 526 5 38 471 31 9
645 4 94 1260 31 7
917 - 15 335 31 9
264 7 94 1259 30 9
706 24 60 1241 30 11
119 1 94 1259 29 8
QuantUtils.h
in include/fbgemm
 224 10 60 1241 29 9
256 1 86 1228 28 10
359 6 15 335 28 12
262 1 86 1259 27 10
340 2 38 471 27 8
FbgemmFP16.h
in include/fbgemm
 40 1 94 1259 25 9
99 1 94 1255 25 10
380 19 15 335 25 8
434 6 39 1046 25 7
1170 1 70 806 24 5
189 6 94 1188 23 6
217 4 94 1241 23 6
256 1 94 1188 23 6
531 34 21 471 23 11
1803 - 17 471 23 10
fbgemm_cuda_utils.cuh
in fbgemm_gpu/include/fbgemm_gpu
 1948 - 31 471 23 7
180 1 15 335 22 7
271 1 86 984 22 9
1085 - 38 471 22 10
QuantUtilsAvx2.h
in include/fbgemm
 111 - 84 1228 21 8
defs.bzl
in root
 118 - 66 736 21 8
153 3 94 1259 21 5
158 1 88 1228 21 5
201 4 94 1241 21 5
989 5 60 935 21 4
Files With Most Contributors (Top 50)
Files sorted by the number of unique email addresses found in commits.

See data for all files...

File# lines# unitslast modified
(days ago)		created
(days ago)		# changes# contributors
1963 48 15 410 78 28
Fbgemm.h
in include/fbgemm
 804 46 66 1259 59 10
1755 31 94 1260 55 12
1916 21 18 471 54 14
Fbgemm.cc
in src
 510 3 60 1258 48 7
sparse_ops_cpu.cpp
in fbgemm_gpu/src
 1683 33 17 340 45 12
sparse_ops.cu
in fbgemm_gpu/src
 1953 - 15 353 44 14
Utils.h
in include/fbgemm
 214 3 60 1255 41 11
Utils.cc
in src
 397 19 60 1255 40 10
1761 15 84 1228 39 10
259 - 94 1258 38 6
204 2 94 1188 38 8
sparse_ops.h
in fbgemm_gpu/include/fbgemm_gpu
 317 - 17 353 38 10
FbgemmEmbedding.h
in include/fbgemm
 234 - 42 859 37 7
486 - 60 1258 34 9
1251 1 70 859 33 8
sparse_ops_gpu.cpp
in fbgemm_gpu/src
 157 6 17 340 33 10
setup.py
in fbgemm_gpu
 144 6 67 471 32 10
308 2 86 1259 32 12
645 4 94 1260 31 7
917 - 15 335 31 9
embedding_forward_split_cpu.cpp
in fbgemm_gpu/codegen
 526 5 38 471 31 9
264 7 94 1259 30 9
706 24 60 1241 30 11
119 1 94 1259 29 8
QuantUtils.h
in include/fbgemm
 224 10 60 1241 29 9
256 1 86 1228 28 10
359 6 15 335 28 12
340 2 38 471 27 8
262 1 86 1259 27 10
434 6 39 1046 25 7
380 19 15 335 25 8
FbgemmFP16.h
in include/fbgemm
 40 1 94 1259 25 9
99 1 94 1255 25 10
1170 1 70 806 24 5
217 4 94 1241 23 6
256 1 94 1188 23 6
189 6 94 1188 23 6
fbgemm_cuda_utils.cuh
in fbgemm_gpu/include/fbgemm_gpu
 1948 - 31 471 23 7
1803 - 17 471 23 10
531 34 21 471 23 11
180 1 15 335 22 7
271 1 86 984 22 9
1085 - 38 471 22 10
989 5 60 935 21 4
153 3 94 1259 21 5
201 4 94 1241 21 5
158 1 88 1228 21 5
defs.bzl
in root
 118 - 66 736 21 8
QuantUtilsAvx2.h
in include/fbgemm
 111 - 84 1228 21 8
Correlations

File Size vs. Number of Changes: 233 points

fbgemm_gpu/codegen/embedding_forward_quantized_cpu_template.cpp x: 359 lines of code y: 28 # changes fbgemm_gpu/codegen/embedding_forward_quantized_host.cpp x: 180 lines of code y: 22 # changes fbgemm_gpu/codegen/embedding_forward_quantized_host_cpu.cpp x: 380 lines of code y: 25 # changes fbgemm_gpu/codegen/embedding_forward_quantized_split_template.cu x: 917 lines of code y: 31 # changes fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops.py x: 1963 lines of code y: 78 # changes fbgemm_gpu/src/sparse_ops.cu x: 1953 lines of code y: 44 # changes fbgemm_gpu/fbgemm_gpu/permute_pooled_embedding_modules.py x: 51 lines of code y: 8 # changes fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embedding_ops.h x: 16 lines of code y: 8 # changes fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embs_function.h x: 1 lines of code y: 3 # changes fbgemm_gpu/src/permute_pooled_embedding_ops_gpu.cpp x: 132 lines of code y: 11 # changes fbgemm_gpu/src/jagged_tensor_ops.cu x: 1150 lines of code y: 12 # changes fbgemm_gpu/src/jagged_tensor_ops_cpu.cpp x: 967 lines of code y: 6 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops.h x: 317 lines of code y: 38 # changes fbgemm_gpu/src/sparse_ops_cpu.cpp x: 1683 lines of code y: 45 # changes fbgemm_gpu/src/sparse_ops_gpu.cpp x: 157 lines of code y: 33 # changes fbgemm_gpu/src/split_embeddings_cache_cuda.cu x: 1803 lines of code y: 23 # changes fbgemm_gpu/include/fbgemm_gpu/embedding_common.h x: 54 lines of code y: 5 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops_utils.h x: 214 lines of code y: 10 # changes fbgemm_gpu/src/split_table_batched_embeddings.cpp x: 101 lines of code y: 15 # changes fbgemm_gpu/codegen/embedding_backward_dense_host.cpp x: 363 lines of code y: 14 # changes fbgemm_gpu/codegen/embedding_backward_dense_host_cpu.cpp x: 164 lines of code y: 16 # changes fbgemm_gpu/codegen/embedding_backward_split_host_cpu_template.cpp x: 200 lines of code y: 17 # changes fbgemm_gpu/codegen/embedding_backward_split_host_template.cpp x: 467 lines of code y: 20 # changes fbgemm_gpu/codegen/embedding_bounds_check_host.cpp x: 16 lines of code y: 10 # changes fbgemm_gpu/src/cumem_utils_host.cpp x: 34 lines of code y: 12 # changes fbgemm_gpu/bench/split_table_batched_embeddings_benchmark.py x: 1916 lines of code y: 54 # changes fbgemm_gpu/pyproject.toml x: 18 lines of code y: 2 # changes fbgemm_gpu/bench/sparse_ops_benchmark.py x: 52 lines of code y: 1 # changes fbgemm_gpu/codegen/embedding_backward_code_generator.py x: 531 lines of code y: 23 # changes fbgemm_gpu/bench/batched_unary_embeddings_benchmark.py x: 137 lines of code y: 2 # changes fbgemm_gpu/bench/quantize_ops_benchmark.py x: 186 lines of code y: 9 # changes fbgemm_gpu/include/fbgemm_gpu/input_combine.h x: 16 lines of code y: 6 # changes fbgemm_gpu/src/input_combine_cpu.cpp x: 278 lines of code y: 11 # changes fbgemm_gpu/include/fbgemm_gpu/quantize_ops_gpu.h x: 14 lines of code y: 1 # changes fbgemm_gpu/src/quantize_ops.cu x: 666 lines of code y: 9 # changes fbgemm_gpu/src/quantize_ops_cpu.cpp x: 303 lines of code y: 12 # changes fbgemm_gpu/src/quantize_ops_gpu.cpp x: 35 lines of code y: 11 # changes bench/GEMMsBenchmark.cc x: 260 lines of code y: 11 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops.cuh x: 35 lines of code y: 19 # changes fbgemm_gpu/bench/jagged_tensor_benchmark.py x: 58 lines of code y: 1 # changes fbgemm_gpu/include/fbgemm_gpu/fbgemm_cuda_utils.cuh x: 1948 lines of code y: 23 # changes fbgemm_gpu/bench/bench_utils.py x: 34 lines of code y: 1 # changes fbgemm_gpu/bench/merge_embeddings_benchmark.py x: 440 lines of code y: 17 # changes fbgemm_gpu/codegen/embedding_backward_split_cpu_approx_template.cpp x: 191 lines of code y: 19 # changes fbgemm_gpu/codegen/embedding_backward_split_cpu_template.cpp x: 340 lines of code y: 27 # changes fbgemm_gpu/codegen/embedding_backward_split_indice_weights_template.cu x: 274 lines of code y: 12 # changes fbgemm_gpu/codegen/embedding_backward_split_template.cu x: 1085 lines of code y: 22 # changes fbgemm_gpu/codegen/embedding_bounds_check.cu x: 138 lines of code y: 14 # changes fbgemm_gpu/codegen/embedding_forward_split_cpu.cpp x: 526 lines of code y: 31 # changes fbgemm_gpu/codegen/embedding_forward_split_template.cu x: 450 lines of code y: 18 # changes fbgemm_gpu/src/histogram_binning_calibration_ops.cu x: 388 lines of code y: 7 # changes fbgemm_gpu/src/layout_transform_ops.cu x: 129 lines of code y: 11 # changes fbgemm_gpu/src/layout_transform_ops_cpu.cpp x: 63 lines of code y: 8 # changes fbgemm_gpu/src/permute_pooled_embedding_ops.cu x: 74 lines of code y: 8 # changes fbgemm_gpu/src/split_embeddings_utils.cu x: 284 lines of code y: 5 # changes src/FbgemmConv.cc x: 434 lines of code y: 25 # changes fbgemm_gpu/include/fbgemm_gpu/cpu_utils.h x: 42 lines of code y: 8 # changes include/fbgemm/FbgemmEmbedding.h x: 234 lines of code y: 37 # changes src/GroupwiseConv.cc x: 938 lines of code y: 14 # changes src/TransposeUtils.cc x: 74 lines of code y: 9 # changes fbgemm_gpu/bench/split_embeddings_cache_benchmark.py x: 448 lines of code y: 5 # changes fbgemm_gpu/src/merge_pooled_embeddings_cpu.cpp x: 33 lines of code y: 9 # changes fbgemm_gpu/src/merge_pooled_embeddings_gpu.cpp x: 341 lines of code y: 14 # changes include/fbgemm/FbgemmFPCommon.h x: 190 lines of code y: 8 # changes include/fbgemm/QuantUtils.h x: 224 lines of code y: 29 # changes include/fbgemm/Utils.h x: 214 lines of code y: 41 # changes src/ExecuteKernelU8S8.cc x: 486 lines of code y: 34 # changes src/Fbgemm.cc x: 510 lines of code y: 48 # changes src/FbgemmI8Depthwise2DAvx2-inl.h x: 656 lines of code y: 18 # changes src/FbgemmI8Depthwise3DAvx2.cc x: 989 lines of code y: 21 # changes src/QuantUtils.cc x: 706 lines of code y: 30 # changes src/Utils.cc x: 397 lines of code y: 40 # changes defs.bzl x: 118 lines of code y: 21 # changes include/fbgemm/Fbgemm.h x: 804 lines of code y: 59 # changes include/fbgemm/FbgemmI8DirectconvAvx2.h x: 33 lines of code y: 1 # changes src/DirectConv.h x: 148 lines of code y: 5 # changes src/GenerateKernelDirectConvU8S8S32ACC32.cc x: 491 lines of code y: 6 # changes src/PackWeightsForConv.cc x: 172 lines of code y: 14 # changes src/PackWeightsForDirectConv.cc x: 412 lines of code y: 1 # changes fbgemm_gpu/setup.py x: 144 lines of code y: 32 # changes fbgemm_gpu/version.py x: 1 lines of code y: 1 # changes fbgemm_gpu/src/cumem_utils.h x: 18 lines of code y: 5 # changes src/CodeCache.h x: 96 lines of code y: 11 # changes src/EmbeddingSpMDM.cc x: 1251 lines of code y: 33 # changes src/EmbeddingSpMDMNBit.cc x: 1170 lines of code y: 24 # changes fbgemm_gpu/include/fbgemm_gpu/embedding_backward_template_helpers.cuh x: 98 lines of code y: 6 # changes fbgemm_gpu/include/fbgemm_gpu/split_embeddings_utils.cuh x: 139 lines of code y: 4 # changes fbgemm_gpu/include/fbgemm_gpu/dispatch_macros.h x: 141 lines of code y: 8 # changes include/fbgemm/QuantUtilsAvx2.h x: 111 lines of code y: 21 # changes src/QuantUtilsAvx2.cc x: 1761 lines of code y: 39 # changes bench/BenchUtils.cc x: 182 lines of code y: 19 # changes bench/BenchUtils.h x: 415 lines of code y: 19 # changes bench/EmbeddingIndexRemappingBenchmark.cc x: 134 lines of code y: 4 # changes bench/EmbeddingSpMDM8BitBenchmark.cc x: 282 lines of code y: 19 # changes bench/EmbeddingSpMDMBenchmark.cc x: 293 lines of code y: 16 # changes bench/EmbeddingSpMDMNBitBenchmark.cc x: 278 lines of code y: 9 # changes bench/EmbeddingSpMDMNBitRowWiseSparseBenchmark.cc x: 309 lines of code y: 10 # changes bench/GEMMsTunableBenchmark.cc x: 286 lines of code y: 9 # changes bench/GroupwiseConvRequantizeBenchmark.cc x: 413 lines of code y: 10 # changes bench/Im2ColFusedRequantizeBenchmark.cc x: 296 lines of code y: 7 # changes bench/PackedFloatInOutBenchmark.cc x: 242 lines of code y: 17 # changes bench/PackedRequantizeAcc16Benchmark.cc x: 359 lines of code y: 17 # changes bench/PackedRequantizeAcc32Benchmark.cc x: 262 lines of code y: 16 # changes bench/RowwiseAdagradBenchmark.cc x: 189 lines of code y: 12 # changes bench/RowwiseAdagradFusedBenchmark.cc x: 162 lines of code y: 7 # changes bench/SparseDenseMMFP32Benchmark.cc x: 93 lines of code y: 5 # changes bench/SparseDenseMMInt8Benchmark.cc x: 112 lines of code y: 5 # changes include/fbgemm/Types.h x: 116 lines of code y: 8 # changes src/CodeGenHelpers.h x: 167 lines of code y: 14 # changes src/FbgemmSparseDenseVectorInt8Avx512.cc x: 217 lines of code y: 6 # changes src/GenerateKernelU8S8S32ACC16.cc x: 262 lines of code y: 27 # changes src/GenerateKernelU8S8S32ACC16Avx512.cc x: 256 lines of code y: 28 # changes src/GenerateKernelU8S8S32ACC32.cc x: 308 lines of code y: 32 # changes src/GenerateKernelU8S8S32ACC32Avx512VNNI.cc x: 271 lines of code y: 22 # changes src/FbgemmI8DepthwiseAvx2.cc x: 158 lines of code y: 21 # changes src/FbgemmFloat16Convert.cc x: 74 lines of code y: 14 # changes include/fbgemm/FbgemmI8DepthwiseAvx2.h x: 63 lines of code y: 17 # changes src/FbgemmFP16UKernelsAvx2.cc x: 886 lines of code y: 16 # changes src/FbgemmFP16UKernelsAvx512.cc x: 3130 lines of code y: 13 # changes src/FbgemmFP16UKernelsAvx512_256.cc x: 2247 lines of code y: 9 # changes src/codegen_fp16fp32.cc x: 461 lines of code y: 19 # changes bench/ConvUnifiedBenchmark.cc x: 449 lines of code y: 17 # changes bench/ConvertBenchmark.cc x: 51 lines of code y: 4 # changes bench/Depthwise3DBenchmark.cc x: 188 lines of code y: 18 # changes bench/DepthwiseBenchmark.cc x: 305 lines of code y: 20 # changes bench/EmbeddingQuantizeBenchmark.cc x: 89 lines of code y: 5 # changes bench/FP16Benchmark.cc x: 40 lines of code y: 18 # changes bench/I64Benchmark.cc x: 115 lines of code y: 6 # changes bench/I8SpmdmBenchmark.cc x: 172 lines of code y: 8 # changes bench/RequantizeBenchmark.cc x: 123 lines of code y: 7 # changes bench/RowOffsetBenchmark.cc x: 37 lines of code y: 6 # changes bench/TransposeBenchmark.cc x: 49 lines of code y: 9 # changes fbgemm_gpu/bench/histogram_binning_calibration_benchmark.py x: 242 lines of code y: 6 # changes fbgemm_gpu/bench/scripts/batch_benchmark_run.py x: 73 lines of code y: 4 # changes fbgemm_gpu/bench/verify_fp16_stochastic_benchmark.cu x: 287 lines of code y: 6 # changes fbgemm_gpu/codegen/embedding_forward_split_cpu.h x: 58 lines of code y: 13 # changes fbgemm_gpu/codegen/embedding_forward_template_helpers.cuh x: 29 lines of code y: 13 # changes fbgemm_gpu/codegen/lookup_args.py x: 39 lines of code y: 7 # changes fbgemm_gpu/fbgemm_gpu/__init__.py x: 4 lines of code y: 4 # changes fbgemm_gpu/fbgemm_gpu/batched_unary_embeddings_ops.py x: 61 lines of code y: 6 # changes fbgemm_gpu/fbgemm_gpu/enums.py x: 10 lines of code y: 3 # changes fbgemm_gpu/fbgemm_gpu/split_embedding_configs.py x: 88 lines of code y: 16 # changes fbgemm_gpu/fbgemm_gpu/split_embedding_inference_converter.py x: 145 lines of code y: 16 # changes fbgemm_gpu/fbgemm_gpu/uvm.py x: 19 lines of code y: 4 # changes fbgemm_gpu/include/fbgemm_gpu/bench_utils.cuh x: 65 lines of code y: 6 # changes fbgemm_gpu/include/fbgemm_gpu/cub_namespace_postfix.cuh x: 14 lines of code y: 4 # changes fbgemm_gpu/include/fbgemm_gpu/cub_namespace_prefix.cuh x: 12 lines of code y: 3 # changes fbgemm_gpu/include/fbgemm_gpu/enum_utils.h x: 57 lines of code y: 4 # changes fbgemm_gpu/include/fbgemm_gpu/layout_transform_ops.cuh x: 105 lines of code y: 5 # changes fbgemm_gpu/include/fbgemm_gpu/quantize_ops.cuh x: 27 lines of code y: 18 # changes fbgemm_gpu/src/cpu_utils.cpp x: 145 lines of code y: 4 # changes fbgemm_gpu/src/layout_transform_ops_gpu.cpp x: 16 lines of code y: 7 # changes include/fbgemm/ConvUtils.h x: 161 lines of code y: 12 # changes include/fbgemm/FbgemmBuild.h x: 60 lines of code y: 9 # changes include/fbgemm/FbgemmFP16.h x: 40 lines of code y: 25 # changes include/fbgemm/FbgemmI8Spmdm.h x: 77 lines of code y: 7 # changes include/fbgemm/FbgemmPackMatrixB.h x: 200 lines of code y: 8 # changes include/fbgemm/FbgemmSparse.h x: 135 lines of code y: 3 # changes include/fbgemm/OutputProcessing-inl.h x: 282 lines of code y: 14 # changes include/fbgemm/PackingTraits-inl.h x: 298 lines of code y: 14 # changes include/fbgemm/UtilsAvx2.h x: 51 lines of code y: 13 # changes include/fbgemm/spmmUtils.h x: 44 lines of code y: 3 # changes include/fbgemm/spmmUtilsAvx2.h x: 28 lines of code y: 3 # changes src/EmbeddingSpMDMAvx2.cc x: 132 lines of code y: 8 # changes src/EmbeddingSpMDMAvx512.cc x: 521 lines of code y: 6 # changes src/ExecuteKernel.cc x: 1 lines of code y: 6 # changes src/ExecuteKernelU8S8.h x: 50 lines of code y: 13 # changes src/FbgemmBfloat16Convert.cc x: 56 lines of code y: 12 # changes src/FbgemmBfloat16ConvertAvx2.cc x: 42 lines of code y: 5 # changes src/FbgemmBfloat16ConvertAvx512.cc x: 38 lines of code y: 5 # changes src/FbgemmFP16.cc x: 119 lines of code y: 29 # changes src/FbgemmFP16UKernelsAvx2.h x: 14 lines of code y: 14 # changes src/FbgemmFP16UKernelsAvx512.h x: 22 lines of code y: 12 # changes src/FbgemmFP16UKernelsIntrinsicAvx2.cc x: 86 lines of code y: 4 # changes src/FbgemmFP16UKernelsIntrinsicAvx512.cc x: 111 lines of code y: 4 # changes src/FbgemmFP16UKernelsIntrinsicAvx512_256.cc x: 92 lines of code y: 4 # changes src/FbgemmFPCommon.cc x: 260 lines of code y: 6 # changes src/FbgemmI64.cc x: 409 lines of code y: 11 # changes src/FbgemmI8DepthwiseAvx2-inl.h x: 487 lines of code y: 9 # changes src/FbgemmI8DepthwisePerChannelQuantAvx2.cc x: 91 lines of code y: 10 # changes src/FbgemmSparseDense.cc x: 253 lines of code y: 6 # changes src/FbgemmSparseDenseAvx2.cc x: 117 lines of code y: 5 # changes src/FbgemmSparseDenseAvx512.cc x: 130 lines of code y: 5 # changes src/FbgemmSparseDenseInt8Avx2.cc x: 221 lines of code y: 5 # changes src/FbgemmSparseDenseInt8Avx512.cc x: 506 lines of code y: 4 # changes src/GenerateI8Depthwise.cc x: 483 lines of code y: 11 # changes src/GenerateI8Depthwise.h x: 30 lines of code y: 6 # changes src/GenerateKernel.h x: 99 lines of code y: 25 # changes src/GenerateKernelU8S8S32ACC16Avx512VNNI.cc x: 27 lines of code y: 9 # changes src/GroupwiseConv.h x: 256 lines of code y: 23 # changes src/GroupwiseConvAcc32Avx2.cc x: 204 lines of code y: 38 # changes src/GroupwiseConvAcc32Avx512.cc x: 206 lines of code y: 7 # changes src/PackAWithIm2Col.cc x: 645 lines of code y: 31 # changes src/PackAWithQuantRowOffset.cc x: 217 lines of code y: 23 # changes src/PackAWithRowOffset.cc x: 201 lines of code y: 21 # changes src/PackBMatrix.cc x: 264 lines of code y: 30 # changes src/PackDepthwiseConvMatrixAvx2.cc x: 113 lines of code y: 13 # changes src/PackMatrix.cc x: 116 lines of code y: 20 # changes src/PackWeightMatrixForGConv.cc x: 189 lines of code y: 23 # changes src/QuantUtilsAvx512.cc x: 356 lines of code y: 7 # changes src/RefImplementations.cc x: 1755 lines of code y: 55 # changes src/RefImplementations.h x: 259 lines of code y: 38 # changes src/RowWiseSparseAdagradFused.cc x: 823 lines of code y: 14 # changes src/SparseAdagrad.cc x: 842 lines of code y: 20 # changes src/TransposeUtils.h x: 31 lines of code y: 10 # changes src/TransposeUtilsAvx2.h x: 343 lines of code y: 7 # changes src/UtilsAvx2.cc x: 269 lines of code y: 10 # changes src/UtilsAvx512.cc x: 845 lines of code y: 12 # changes src/spmmUtils.cc x: 292 lines of code y: 4 # changes src/spmmUtilsAvx2.cc x: 271 lines of code y: 4 # changes cmake/modules/FindSphinx.cmake x: 8 lines of code y: 1 # changes bench/AlignedVec.h x: 86 lines of code y: 3 # changes
78.0
# changes
  min: 1.0
  average: 13.88
  25th percentile: 6.0
  median: 10.0
  75th percentile: 19.0
  max: 78.0
0 3130.0
lines of code
min: 1.0 | average: 303.43 | 25th percentile: 54.5 | median: 162.0 | 75th percentile: 313.0 | max: 3130.0

Number of Contributors vs. Number of Changes: 233 points

fbgemm_gpu/codegen/embedding_forward_quantized_cpu_template.cpp x: 12 # contributors y: 28 # changes fbgemm_gpu/codegen/embedding_forward_quantized_host.cpp x: 7 # contributors y: 22 # changes fbgemm_gpu/codegen/embedding_forward_quantized_host_cpu.cpp x: 8 # contributors y: 25 # changes fbgemm_gpu/codegen/embedding_forward_quantized_split_template.cu x: 9 # contributors y: 31 # changes fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops.py x: 28 # contributors y: 78 # changes fbgemm_gpu/src/sparse_ops.cu x: 14 # contributors y: 44 # changes fbgemm_gpu/fbgemm_gpu/permute_pooled_embedding_modules.py x: 6 # contributors y: 8 # changes fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embedding_ops.h x: 5 # contributors y: 8 # changes fbgemm_gpu/include/fbgemm_gpu/permute_pooled_embs_function.h x: 2 # contributors y: 3 # changes fbgemm_gpu/src/permute_pooled_embedding_ops_gpu.cpp x: 7 # contributors y: 11 # changes fbgemm_gpu/src/jagged_tensor_ops.cu x: 4 # contributors y: 12 # changes fbgemm_gpu/src/jagged_tensor_ops_cpu.cpp x: 2 # contributors y: 6 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops.h x: 10 # contributors y: 38 # changes fbgemm_gpu/src/sparse_ops_cpu.cpp x: 12 # contributors y: 45 # changes fbgemm_gpu/src/sparse_ops_gpu.cpp x: 10 # contributors y: 33 # changes fbgemm_gpu/src/split_embeddings_cache_cuda.cu x: 10 # contributors y: 23 # changes fbgemm_gpu/include/fbgemm_gpu/embedding_common.h x: 4 # contributors y: 5 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops_utils.h x: 5 # contributors y: 10 # changes fbgemm_gpu/src/split_table_batched_embeddings.cpp x: 5 # contributors y: 15 # changes fbgemm_gpu/codegen/embedding_backward_dense_host.cpp x: 4 # contributors y: 14 # changes fbgemm_gpu/codegen/embedding_backward_dense_host_cpu.cpp x: 7 # contributors y: 16 # changes fbgemm_gpu/codegen/embedding_backward_split_host_cpu_template.cpp x: 8 # contributors y: 17 # changes fbgemm_gpu/codegen/embedding_backward_split_host_template.cpp x: 6 # contributors y: 20 # changes fbgemm_gpu/codegen/embedding_bounds_check_host.cpp x: 4 # contributors y: 10 # changes fbgemm_gpu/codegen/embedding_bounds_check_host_cpu.cpp x: 6 # contributors y: 15 # changes fbgemm_gpu/src/cumem_utils_host.cpp x: 5 # contributors y: 12 # changes fbgemm_gpu/bench/split_table_batched_embeddings_benchmark.py x: 14 # contributors y: 54 # changes fbgemm_gpu/pyproject.toml x: 2 # contributors y: 2 # changes fbgemm_gpu/bench/sparse_ops_benchmark.py x: 1 # contributors y: 1 # changes fbgemm_gpu/codegen/embedding_backward_code_generator.py x: 11 # contributors y: 23 # changes fbgemm_gpu/bench/batched_unary_embeddings_benchmark.py x: 1 # contributors y: 2 # changes fbgemm_gpu/bench/quantize_ops_benchmark.py x: 6 # contributors y: 9 # changes fbgemm_gpu/include/fbgemm_gpu/input_combine.h x: 4 # contributors y: 6 # changes fbgemm_gpu/src/input_combine_cpu.cpp x: 8 # contributors y: 11 # changes fbgemm_gpu/src/quantize_ops_cpu.cpp x: 7 # contributors y: 12 # changes fbgemm_gpu/src/quantize_ops_gpu.cpp x: 5 # contributors y: 11 # changes bench/GEMMsBenchmark.cc x: 6 # contributors y: 11 # changes fbgemm_gpu/include/fbgemm_gpu/sparse_ops.cuh x: 8 # contributors y: 19 # changes fbgemm_gpu/include/fbgemm_gpu/fbgemm_cuda_utils.cuh x: 7 # contributors y: 23 # changes fbgemm_gpu/codegen/embedding_backward_split_cpu_approx_template.cpp x: 6 # contributors y: 19 # changes fbgemm_gpu/codegen/embedding_backward_split_cpu_template.cpp x: 8 # contributors y: 27 # changes fbgemm_gpu/codegen/embedding_backward_split_template.cu x: 10 # contributors y: 22 # changes fbgemm_gpu/codegen/embedding_bounds_check.cu x: 6 # contributors y: 14 # changes fbgemm_gpu/codegen/embedding_forward_split_template.cu x: 6 # contributors y: 18 # changes fbgemm_gpu/src/histogram_binning_calibration_ops.cu x: 3 # contributors y: 7 # changes fbgemm_gpu/src/permute_pooled_embedding_ops.cu x: 4 # contributors y: 8 # changes fbgemm_gpu/src/split_embeddings_utils.cu x: 3 # contributors y: 5 # changes src/FbgemmConv.cc x: 7 # contributors y: 25 # changes include/fbgemm/FbgemmEmbedding.h x: 7 # contributors y: 37 # changes src/GroupwiseConv.cc x: 5 # contributors y: 14 # changes src/TransposeUtils.cc x: 4 # contributors y: 9 # changes fbgemm_gpu/bench/split_embeddings_cache_benchmark.py x: 1 # contributors y: 5 # changes fbgemm_gpu/src/merge_pooled_embeddings_cpu.cpp x: 5 # contributors y: 9 # changes fbgemm_gpu/src/merge_pooled_embeddings_gpu.cpp x: 7 # contributors y: 14 # changes include/fbgemm/QuantUtils.h x: 9 # contributors y: 29 # changes include/fbgemm/Utils.h x: 11 # contributors y: 41 # changes src/ExecuteKernelU8S8.cc x: 9 # contributors y: 34 # changes src/Fbgemm.cc x: 7 # contributors y: 48 # changes src/FbgemmI8Depthwise2DAvx2-inl.h x: 4 # contributors y: 18 # changes src/FbgemmI8Depthwise3DAvx2.cc x: 4 # contributors y: 21 # changes src/QuantUtils.cc x: 11 # contributors y: 30 # changes src/Utils.cc x: 10 # contributors y: 40 # changes defs.bzl x: 8 # contributors y: 21 # changes include/fbgemm/Fbgemm.h x: 10 # contributors y: 59 # changes src/DirectConv.h x: 2 # contributors y: 5 # changes src/GenerateKernelDirectConvU8S8S32ACC32.cc x: 3 # contributors y: 6 # changes fbgemm_gpu/setup.py x: 10 # contributors y: 32 # changes src/EmbeddingSpMDM.cc x: 8 # contributors y: 33 # changes src/EmbeddingSpMDMNBit.cc x: 5 # contributors y: 24 # changes fbgemm_gpu/include/fbgemm_gpu/split_embeddings_utils.cuh x: 3 # contributors y: 4 # changes fbgemm_gpu/include/fbgemm_gpu/dispatch_macros.h x: 3 # contributors y: 8 # changes src/QuantUtilsAvx2.cc x: 10 # contributors y: 39 # changes bench/BenchUtils.cc x: 5 # contributors y: 19 # changes bench/BenchUtils.h x: 4 # contributors y: 19 # changes bench/EmbeddingSpMDMBenchmark.cc x: 4 # contributors y: 16 # changes bench/EmbeddingSpMDMNBitBenchmark.cc x: 2 # contributors y: 9 # changes bench/EmbeddingSpMDMNBitRowWiseSparseBenchmark.cc x: 2 # contributors y: 10 # changes bench/GEMMsTunableBenchmark.cc x: 3 # contributors y: 9 # changes bench/GroupwiseConvRequantizeBenchmark.cc x: 3 # contributors y: 10 # changes bench/PackedFloatInOutBenchmark.cc x: 4 # contributors y: 17 # changes bench/RowwiseAdagradFusedBenchmark.cc x: 2 # contributors y: 7 # changes bench/SparseAdagradBenchmark.cc x: 6 # contributors y: 12 # changes src/GenerateKernelU8S8S32ACC16.cc x: 10 # contributors y: 27 # changes src/GenerateKernelU8S8S32ACC16Avx512.cc x: 10 # contributors y: 28 # changes src/GenerateKernelU8S8S32ACC32.cc x: 12 # contributors y: 32 # changes src/GenerateKernelU8S8S32ACC32Avx512VNNI.cc x: 9 # contributors y: 22 # changes src/FbgemmI8DepthwiseAvx2.cc x: 5 # contributors y: 21 # changes src/FbgemmFloat16Convert.cc x: 8 # contributors y: 14 # changes include/fbgemm/FbgemmI8DepthwiseAvx2.h x: 3 # contributors y: 17 # changes src/FbgemmFP16UKernelsAvx512.cc x: 6 # contributors y: 13 # changes bench/ConvUnifiedBenchmark.cc x: 7 # contributors y: 17 # changes bench/ConvertBenchmark.cc x: 1 # contributors y: 4 # changes bench/DepthwiseBenchmark.cc x: 4 # contributors y: 20 # changes bench/FP16Benchmark.cc x: 8 # contributors y: 18 # changes bench/I8SpmdmBenchmark.cc x: 2 # contributors y: 8 # changes fbgemm_gpu/codegen/embedding_forward_split_cpu.h x: 7 # contributors y: 13 # changes fbgemm_gpu/fbgemm_gpu/__init__.py x: 2 # contributors y: 4 # changes fbgemm_gpu/fbgemm_gpu/split_embedding_configs.py x: 8 # contributors y: 16 # changes fbgemm_gpu/fbgemm_gpu/split_embedding_inference_converter.py x: 9 # contributors y: 16 # changes fbgemm_gpu/include/fbgemm_gpu/bench_utils.cuh x: 1 # contributors y: 6 # changes fbgemm_gpu/include/fbgemm_gpu/cuda_utils.cuh x: 1 # contributors y: 3 # changes fbgemm_gpu/include/fbgemm_gpu/quantize_ops.cuh x: 7 # contributors y: 18 # changes include/fbgemm/FbgemmFP16.h x: 9 # contributors y: 25 # changes include/fbgemm/OutputProcessing-inl.h x: 3 # contributors y: 14 # changes include/fbgemm/UtilsAvx2.h x: 4 # contributors y: 13 # changes src/ExecuteKernelU8S8.h x: 5 # contributors y: 13 # changes src/FbgemmFP16.cc x: 8 # contributors y: 29 # changes src/GenerateI8Depthwise.cc x: 4 # contributors y: 11 # changes src/GenerateKernel.h x: 10 # contributors y: 25 # changes src/GroupwiseConv.h x: 6 # contributors y: 23 # changes src/GroupwiseConvAcc32Avx2.cc x: 8 # contributors y: 38 # changes src/PackAWithIm2Col.cc x: 7 # contributors y: 31 # changes src/PackBMatrix.cc x: 9 # contributors y: 30 # changes src/PackMatrix.cc x: 9 # contributors y: 20 # changes src/RefImplementations.cc x: 12 # contributors y: 55 # changes src/RefImplementations.h x: 6 # contributors y: 38 # changes src/SparseAdagrad.cc x: 8 # contributors y: 20 # changes src/TransposeUtilsAvx2.h x: 4 # contributors y: 7 # changes
78.0
# changes
  min: 1.0
  average: 13.88
  25th percentile: 6.0
  median: 10.0
  75th percentile: 19.0
  max: 78.0
0 28.0
# contributors
min: 1.0 | average: 4.97 | 25th percentile: 3.0 | median: 4.0 | 75th percentile: 7.0 | max: 28.0