G

Intro

For duplication, we look at places in code where there are 6 or more lines of code that are exactly the same.
Before duplication is calculated, the code is cleaned to remove empty lines, comments, and frequently duplicated constructs such as imports.
You should aim at having as little as possible (<5%) of duplicated code as high-level of duplication can lead to maintenance difficulties, poor factoring, and logical contradictions.

Learn more...

Duplication Overall

40% duplication:

64,386 cleaned lines of cleaned code (without empty lines, comments, and frequently duplicated constructs such as imports)
26,034 duplicated lines

5,193 duplicates

Duplication per Extension

Duplication per Component (primary)

Duplication Between Components (50+ lines)

Download: SVG DOT (open online Graphviz editor)

From Component --> To Component	Duplicated Lines	File Pairs	Details
include/fbgemm (12%) --> src (1%)	696	37 file pairs	details...
fbgemm_gpu/include/fbgemm_gpu (2%) --> fbgemm_gpu/src (<1%)	190	92 file pairs	details...
fbgemm_gpu/codegen (1%) --> fbgemm_gpu/src (<1%)	145	54 file pairs	details...
fbgemm_gpu/codegen (<1%) --> fbgemm_gpu/include/fbgemm_gpu (1%)	106	60 file pairs	details...
fbgemm_gpu/bench (<1%) --> fbgemm_gpu/include/fbgemm_gpu (1%)	66	10 file pairs	details...
fbgemm_gpu/bench (<1%) --> fbgemm_gpu/src (<1%)	60	9 file pairs	details...

Open 3D force graph...

Show more details on duplication between components...

Longest Duplicates

The list of 50 longest duplicates.

See data for all 5,193 duplicates...

Size	#	Folders	Files	Lines	Code
179	x 2	src src	FbgemmI8Depthwise3DAvx2.cc FbgemmI8Depthwise3DAvx2.cc	194:384 (19%) 584:774 (19%)	view
93	x 2	bench bench	EmbeddingSpMDMNBitBenchmark.cc EmbeddingSpMDMNBitRowWiseSparseBenchm...	215:363 (39%) 253:399 (35%)	view
84	x 2	src src	FbgemmI8Depthwise2DAvx2-inl.h FbgemmI8Depthwise2DAvx2-inl.h	181:268 (13%) 396:483 (13%)	view
71	x 2	src src	UtilsAvx2.cc UtilsAvx512.cc	65:156 (31%) 372:463 (9%)	view
66	x 2	src src	FbgemmI8Depthwise3DAvx2.cc FbgemmI8Depthwise3DAvx2.cc	390:458 (7%) 584:652 (7%)	view
66	x 2	src src	FbgemmI8Depthwise3DAvx2.cc FbgemmI8Depthwise3DAvx2.cc	194:262 (7%) 390:458 (7%)	view
63	x 2	src src	FbgemmI8Depthwise3DAvx2.cc FbgemmI8Depthwise3DAvx2.cc	319:384 (6%) 513:578 (6%)	view
63	x 2	src src	EmbeddingSpMDM.cc EmbeddingSpMDMNBit.cc	821:907 (5%) 859:945 (6%)	view
63	x 2	src src	FbgemmI8Depthwise3DAvx2.cc FbgemmI8Depthwise3DAvx2.cc	513:578 (6%) 709:774 (6%)	view
50	x 2	src src	PackAWithQuantRowOffset.cc PackAWithRowOffset.cc	55:117 (27%) 43:105 (29%)	view
49	x 2	src src	EmbeddingSpMDM.cc EmbeddingSpMDMNBit.cc	585:656 (4%) 624:696 (4%)	view
49	x 2	src src	FbgemmFP16UKernelsIntrinsicAvx2.cc FbgemmFP16UKernelsIntrinsicAvx512_256.cc	20:91 (74%) 20:91 (70%)	view
49	x 2	fbgemm_gpu/src fbgemm_gpu/src	split_embeddings_cache_cuda.cu split_embeddings_cache_cuda.cu	640:688 (2%) 1375:1425 (2%)	view
48	x 2	bench bench	PackedFloatInOutBenchmark.cc PackedRequantizeAcc32Benchmark.cc	35:97 (24%) 39:101 (22%)	view
47	x 2	src src	OptimizedKernelsAvx2.cc OptimizedKernelsAvx2.cc	226:277 (16%) 298:349 (16%)	view
46	x 2	src src	EmbeddingSpMDM.cc EmbeddingSpMDMNBit.cc	512:580 (4%) 551:619 (4%)	view
44	x 2	fbgemm_gpu/include/fbgemm_gpu fbgemm_gpu/include/fbgemm_gpu	fbgemm_cuda_utils.cuh fbgemm_cuda_utils.cuh	234:286 (2%) 373:425 (2%)	view
43	x 2	src src	EmbeddingSpMDM.cc EmbeddingSpMDMNBit.cc	343:393 (3%) 323:372 (4%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	3191:3232 (1%) 3311:3352 (1%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	3191:3232 (1%) 3362:3403 (1%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	3311:3352 (1%) 3362:3403 (1%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	2164:2205 (1%) 2335:2376 (1%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	2164:2205 (1%) 2284:2325 (1%)	view
42	x 2	bench bench	EmbeddingSpMDM8BitBenchmark.cc EmbeddingSpMDMNBitBenchmark.cc	194:255 (18%) 193:254 (17%)	view
42	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	2284:2325 (1%) 2335:2376 (1%)	view
41	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1762:1804 (1%) 2114:2156 (1%)	view
41	x 2	fbgemm_gpu/src fbgemm_gpu/src	split_embeddings_cache_cuda.cu split_embeddings_cache_cuda.cu	1305:1350 (2%) 1642:1687 (2%)	view
41	x 2	src src	GenerateKernelU8S8S32ACC16Avx512.cc GenerateKernelU8S8S32ACC32Avx512VNNI.cc	134:184 (17%) 125:175 (16%)	view
41	x 2	src src	GenerateKernelU8S8S32ACC32.cc GenerateKernelU8S8S32ACC32Avx512VNNI.cc	115:164 (14%) 67:117 (16%)	view
41	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2789:2831 (1%) 3141:3183 (1%)	view
40	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	2010:2050 (1%) 2378:2418 (1%)	view
40	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	3037:3077 (1%) 3405:3445 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1809:1847 (1%) 1922:1960 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1809:1847 (1%) 1970:2008 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1922:1960 (1%) 1970:2008 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2836:2874 (1%) 2997:3035 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2836:2874 (1%) 2949:2987 (1%)	view
39	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2949:2987 (1%) 2997:3035 (1%)	view
38	x 2	src src	QuantUtilsAvx2.cc QuantUtilsAvx2.cc	457:506 (2%) 1143:1191 (2%)	view
38	x 2	fbgemm_gpu/include/fbgemm_gpu fbgemm_gpu/include/fbgemm_gpu	fbgemm_cuda_utils.cuh fbgemm_cuda_utils.cuh	198:240 (1%) 477:519 (1%)	view
38	x 2	src src	PackAMatrix.cc PackAWithRowOffset.cc	35:80 (29%) 40:85 (22%)	view
38	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2456:2495 (1%) 2789:2828 (1%)	view
38	x 2	src src	FbgemmFP16UKernelsAvx512.cc FbgemmFP16UKernelsAvx512.cc	2456:2495 (1%) 3141:3180 (1%)	view
38	x 2	src src	PackAWithIm2Col.cc PackAWithRowOffset.cc	48:93 (6%) 40:85 (22%)	view
38	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1429:1468 (1%) 1762:1801 (1%)	view
38	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1429:1468 (1%) 2114:2153 (1%)	view
38	x 2	src src	PackAMatrix.cc PackAWithIm2Col.cc	35:80 (29%) 48:93 (6%)	view
37	x 2	fbgemm_gpu/src fbgemm_gpu/src	histogram_binning_calibration_ops.cu histogram_binning_calibration_ops.cu	185:225 (9%) 346:386 (9%)	view
37	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512_256.cc	1661:1698 (1%) 2378:2415 (1%)	view
37	x 2	fbgemm_gpu/include/fbgemm_gpu fbgemm_gpu/include/fbgemm_gpu	fbgemm_cuda_utils.cuh fbgemm_cuda_utils.cuh	313:355 (1%) 432:474 (1%)	view

Duplicated Units

The list of top 23 duplicated units.

See data for all 23 unit duplicates...

Size	#	Folders	Files	Lines	Code
345	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	2076:2446 3103:3473	view
326	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	1724:2075 2751:3102	view
307	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	1391:1723 2418:2750	view
288	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	1077:1390 2104:2417	view
269	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	782:1076 1809:2103	view
250	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	506:781 1533:1808	view
231	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	249:505 1276:1532	view
212	x 2	src src	FbgemmFP16UKernelsAvx512_256.cc FbgemmFP16UKernelsAvx512.cc	11:248 1038:1275	view
193	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	819:1037 819:1037	view
174	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	619:818 619:818	view
155	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	438:618 438:618	view
136	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	276:437 276:437	view
117	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	133:275 133:275	view
96	x 2	src src	FbgemmFP16UKernelsAvx2.cc FbgemmFP16UKernelsAvx512.cc	11:132 11:132	view
56	x 2	bench bench	EmbeddingSpMDMNBitBenchmark.cc EmbeddingSpMDMNBitRowWiseSparseBenchm...	285:364 321:400	view
31	x 2	src src	GenerateKernel.h DirectConv.h	97:129 104:136	view
27	x 2	src src	GenerateKernelDirectConvU8S8S32ACC32.cc GenerateKernelU8S8S32ACC32.cc	41:70 67:96	view
19	x 3	src src src	PackAWithRowOffset.cc PackAWithIm2Col.cc PackAWithQuantRowOffset.cc	211:232 725:746 232:252	view
16	x 3	src src src	PackAWithRowOffset.cc PackAWithQuantRowOffset.cc PackAMatrix.cc	189:208 210:229 189:208	view
12	x 3	src src src	PackAWithRowOffset.cc PackAWithQuantRowOffset.cc PackAMatrix.cc	170:186 191:207 170:186	view
9	x 2	bench bench	EmbeddingQuantizeBenchmark.cc Im2ColFusedRequantizeBenchmark.cc	100:111 355:366	view
8	x 7	bench bench bench bench bench bench bench	RequantizeBenchmark.cc GEMMsBenchmark.cc PackedRequantizeAcc16Benchmark.cc PackedRequantizeAcc32Benchmark.cc RowOffsetBenchmark.cc GroupwiseConvRequantizeBenchmark.cc PackedFloatInOutBenchmark.cc	139:149 326:336 458:468 333:343 43:53 540:550 309:319	view
8	x 2	bench bench	EmbeddingSpMDMNBitBenchmark.cc EmbeddingSpMDMNBitRowWiseSparseBenchm...	29:38 29:38	view