anthropics / triton
File Churn

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

Overview
File Change Frequency Overall
  • There are 201 files with 44,383 lines of code.
    • 1 file changed more than 100 times (1,185 lines of code)
    • 4 files changed 51-100 times (4,621 lines of code)
    • 32 files changed 21-50 times (14,874 lines of code)
    • 70 files changed 6-20 times (15,740 lines of code)
    • 94 files changed 1-5 times (7,963 lines of code)
2% | 10% | 33% | 35% | 17%
Legend:
101+
51-100
21-50
6-20
1-5

explore: grouped by folders | grouped by update frequency | data
Contributors Count Frequency Overall
  • There are 201 files with 44,383 lines of code.
    • 3 files changed by more than 25 contributors (2,709 lines of code)
    • 28 files changed by 11-25 contributors (15,943 lines of code)
    • 43 files changed by 6-10 contributors (12,856 lines of code)
    • 84 files changed by 2-5 contributors (9,853 lines of code)
    • 43 files changed by 1 contributor (3,022 lines of code)
6% | 35% | 28% | 22% | 6%
Legend:
26+
11-25
6-10
2-5
1

explore: grouped by folders | grouped by contributors count | data
File Change Frequency per File Extension
py, h, cpp, txt, mlir, yaml, rst, md, cc, sh, c, hpp, svg, in, toml, cmake, ll, editorconfig, html, gitignore, clang-format, gitmodules
File Change Frequency per Extension
The number of recorded file updates
101+
51-100
21-50
6-20
1-5
py10% | 13% | 28% | 41% | 6%
cpp0% | 11% | 37% | 36% | 14%
h0% | 0% | 44% | 18% | 36%
c0% | 0% | 0% | 76% | 23%
yaml0% | 0% | 0% | 100% | 0%
cc0% | 0% | 0% | <1% | 99%
in0% | 0% | 0% | 100% | 0%
hpp0% | 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
python8% | 11% | 23% | 37% | 19%
lib0% | 11% | 41% | 36% | 11%
include0% | 0% | 15% | 29% | 55%
ROOT0% | 0% | 0% | 100% | 0%
unittest0% | 0% | 0% | 0% | 100%
.github0% | 0% | 0% | 0% | 100%
Most Frequently Changed Files (Top 50)

See data for all files...

File# lines# unitscreatedlast modified# changes
(days)		# contributorsline
churn		first
contributor		latest
contributor
core.py
in python/triton/language
 1185 204 2021-09-02 2023-12-17 121 44 +4257 / -2067 phil@openai.com phil@openai.com
semantic.py
in python/triton/language
 1095 95 2022-03-17 2023-12-21 87 36 +3833 / -2239 dyanab@connect.ust.hk phil@openai.com
jit.py
in python/triton/runtime
 429 44 2022-09-18 2023-12-25 65 32 +1482 / -867 phil@openai.com phil@openai.com
ElementwiseOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1465 39 2022-12-21 2023-12-15 55 19 +4329 / -2574 phil@openai.com 153013546+pawelszczerbuk@us...
Dialect.cpp
in lib/Dialect/TritonGPU/IR
 1632 108 2022-12-21 2023-12-13 53 19 +2766 / -857 phil@openai.com kerenzhou@openai.com
Utility.cpp
in lib/Analysis
 609 53 2022-12-21 2023-12-15 47 14 +1093 / -308 phil@openai.com manman.ren@gmail.com
compiler.py
in python/triton/compiler
 198 19 2023-04-02 2023-12-26 45 25 +1347 / -1074 phil@openai.com phil@openai.com
RemoveLayoutConversions.cpp
in lib/Dialect/TritonGPU/Transforms
 853 26 2023-02-22 2023-12-18 42 14 +2101 / -1099 phil@openai.com phil@openai.com
code_generator.py
in python/triton/compiler
 963 81 2023-04-02 2023-12-26 42 14 +1529 / -295 phil@openai.com phil@openai.com
Utility.cpp
in lib/Dialect/TritonGPU/Transforms
 599 33 2023-01-14 2023-12-12 41 13 +1283 / -555 yanchunwei@outlook.com phil@openai.com
TritonGPUToLLVMPass.cpp
in lib/Conversion/TritonGPUToLLVM
 748 14 2022-12-21 2023-12-15 40 16 +1329 / -404 phil@openai.com 153013546+pawelszczerbuk@us...
ConvertLayoutOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 915 4 2022-12-21 2023-12-12 40 14 +1586 / -526 phil@openai.com phil@openai.com
TritonGPUToLLVMBase.h
in lib/Conversion/TritonGPUToLLVM
 991 37 2022-12-21 2023-12-12 39 17 +1849 / -612 phil@openai.com phil@openai.com
__init__.py
in python/triton
 51 2 2020-02-21 2023-11-30 38 11 +126 / -69 ptillet@g.harvard.edu phil@openai.com
TritonToTritonGPUPass.cpp
in lib/Conversion/TritonToTritonGPU
 608 10 2022-12-21 2023-12-09 38 17 +1348 / -560 phil@openai.com justin.lebar@gmail.com
LoadStoreOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1500 14 2022-12-21 2023-12-13 38 15 +2269 / -419 phil@openai.com kerenzhou@openai.com
matmul.py
in python/triton/ops
 174 7 2021-01-29 2023-12-12 37 18 +776 / -552 phil@openai.com karupayun@google.com
03-matrix-multiplication.py
in python/tutorials
 119 4 2021-03-14 2023-11-03 36 16 +1153 / -802 phil@openai.com justin.lebar@gmail.com
06-fused-attention.py
in python/tutorials
 481 10 2022-06-27 2023-11-13 36 14 +1535 / -901 phil@openai.com thomas.raoux@openai.com
__init__.py
in python/triton/language
 217 - 2021-09-02 2023-12-09 35 15 +270 / -48 phil@openai.com justin.lebar@gmail.com
01-vector-add.py
in python/tutorials
 59 3 2021-03-04 2023-12-26 31 13 +541 / -407 phil@openai.com phil@openai.com
TritonGPUToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 695 7 2022-12-21 2023-12-07 30 14 +1202 / -304 phil@openai.com kerenzhou@openai.com
Ops.cpp
in lib/Dialect/Triton/IR
 843 51 2022-12-21 2023-12-15 30 12 +1244 / -192 phil@openai.com 153013546+pawelszczerbuk@us...
matmul.py
in python/triton/ops/blocksparse
 306 11 2021-01-29 2023-11-03 29 15 +1905 / -1567 phil@openai.com justin.lebar@gmail.com
02-fused-softmax.py
in python/tutorials
 81 4 2021-03-06 2023-11-03 26 12 +544 / -350 phil@openai.com justin.lebar@gmail.com
autotuner.py
in python/triton/runtime
 207 15 2022-09-18 2023-11-30 26 15 +493 / -159 phil@openai.com kerenzhou@openai.com
ReduceOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 425 12 2022-12-21 2023-11-29 26 11 +1369 / -844 phil@openai.com karupayun@google.com
AxisInfo.cpp
in lib/Analysis
 905 47 2022-12-21 2023-12-05 26 10 +1442 / -343 phil@openai.com thomas.raoux@openai.com
random.py
in python/triton/language
 95 10 2021-09-03 2023-11-16 25 9 +533 / -352 informatyker@gmail.com phil@openai.com
Utility.h
in include/triton/Analysis
 249 24 2022-12-21 2023-10-06 25 8 +400 / -46 phil@openai.com 128723247+zahimoud@users.no...
AccelerateMatmul.cpp
in lib/Dialect/TritonGPU/Transforms
 301 5 2023-02-22 2023-12-18 24 11 +644 / -282 phil@openai.com phil@openai.com
Allocation.cpp
in lib/Analysis
 467 19 2022-12-21 2023-12-12 24 14 +747 / -98 phil@openai.com phil@openai.com
softmax.py
in python/triton/ops/blocksparse
 183 8 2021-01-31 2023-11-03 23 10 +745 / -521 phil@openai.com justin.lebar@gmail.com
Dialect.h
in include/triton/Dialect/TritonGPU/IR
 55 - 2022-12-21 2023-11-29 21 11 +150 / -29 phil@openai.com chengjun.lu@intel.com
OptimizeDotOperands.cpp
in lib/Dialect/TritonGPU/Transforms
 250 3 2023-03-07 2023-12-18 21 9 +702 / -390 phil@openai.com phil@openai.com
Utility.h
in lib/Conversion/TritonGPUToLLVM
 272 8 2022-12-21 2023-11-29 21 8 +515 / -168 phil@openai.com karupayun@google.com
SharedToDotOperandMMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM
 455 7 2023-03-21 2023-12-12 21 10 +1205 / -565 phil@openai.com phil@openai.com
TypeConverter.cpp
in lib/Conversion/TritonGPUToLLVM
 138 5 2023-03-06 2023-12-12 19 9 +294 / -134 phil@openai.com phil@openai.com
flash_attention.py
in python/triton/ops
 381 6 2023-02-01 2023-11-03 18 12 +896 / -440 phil@openai.com justin.lebar@gmail.com
Membar.cpp
in lib/Analysis
 157 5 2022-12-21 2023-12-13 17 8 +363 / -165 phil@openai.com kerenzhou@openai.com
05-layer-norm.py
in python/tutorials
 237 7 2021-10-08 2023-12-26 17 10 +820 / -427 phil@openai.com daniel.hesslow@gmail.com
Prefetch.cpp
in lib/Dialect/TritonGPU/Transforms
 280 7 2022-12-21 2023-12-18 17 9 +450 / -63 phil@openai.com phil@openai.com
Utility.cpp
in lib/Conversion/TritonGPUToLLVM
 315 25 2023-01-28 2023-11-29 17 9 +472 / -98 kerenzhou@openai.com karupayun@google.com
matmul_perf_model.py
in python/triton/ops
 117 5 2021-12-21 2023-12-18 16 9 +360 / -76 dyanab@connect.ust.hk phil@openai.com
DotOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 165 1 2022-12-21 2023-12-12 16 9 +538 / -343 phil@openai.com phil@openai.com
ViewOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 242 7 2022-12-21 2023-12-09 16 9 +484 / -179 phil@openai.com justin.lebar@gmail.com
WGMMA.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 481 16 2023-08-07 2023-12-12 16 7 +711 / -136 thomas.raoux@openai.com phil@openai.com
cross_entropy.py
in python/triton/ops
 62 5 2021-02-07 2023-11-03 15 10 +216 / -124 jareddk@gmail.com justin.lebar@gmail.com
Coalesce.cpp
in lib/Dialect/TritonGPU/Transforms
 242 9 2022-12-21 2023-12-21 15 9 +592 / -281 phil@openai.com manman.ren@gmail.com
build_extern.py
in python/triton/tools
 307 26 2022-07-13 2023-11-03 15 8 +552 / -176 kerenzhou@openai.com justin.lebar@gmail.com
Files With Most Churn (Top 50)
Based on the number of lines added and deleted across all commits.

See data for all files...

File# lines# unitscreatedlast modified# changes
(days)		# contributorsline
churn		first
contributor		latest
contributor
ElementwiseOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1465 39 2022-12-21 2023-12-15 55 19 +4329 / -2574 phil@openai.com 153013546+pawelszczerbuk@us...
core.py
in python/triton/language
 1185 204 2021-09-02 2023-12-17 121 44 +4257 / -2067 phil@openai.com phil@openai.com
semantic.py
in python/triton/language
 1095 95 2022-03-17 2023-12-21 87 36 +3833 / -2239 dyanab@connect.ust.hk phil@openai.com
math.py
in python/triton/language
 1292 200 2023-03-19 2023-12-14 12 7 +3175 / -1474 phil@openai.com 128723247+zahimoud@users.no...
Dialect.cpp
in lib/Dialect/TritonGPU/IR
 1632 108 2022-12-21 2023-12-13 53 19 +2766 / -857 phil@openai.com kerenzhou@openai.com
matmul.py
in python/triton/ops/blocksparse
 306 11 2021-01-29 2023-11-03 29 15 +1905 / -1567 phil@openai.com justin.lebar@gmail.com
RemoveLayoutConversions.cpp
in lib/Dialect/TritonGPU/Transforms
 853 26 2023-02-22 2023-12-18 42 14 +2101 / -1099 phil@openai.com phil@openai.com
NVGPUToLLVMPass.cpp
in lib/Conversion/NVGPUToLLVM
 982 31 2023-08-10 2023-11-30 9 4 +2077 / -944 128723247+zahimoud@users.no... thomas.raoux@openai.com
LoadStoreOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1500 14 2022-12-21 2023-12-13 38 15 +2269 / -419 phil@openai.com kerenzhou@openai.com
TritonGPUToLLVMBase.h
in lib/Conversion/TritonGPUToLLVM
 991 37 2022-12-21 2023-12-12 39 17 +1849 / -612 phil@openai.com phil@openai.com
06-fused-attention.py
in python/tutorials
 481 10 2022-06-27 2023-11-13 36 14 +1535 / -901 phil@openai.com thomas.raoux@openai.com
compiler.py
in python/triton/compiler
 198 19 2023-04-02 2023-12-26 45 25 +1347 / -1074 phil@openai.com phil@openai.com
jit.py
in python/triton/runtime
 429 44 2022-09-18 2023-12-25 65 32 +1482 / -867 phil@openai.com phil@openai.com
ReduceOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 425 12 2022-12-21 2023-11-29 26 11 +1369 / -844 phil@openai.com karupayun@google.com
ConvertLayoutOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 915 4 2022-12-21 2023-12-12 40 14 +1586 / -526 phil@openai.com phil@openai.com
03-matrix-multiplication.py
in python/tutorials
 119 4 2021-03-14 2023-11-03 36 16 +1153 / -802 phil@openai.com justin.lebar@gmail.com
TritonToTritonGPUPass.cpp
in lib/Conversion/TritonToTritonGPU
 608 10 2022-12-21 2023-12-09 38 17 +1348 / -560 phil@openai.com justin.lebar@gmail.com
Utility.cpp
in lib/Dialect/TritonGPU/Transforms
 599 33 2023-01-14 2023-12-12 41 13 +1283 / -555 yanchunwei@outlook.com phil@openai.com
code_generator.py
in python/triton/compiler
 963 81 2023-04-02 2023-12-26 42 14 +1529 / -295 phil@openai.com phil@openai.com
AxisInfo.cpp
in lib/Analysis
 905 47 2022-12-21 2023-12-05 26 10 +1442 / -343 phil@openai.com thomas.raoux@openai.com
SharedToDotOperandMMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM
 455 7 2023-03-21 2023-12-12 21 10 +1205 / -565 phil@openai.com phil@openai.com
TritonGPUToLLVMPass.cpp
in lib/Conversion/TritonGPUToLLVM
 748 14 2022-12-21 2023-12-15 40 16 +1329 / -404 phil@openai.com 153013546+pawelszczerbuk@us...
ir.cc
in python/src
 1443 13 2023-12-18 2023-12-26 3 1 +1602 / -41 phil@openai.com phil@openai.com
TritonGPUToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 695 7 2022-12-21 2023-12-07 30 14 +1202 / -304 phil@openai.com kerenzhou@openai.com
Ops.cpp
in lib/Dialect/Triton/IR
 843 51 2022-12-21 2023-12-15 30 12 +1244 / -192 phil@openai.com 153013546+pawelszczerbuk@us...
Utility.cpp
in lib/Analysis
 609 53 2022-12-21 2023-12-15 47 14 +1093 / -308 phil@openai.com manman.ren@gmail.com
flash_attention.py
in python/triton/ops
 381 6 2023-02-01 2023-11-03 18 12 +896 / -440 phil@openai.com justin.lebar@gmail.com
matmul.py
in python/triton/ops
 174 7 2021-01-29 2023-12-12 37 18 +776 / -552 phil@openai.com karupayun@google.com
softmax.py
in python/triton/ops/blocksparse
 183 8 2021-01-31 2023-11-03 23 10 +745 / -521 phil@openai.com justin.lebar@gmail.com
WSPipeline.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 811 15 2023-08-07 2023-12-13 12 7 +1217 / -41 109190422+goostavz@users.no... kerenzhou@openai.com
05-layer-norm.py
in python/tutorials
 237 7 2021-10-08 2023-12-26 17 10 +820 / -427 phil@openai.com daniel.hesslow@gmail.com
OptimizeDotOperands.cpp
in lib/Dialect/TritonGPU/Transforms
 250 3 2023-03-07 2023-12-18 21 9 +702 / -390 phil@openai.com phil@openai.com
PlanCTA.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 821 45 2023-08-07 2023-12-08 8 7 +1062 / -26 109190422+goostavz@users.no... 153013546+pawelszczerbuk@us...
MatmulLoopPipeline.cpp
in lib/Dialect/TritonGPU/Transforms/Pipeliner
 691 16 2023-11-02 2023-12-13 7 4 +959 / -123 thomas.raoux@openai.com kerenzhou@openai.com
RewriteTensorPointer.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 645 19 2023-08-07 2023-12-18 9 7 +958 / -81 109190422+goostavz@users.no... thomas.raoux@openai.com
WSMaterialization.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 555 18 2023-08-07 2023-12-13 8 9 +874 / -106 109190422+goostavz@users.no... kerenzhou@openai.com
01-vector-add.py
in python/tutorials
 59 3 2021-03-04 2023-12-26 31 13 +541 / -407 phil@openai.com phil@openai.com
AccelerateMatmul.cpp
in lib/Dialect/TritonGPU/Transforms
 301 5 2023-02-22 2023-12-18 24 11 +644 / -282 phil@openai.com phil@openai.com
02-fused-softmax.py
in python/tutorials
 81 4 2021-03-06 2023-11-03 26 12 +544 / -350 phil@openai.com justin.lebar@gmail.com
random.py
in python/triton/language
 95 10 2021-09-03 2023-11-16 25 9 +533 / -352 informatyker@gmail.com phil@openai.com
DotOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 165 1 2022-12-21 2023-12-12 16 9 +538 / -343 phil@openai.com phil@openai.com
Coalesce.cpp
in lib/Dialect/TritonGPU/Transforms
 242 9 2022-12-21 2023-12-21 15 9 +592 / -281 phil@openai.com manman.ren@gmail.com
PipelineExpander.cpp
in lib/Dialect/TritonGPU/Transforms/Pipeliner
 555 6 2023-11-02 2023-12-14 5 4 +793 / -70 thomas.raoux@openai.com kerenzhou@openai.com
WGMMA.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 481 16 2023-08-07 2023-12-12 16 7 +711 / -136 thomas.raoux@openai.com phil@openai.com
Allocation.cpp
in lib/Analysis
 467 19 2022-12-21 2023-12-12 24 14 +747 / -98 phil@openai.com phil@openai.com
build_extern.py
in python/triton/tools
 307 26 2022-07-13 2023-11-03 15 8 +552 / -176 kerenzhou@openai.com justin.lebar@gmail.com
EmitIndicesTest.cpp
in unittest/Conversion/TritonGPUToLLVM
 509 19 2023-08-07 2023-12-12 4 4 +686 / -7 109190422+goostavz@users.no... phil@openai.com
make_launcher.py
in python/triton/compiler
 85 4 2023-04-02 2023-11-03 13 9 +490 / -193 phil@openai.com justin.lebar@gmail.com
Utility.h
in lib/Conversion/TritonGPUToLLVM
 272 8 2022-12-21 2023-11-29 21 8 +515 / -168 phil@openai.com karupayun@google.com
ViewOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 242 7 2022-12-21 2023-12-09 16 9 +484 / -179 phil@openai.com justin.lebar@gmail.com
Files With Most Contributors (Top 50)
Based on the number of unique email addresses found in commits.

See data for all files...

File# lines# unitscreatedlast modified# changes
(days)		# contributorsline
churn		first
contributor		latest
contributor
core.py
in python/triton/language
 1185 204 2021-09-02 2023-12-17 121 44 +4257 / -2067 phil@openai.com phil@openai.com
semantic.py
in python/triton/language
 1095 95 2022-03-17 2023-12-21 87 36 +3833 / -2239 dyanab@connect.ust.hk phil@openai.com
jit.py
in python/triton/runtime
 429 44 2022-09-18 2023-12-25 65 32 +1482 / -867 phil@openai.com phil@openai.com
compiler.py
in python/triton/compiler
 198 19 2023-04-02 2023-12-26 45 25 +1347 / -1074 phil@openai.com phil@openai.com
ElementwiseOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1465 39 2022-12-21 2023-12-15 55 19 +4329 / -2574 phil@openai.com 153013546+pawelszczerbuk@us...
Dialect.cpp
in lib/Dialect/TritonGPU/IR
 1632 108 2022-12-21 2023-12-13 53 19 +2766 / -857 phil@openai.com kerenzhou@openai.com
matmul.py
in python/triton/ops
 174 7 2021-01-29 2023-12-12 37 18 +776 / -552 phil@openai.com karupayun@google.com
TritonGPUToLLVMBase.h
in lib/Conversion/TritonGPUToLLVM
 991 37 2022-12-21 2023-12-12 39 17 +1849 / -612 phil@openai.com phil@openai.com
TritonToTritonGPUPass.cpp
in lib/Conversion/TritonToTritonGPU
 608 10 2022-12-21 2023-12-09 38 17 +1348 / -560 phil@openai.com justin.lebar@gmail.com
TritonGPUToLLVMPass.cpp
in lib/Conversion/TritonGPUToLLVM
 748 14 2022-12-21 2023-12-15 40 16 +1329 / -404 phil@openai.com 153013546+pawelszczerbuk@us...
03-matrix-multiplication.py
in python/tutorials
 119 4 2021-03-14 2023-11-03 36 16 +1153 / -802 phil@openai.com justin.lebar@gmail.com
LoadStoreOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1500 14 2022-12-21 2023-12-13 38 15 +2269 / -419 phil@openai.com kerenzhou@openai.com
__init__.py
in python/triton/language
 217 - 2021-09-02 2023-12-09 35 15 +270 / -48 phil@openai.com justin.lebar@gmail.com
matmul.py
in python/triton/ops/blocksparse
 306 11 2021-01-29 2023-11-03 29 15 +1905 / -1567 phil@openai.com justin.lebar@gmail.com
autotuner.py
in python/triton/runtime
 207 15 2022-09-18 2023-11-30 26 15 +493 / -159 phil@openai.com kerenzhou@openai.com
Utility.cpp
in lib/Analysis
 609 53 2022-12-21 2023-12-15 47 14 +1093 / -308 phil@openai.com manman.ren@gmail.com
RemoveLayoutConversions.cpp
in lib/Dialect/TritonGPU/Transforms
 853 26 2023-02-22 2023-12-18 42 14 +2101 / -1099 phil@openai.com phil@openai.com
code_generator.py
in python/triton/compiler
 963 81 2023-04-02 2023-12-26 42 14 +1529 / -295 phil@openai.com phil@openai.com
ConvertLayoutOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 915 4 2022-12-21 2023-12-12 40 14 +1586 / -526 phil@openai.com phil@openai.com
06-fused-attention.py
in python/tutorials
 481 10 2022-06-27 2023-11-13 36 14 +1535 / -901 phil@openai.com thomas.raoux@openai.com
TritonGPUToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 695 7 2022-12-21 2023-12-07 30 14 +1202 / -304 phil@openai.com kerenzhou@openai.com
Allocation.cpp
in lib/Analysis
 467 19 2022-12-21 2023-12-12 24 14 +747 / -98 phil@openai.com phil@openai.com
Utility.cpp
in lib/Dialect/TritonGPU/Transforms
 599 33 2023-01-14 2023-12-12 41 13 +1283 / -555 yanchunwei@outlook.com phil@openai.com
01-vector-add.py
in python/tutorials
 59 3 2021-03-04 2023-12-26 31 13 +541 / -407 phil@openai.com phil@openai.com
Ops.cpp
in lib/Dialect/Triton/IR
 843 51 2022-12-21 2023-12-15 30 12 +1244 / -192 phil@openai.com 153013546+pawelszczerbuk@us...
02-fused-softmax.py
in python/tutorials
 81 4 2021-03-06 2023-11-03 26 12 +544 / -350 phil@openai.com justin.lebar@gmail.com
flash_attention.py
in python/triton/ops
 381 6 2023-02-01 2023-11-03 18 12 +896 / -440 phil@openai.com justin.lebar@gmail.com
__init__.py
in python/triton
 51 2 2020-02-21 2023-11-30 38 11 +126 / -69 ptillet@g.harvard.edu phil@openai.com
ReduceOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 425 12 2022-12-21 2023-11-29 26 11 +1369 / -844 phil@openai.com karupayun@google.com
AccelerateMatmul.cpp
in lib/Dialect/TritonGPU/Transforms
 301 5 2023-02-22 2023-12-18 24 11 +644 / -282 phil@openai.com phil@openai.com
Dialect.h
in include/triton/Dialect/TritonGPU/IR
 55 - 2022-12-21 2023-11-29 21 11 +150 / -29 phil@openai.com chengjun.lu@intel.com
AxisInfo.cpp
in lib/Analysis
 905 47 2022-12-21 2023-12-05 26 10 +1442 / -343 phil@openai.com thomas.raoux@openai.com
softmax.py
in python/triton/ops/blocksparse
 183 8 2021-01-31 2023-11-03 23 10 +745 / -521 phil@openai.com justin.lebar@gmail.com
SharedToDotOperandMMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM
 455 7 2023-03-21 2023-12-12 21 10 +1205 / -565 phil@openai.com phil@openai.com
05-layer-norm.py
in python/tutorials
 237 7 2021-10-08 2023-12-26 17 10 +820 / -427 phil@openai.com daniel.hesslow@gmail.com
cross_entropy.py
in python/triton/ops
 62 5 2021-02-07 2023-11-03 15 10 +216 / -124 jareddk@gmail.com justin.lebar@gmail.com
random.py
in python/triton/language
 95 10 2021-09-03 2023-11-16 25 9 +533 / -352 informatyker@gmail.com phil@openai.com
OptimizeDotOperands.cpp
in lib/Dialect/TritonGPU/Transforms
 250 3 2023-03-07 2023-12-18 21 9 +702 / -390 phil@openai.com phil@openai.com
TypeConverter.cpp
in lib/Conversion/TritonGPUToLLVM
 138 5 2023-03-06 2023-12-12 19 9 +294 / -134 phil@openai.com phil@openai.com
Prefetch.cpp
in lib/Dialect/TritonGPU/Transforms
 280 7 2022-12-21 2023-12-18 17 9 +450 / -63 phil@openai.com phil@openai.com
Utility.cpp
in lib/Conversion/TritonGPUToLLVM
 315 25 2023-01-28 2023-11-29 17 9 +472 / -98 kerenzhou@openai.com karupayun@google.com
ViewOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 242 7 2022-12-21 2023-12-09 16 9 +484 / -179 phil@openai.com justin.lebar@gmail.com
DotOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 165 1 2022-12-21 2023-12-12 16 9 +538 / -343 phil@openai.com phil@openai.com
matmul_perf_model.py
in python/triton/ops
 117 5 2021-12-21 2023-12-18 16 9 +360 / -76 dyanab@connect.ust.hk phil@openai.com
Coalesce.cpp
in lib/Dialect/TritonGPU/Transforms
 242 9 2022-12-21 2023-12-21 15 9 +592 / -281 phil@openai.com manman.ren@gmail.com
make_launcher.py
in python/triton/compiler
 85 4 2023-04-02 2023-11-03 13 9 +490 / -193 phil@openai.com justin.lebar@gmail.com
backend.py
in python/triton/common
 115 21 2023-06-09 2023-12-20 11 9 +203 / -19 eikan.wang@intel.com phil@openai.com
WSMaterialization.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 555 18 2023-08-07 2023-12-13 8 9 +874 / -106 109190422+goostavz@users.no... kerenzhou@openai.com
Utility.h
in include/triton/Analysis
 249 24 2022-12-21 2023-10-06 25 8 +400 / -46 phil@openai.com 128723247+zahimoud@users.no...
Utility.h
in lib/Conversion/TritonGPUToLLVM
 272 8 2022-12-21 2023-11-29 21 8 +515 / -168 phil@openai.com karupayun@google.com
Files With Least Contributors (Top 50)
Based on the number of unique email addresses found in commits.

See data for all files...

File# lines# unitscreatedlast modified# changes
(days)		# contributorsline
churn		first
contributor		latest
contributor
Utility.h
in python/triton/_C/include/triton/Analysis
 249 24
AxisInfo.h
in python/triton/_C/include/triton/Analysis
 247 19
GCNAsmFormat.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 241 7
PTXAsmFormat.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 180 7
Allocation.h
in python/triton/_C/include/triton/Analysis
 177 21
Traits.h
in python/triton/_C/include/triton/Dialect/Triton/IR
 89 8
Membar.h
in python/triton/_C/include/triton/Analysis
 82 5
Utility.h
in python/triton/_C/include/triton/Dialect/TritonGPU/Transforms
 72 -
TmaMetadata.h
in python/triton/_C/include/triton/Target/PTX
 55 2
Dialect.h
in python/triton/_C/include/triton/Dialect/TritonGPU/IR
 55 -
Utility.h
in python/triton/_C/include/triton/Dialect/TritonNvidiaGPU/Transforms
 53 6
Alias.h
in python/triton/_C/include/triton/Analysis
 48 5
Dialect.h
in python/triton/_C/include/triton/Dialect/Triton/IR
 45 -
Passes.h
in python/triton/_C/include/triton/Dialect/TritonNvidiaGPU/Transforms
 41 -
GetEnv.hpp
in python/triton/_C/include/triton/Tools/Sys
 32 2
Traits.h
in python/triton/_C/include/triton/Dialect/TritonGPU/IR
 30 2
MLIRTypes.h
in python/triton/_C/include/triton/Conversion
 28 11
Passes.h
in python/triton/_C/include/triton/Dialect/TritonGPU/Transforms
 27 -
TritonGPUConversion.h
in python/triton/_C/include/triton/Dialect/TritonGPU/Transforms
 24 3
Types.h
in python/triton/_C/include/triton/Dialect/Triton/IR
 21 -
Traits.h
in python/triton/_C/include/triton/Dialect/TritonNvidiaGPU/IR
 21 1
Dialect.h
in python/triton/_C/include/triton/Dialect/TritonNvidiaGPU/IR
 21 -
AsmFormat.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 21 1
TritonGPUToLLVMPass.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 20 -
Dialect.h
in python/triton/_C/include/triton/Dialect/NVGPU/IR
 19 -
TritonToTritonGPUPass.h
in python/triton/_C/include/triton/Conversion/TritonToTritonGPU
 19 -
Passes.h
in python/triton/_C/include/triton/Dialect/Triton/Transforms
 14 -
Passes.h
in python/triton/_C/include/triton/Conversion/TritonToTritonGPU
 11 -
Passes.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 11 -
NVGPUToLLVMPass.h
in python/triton/_C/include/triton/Conversion/NVGPUToLLVM
 11 -
Passes.h
in python/triton/_C/include/triton/Conversion/NVGPUToLLVM
 11 -
Passes.h
in python/triton/_C/include/triton/Target/LLVMIR
 9 -
GetPlatform.hpp
in python/triton/_C/include/triton/Tools/Sys
 9 -
Types.h
in python/triton/_C/include/triton/Dialect/TritonGPU/IR
 7 -
Types.h
in python/triton/_C/include/triton/Dialect/TritonNvidiaGPU/IR
 7 -
Attributes.h
in python/triton/_C/include/triton/Dialect/TritonGPU/IR
 6 -
Interfaces.h
in python/triton/_C/include/triton/Dialect/Triton/IR
 6 -
TritonGPUInterfaces.h
in python/triton/_C/include/triton/Dialect/TritonGPU/IR
 4 -
ir.cc
in python/src
 1443 13 2023-12-18 2023-12-26 3 1 +1602 / -41 phil@openai.com phil@openai.com
GCNAsmFormat.h
in include/triton/Conversion/TritonGPUToLLVM
 241 7 2023-03-25 2023-03-29 2 1 +381 / -0 micmelesse@gmail.com micmelesse@gmail.com
llvm.cc
in python/src
 172 3 2023-12-25 2023-12-26 2 1 +326 / -118 phil@openai.com phil@openai.com
GCNAsmFormat.cpp
in lib/Conversion/TritonGPUToLLVM
 161 8 2023-03-25 2023-03-25 1 1 +191 / -0 micmelesse@gmail.com micmelesse@gmail.com
nvidia.cc
in python/src
 120 3 2023-12-25 2023-12-26 2 1 +143 / -1 phil@openai.com phil@openai.com
LLVMDIScope.cpp
in lib/Target/LLVMIR
 115 6 2023-07-07 2023-07-07 1 1 +148 / -0 kerenzhou@openai.com kerenzhou@openai.com
TensorPtrOpsToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 63 1 2023-08-07 2023-08-07 1 1 +104 / -0 109190422+goostavz@users.no... 109190422+goostavz@users.no...
passes.cc
in python/src
 62 6 2023-12-18 2023-12-25 2 1 +125 / -56 phil@openai.com phil@openai.com
TmaMetadata.h
in include/triton/Target/PTX
 55 2 2023-08-07 2023-08-07 1 1 +107 / -0 109190422+goostavz@users.no... 109190422+goostavz@users.no...
WSFixupMissingAttrs.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 54 2 2023-08-08 2023-08-08 1 1 +69 / -0 138413300+allatit23@users.n... 138413300+allatit23@users.n...
LLVMIRBreakPhiStruct.cpp
in lib/Target/LLVMIR
 50 3 2023-09-13 2023-09-13 1 1 +60 / -0 thomas.raoux@openai.com thomas.raoux@openai.com
interpreter.cc
in python/src
 40 1 2023-12-17 2023-12-17 1 1 +44 / -0 phil@openai.com phil@openai.com
Correlations

File Size vs. Number of Changes: 201 points

python/tutorials/04-low-memory-dropout.py x: 73 lines of code y: 8 # changes python/src/ir.cc x: 1443 lines of code y: 3 # changes python/src/llvm.cc x: 172 lines of code y: 2 # changes python/src/nvidia.cc x: 120 lines of code y: 2 # changes python/triton/compiler/backends/cuda.py x: 248 lines of code y: 9 # changes python/triton/compiler/code_generator.py x: 963 lines of code y: 42 # changes python/triton/compiler/compiler.py x: 198 lines of code y: 45 # changes python/tutorials/01-vector-add.py x: 59 lines of code y: 31 # changes python/tutorials/05-layer-norm.py x: 237 lines of code y: 17 # changes python/src/main.cc x: 17 lines of code y: 8 # changes python/src/passes.cc x: 62 lines of code y: 2 # changes python/src/passes.h x: 16 lines of code y: 1 # changes python/triton/runtime/jit.py x: 429 lines of code y: 65 # changes python/triton/language/semantic.py x: 1095 lines of code y: 87 # changes lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 242 lines of code y: 15 # changes lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 127 lines of code y: 10 # changes python/triton/common/backend.py x: 115 lines of code y: 11 # changes lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 645 lines of code y: 9 # changes include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 27 lines of code y: 9 # changes lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 301 lines of code y: 24 # changes lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 65 lines of code y: 6 # changes lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 250 lines of code y: 21 # changes lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 82 lines of code y: 4 # changes lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 375 lines of code y: 6 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 79 lines of code y: 3 # changes lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 280 lines of code y: 17 # changes lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 853 lines of code y: 42 # changes python/triton/ops/matmul_perf_model.py x: 117 lines of code y: 16 # changes python/triton/compiler/backends/__init__.py x: 3 lines of code y: 2 # changes python/triton/runtime/backends/hip.c x: 92 lines of code y: 3 # changes python/triton/runtime/driver.py x: 190 lines of code y: 12 # changes python/src/interpreter.cc x: 40 lines of code y: 1 # changes python/triton/language/core.py x: 1185 lines of code y: 121 # changes python/triton/runtime/interpreter.py x: 371 lines of code y: 6 # changes lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 1465 lines of code y: 55 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 748 lines of code y: 40 # changes lib/Dialect/Triton/IR/Ops.cpp x: 843 lines of code y: 30 # changes lib/Analysis/Utility.cpp x: 609 lines of code y: 47 # changes python/triton/language/math.py x: 1292 lines of code y: 12 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 555 lines of code y: 5 # changes python/triton/language/standard.py x: 240 lines of code y: 9 # changes include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 53 lines of code y: 2 # changes lib/Analysis/Alias.cpp x: 55 lines of code y: 9 # changes lib/Analysis/Membar.cpp x: 157 lines of code y: 17 # changes lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 1500 lines of code y: 38 # changes lib/Dialect/TritonGPU/IR/Dialect.cpp x: 1632 lines of code y: 53 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 691 lines of code y: 7 # changes lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 225 lines of code y: 5 # changes lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 396 lines of code y: 4 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 555 lines of code y: 8 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSMutex.cpp x: 277 lines of code y: 6 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 811 lines of code y: 12 # changes python/triton/ops/matmul.py x: 174 lines of code y: 37 # changes lib/Analysis/Allocation.cpp x: 467 lines of code y: 24 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 915 lines of code y: 40 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 369 lines of code y: 11 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 455 lines of code y: 21 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.cpp x: 165 lines of code y: 16 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/FMA.cpp x: 83 lines of code y: 5 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 124 lines of code y: 6 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 350 lines of code y: 10 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 481 lines of code y: 16 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 991 lines of code y: 39 # changes lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 138 lines of code y: 19 # changes lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 599 lines of code y: 41 # changes lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 96 lines of code y: 7 # changes unittest/Conversion/TritonGPUToLLVM/EmitIndicesTest.cpp x: 509 lines of code y: 4 # changes unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 45 lines of code y: 5 # changes lib/Conversion/TritonGPUToLLVM/ViewOpToLLVM.cpp x: 242 lines of code y: 16 # changes lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 608 lines of code y: 38 # changes python/triton/language/__init__.py x: 217 lines of code y: 35 # changes lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 821 lines of code y: 8 # changes include/triton/Dialect/TritonGPU/IR/Traits.h x: 30 lines of code y: 2 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 695 lines of code y: 30 # changes lib/Dialect/TritonGPU/IR/Traits.cpp x: 18 lines of code y: 5 # changes python/triton/runtime/cache.py x: 116 lines of code y: 9 # changes python/triton/runtime/backends/cuda.c x: 458 lines of code y: 10 # changes python/triton/compiler/__init__.py x: 3 lines of code y: 6 # changes python/triton/ops/__init__.py x: 5 lines of code y: 10 # changes python/triton/tools/compile.py x: 104 lines of code y: 9 # changes lib/Analysis/AxisInfo.cpp x: 905 lines of code y: 26 # changes python/triton/runtime/autotuner.py x: 207 lines of code y: 26 # changes python/triton/__init__.py x: 51 lines of code y: 38 # changes python/triton/compiler/utils.py x: 210 lines of code y: 6 # changes python/triton/language/extra/cuda.py x: 12 lines of code y: 6 # changes lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 982 lines of code y: 9 # changes lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 200 lines of code y: 3 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 195 lines of code y: 7 # changes lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 425 lines of code y: 26 # changes lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.cpp x: 373 lines of code y: 9 # changes lib/Conversion/TritonGPUToLLVM/Utility.cpp x: 315 lines of code y: 17 # changes lib/Conversion/TritonGPUToLLVM/Utility.h x: 272 lines of code y: 21 # changes lib/Dialect/NVGPU/IR/Dialect.cpp x: 71 lines of code y: 2 # changes lib/Dialect/Triton/IR/Dialect.cpp x: 63 lines of code y: 8 # changes unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 296 lines of code y: 4 # changes include/triton/Dialect/TritonGPU/IR/Dialect.h x: 55 lines of code y: 21 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSFeasibilityChecking.cpp x: 39 lines of code y: 2 # changes lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 83 lines of code y: 11 # changes include/triton/Analysis/Membar.h x: 82 lines of code y: 5 # changes include/triton/Dialect/Triton/IR/Traits.h x: 89 lines of code y: 3 # changes lib/Dialect/Triton/IR/Traits.cpp x: 211 lines of code y: 7 # changes lib/Dialect/TritonNvidiaGPU/IR/Traits.cpp x: 10 lines of code y: 2 # changes include/triton/Dialect/TritonGPU/IR/Attributes.h x: 6 lines of code y: 2 # changes include/triton/Dialect/TritonGPU/IR/TritonGPUInterfaces.h x: 4 lines of code y: 1 # changes python/triton/language/random.py x: 95 lines of code y: 25 # changes include/triton/Analysis/AxisInfo.h x: 247 lines of code y: 10 # changes python/tutorials/06-fused-attention.py x: 481 lines of code y: 36 # changes .pre-commit-config.yaml x: 52 lines of code y: 9 # changes python/triton/common/build.py x: 109 lines of code y: 6 # changes lib/Dialect/Triton/Transforms/RewriteTensorPointer.cpp x: 364 lines of code y: 7 # changes lib/Dialect/TritonNvidiaGPU/IR/Ops.cpp x: 57 lines of code y: 2 # changes .github/workflows/torch-inductor/scripts/check_perf.py x: 57 lines of code y: 3 # changes python/triton/compiler/make_launcher.py x: 85 lines of code y: 13 # changes python/triton/ops/blocksparse/matmul.py x: 306 lines of code y: 29 # changes python/triton/ops/blocksparse/softmax.py x: 183 lines of code y: 23 # changes python/triton/ops/cross_entropy.py x: 62 lines of code y: 15 # changes python/triton/ops/flash_attention.py x: 381 lines of code y: 18 # changes python/triton/runtime/__init__.py x: 17 lines of code y: 9 # changes python/triton/runtime/errors.py x: 10 lines of code y: 3 # changes python/triton/tools/build_extern.py x: 307 lines of code y: 15 # changes python/triton/tools/link.py x: 235 lines of code y: 7 # changes python/tutorials/02-fused-softmax.py x: 81 lines of code y: 26 # changes python/tutorials/03-matrix-multiplication.py x: 119 lines of code y: 36 # changes python/tutorials/07-math-functions.py x: 37 lines of code y: 3 # changes python/tutorials/08-experimental-block-pointer.py x: 85 lines of code y: 3 # changes python/tutorials/09-experimental-tma-matrix-multiplication.py x: 138 lines of code y: 5 # changes python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 104 lines of code y: 4 # changes python/tutorials/11-grouped-gemm.py x: 204 lines of code y: 3 # changes include/triton/Dialect/TritonGPU/Transforms/Utility.h x: 72 lines of code y: 11 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.h x: 37 lines of code y: 1 # changes .github/workflows/torch-inductor/scripts/check_acc.py x: 10 lines of code y: 1 # changes include/triton/Analysis/Alias.h x: 48 lines of code y: 4 # changes include/triton/Dialect/Triton/IR/Dialect.h x: 45 lines of code y: 8 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSDecomposing.cpp x: 179 lines of code y: 2 # changes include/triton/Analysis/Utility.h x: 249 lines of code y: 25 # changes include/triton/Tools/Sys/GetEnv.hpp x: 32 lines of code y: 5 # changes lib/Dialect/Triton/Transforms/ReorderBroadcast.cpp x: 201 lines of code y: 4 # changes lib/Dialect/Triton/Transforms/Combine.cpp x: 178 lines of code y: 11 # changes python/triton/tools/disasm.py x: 89 lines of code y: 8 # changes include/triton/Dialect/TritonNvidiaGPU/IR/Dialect.h x: 21 lines of code y: 3 # changes lib/Target/LLVMIR/LLVMIRBreakPhiStruct.cpp x: 50 lines of code y: 1 # changes include/triton/Analysis/Allocation.h x: 177 lines of code y: 9 # changes include/triton/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.h x: 20 lines of code y: 6 # changes lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.h x: 12 lines of code y: 8 # changes python/triton/tools/compile.c x: 50 lines of code y: 3 # changes python/triton/tools/compile.h x: 10 lines of code y: 4 # changes include/triton/Conversion/TritonGPUToLLVM/PTXAsmFormat.h x: 180 lines of code y: 5 # changes lib/Conversion/TritonGPUToLLVM/PTXAsmFormat.cpp x: 192 lines of code y: 5 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSFixupMissingAttrs.cpp x: 54 lines of code y: 1 # changes include/triton/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.h x: 19 lines of code y: 3 # changes include/triton/Dialect/NVGPU/IR/Dialect.h x: 19 lines of code y: 1 # changes include/triton/Dialect/Triton/IR/Types.h x: 21 lines of code y: 5 # changes include/triton/Dialect/Triton/Transforms/Passes.h x: 14 lines of code y: 4 # changes include/triton/Dialect/TritonGPU/IR/Types.h x: 7 lines of code y: 1 # changes include/triton/Dialect/TritonGPU/Transforms/TritonGPUConversion.h x: 24 lines of code y: 3 # changes include/triton/Dialect/TritonNvidiaGPU/IR/Traits.h x: 21 lines of code y: 1 # changes lib/Conversion/TritonGPUToLLVM/ClusterOpsToLLVM.cpp x: 34 lines of code y: 1 # changes lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.h x: 13 lines of code y: 5 # changes lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.h x: 12 lines of code y: 2 # changes lib/Conversion/TritonGPUToLLVM/TensorPtrOpsToLLVM.cpp x: 63 lines of code y: 1 # changes lib/Dialect/Triton/IR/Types.cpp x: 93 lines of code y: 6 # changes lib/Dialect/TritonGPU/IR/Types.cpp x: 27 lines of code y: 1 # changes lib/Dialect/TritonNvidiaGPU/IR/Types.cpp x: 14 lines of code y: 1 # changes lib/Target/LLVMIR/LLVMDIScope.cpp x: 115 lines of code y: 1 # changes python/MANIFEST.in x: 5 lines of code y: 7 # changes python/triton/common/__init__.py x: 2 lines of code y: 2 # changes python/triton/language/extra/__init__.py x: 2 lines of code y: 1 # changes include/triton/Conversion/TritonGPUToLLVM/AsmFormat.h x: 21 lines of code y: 2 # changes include/triton/Conversion/TritonGPUToLLVM/GCNAsmFormat.h x: 241 lines of code y: 2 # changes lib/Conversion/TritonGPUToLLVM/GCNAsmFormat.cpp x: 161 lines of code y: 1 # changes lib/Conversion/TritonGPUToLLVM/TypeConverter.h x: 22 lines of code y: 7 # changes unittest/Conversion/TritonGPUToLLVM/PTXAsmFormatTest.cpp x: 110 lines of code y: 2 # changes include/triton/Conversion/MLIRTypes.h x: 28 lines of code y: 2 # changes python/triton/ops/blocksparse/__init__.py x: 6 lines of code y: 6 # changes
121.0
# changes
  min: 1.0
  average: 11.53
  25th percentile: 2.0
  median: 6.0
  75th percentile: 14.0
  max: 121.0
0 1632.0
lines of code
min: 1.0 | average: 220.81 | 25th percentile: 21.0 | median: 89.0 | 75th percentile: 249.5 | max: 1632.0

Number of Contributors vs. Number of Changes: 201 points

python/tutorials/04-low-memory-dropout.py x: 7 # contributors y: 8 # changes python/src/ir.cc x: 1 # contributors y: 3 # changes python/src/llvm.cc x: 1 # contributors y: 2 # changes python/triton/compiler/backends/cuda.py x: 3 # contributors y: 9 # changes python/triton/compiler/code_generator.py x: 14 # contributors y: 42 # changes python/triton/compiler/compiler.py x: 25 # contributors y: 45 # changes python/tutorials/01-vector-add.py x: 13 # contributors y: 31 # changes python/tutorials/05-layer-norm.py x: 10 # contributors y: 17 # changes python/src/main.cc x: 3 # contributors y: 8 # changes python/src/passes.h x: 1 # contributors y: 1 # changes python/triton/runtime/jit.py x: 32 # contributors y: 65 # changes python/triton/language/semantic.py x: 36 # contributors y: 87 # changes lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 9 # contributors y: 15 # changes lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 4 # contributors y: 10 # changes python/triton/common/backend.py x: 9 # contributors y: 11 # changes lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 7 # contributors y: 9 # changes include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 5 # contributors y: 9 # changes lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 11 # contributors y: 24 # changes lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 3 # contributors y: 6 # changes lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 9 # contributors y: 21 # changes lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 3 # contributors y: 4 # changes lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 5 # contributors y: 6 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 3 # contributors y: 3 # changes lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 9 # contributors y: 17 # changes python/triton/ops/matmul_perf_model.py x: 9 # contributors y: 16 # changes python/triton/runtime/backends/hip.c x: 2 # contributors y: 3 # changes python/triton/runtime/driver.py x: 5 # contributors y: 12 # changes python/triton/language/core.py x: 44 # contributors y: 121 # changes python/triton/runtime/interpreter.py x: 4 # contributors y: 6 # changes lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 19 # contributors y: 55 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 16 # contributors y: 40 # changes lib/Dialect/Triton/IR/Ops.cpp x: 12 # contributors y: 30 # changes lib/Analysis/Utility.cpp x: 14 # contributors y: 47 # changes python/triton/language/math.py x: 7 # contributors y: 12 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 4 # contributors y: 5 # changes python/triton/language/standard.py x: 8 # contributors y: 9 # changes include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 2 # contributors y: 2 # changes lib/Analysis/Membar.cpp x: 8 # contributors y: 17 # changes lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 15 # contributors y: 38 # changes lib/Dialect/TritonGPU/IR/Dialect.cpp x: 19 # contributors y: 53 # changes lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 4 # contributors y: 7 # changes lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 5 # contributors y: 5 # changes lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 4 # contributors y: 4 # changes lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 9 # contributors y: 8 # changes python/triton/ops/matmul.py x: 18 # contributors y: 37 # changes lib/Analysis/Allocation.cpp x: 14 # contributors y: 24 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 14 # contributors y: 40 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 6 # contributors y: 11 # changes lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 10 # contributors y: 21 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 7 # contributors y: 10 # changes lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 7 # contributors y: 16 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 17 # contributors y: 39 # changes lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 9 # contributors y: 19 # changes lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 13 # contributors y: 41 # changes lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 6 # contributors y: 7 # changes unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 2 # contributors y: 5 # changes lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 17 # contributors y: 38 # changes python/triton/language/__init__.py x: 15 # contributors y: 35 # changes lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 14 # contributors y: 30 # changes python/triton/runtime/backends/cuda.c x: 8 # contributors y: 10 # changes python/triton/ops/__init__.py x: 5 # contributors y: 10 # changes lib/Analysis/AxisInfo.cpp x: 10 # contributors y: 26 # changes python/triton/runtime/autotuner.py x: 15 # contributors y: 26 # changes python/triton/__init__.py x: 11 # contributors y: 38 # changes python/triton/compiler/utils.py x: 6 # contributors y: 6 # changes lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 4 # contributors y: 9 # changes lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 11 # contributors y: 26 # changes lib/Conversion/TritonGPUToLLVM/Utility.h x: 8 # contributors y: 21 # changes include/triton/Dialect/TritonGPU/IR/Dialect.h x: 11 # contributors y: 21 # changes lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 7 # contributors y: 11 # changes include/triton/Analysis/Membar.h x: 3 # contributors y: 5 # changes lib/Dialect/Triton/IR/Traits.cpp x: 5 # contributors y: 7 # changes python/triton/language/random.py x: 9 # contributors y: 25 # changes include/triton/Analysis/AxisInfo.h x: 6 # contributors y: 10 # changes python/tutorials/06-fused-attention.py x: 14 # contributors y: 36 # changes lib/Dialect/Triton/Transforms/RewriteTensorPointer.cpp x: 7 # contributors y: 7 # changes python/triton/compiler/make_launcher.py x: 9 # contributors y: 13 # changes python/triton/ops/blocksparse/matmul.py x: 15 # contributors y: 29 # changes python/triton/ops/blocksparse/softmax.py x: 10 # contributors y: 23 # changes python/triton/ops/cross_entropy.py x: 10 # contributors y: 15 # changes python/triton/ops/flash_attention.py x: 12 # contributors y: 18 # changes python/triton/tools/build_extern.py x: 8 # contributors y: 15 # changes python/tutorials/02-fused-softmax.py x: 12 # contributors y: 26 # changes python/tutorials/03-matrix-multiplication.py x: 16 # contributors y: 36 # changes include/triton/Dialect/Triton/IR/Dialect.h x: 5 # contributors y: 8 # changes include/triton/Analysis/Utility.h x: 8 # contributors y: 25 # changes python/triton/tools/disasm.py x: 6 # contributors y: 8 # changes python/triton/ops/blocksparse/__init__.py x: 2 # contributors y: 6 # changes
121.0
# changes
  min: 1.0
  average: 11.53
  25th percentile: 2.0
  median: 6.0
  75th percentile: 14.0
  max: 121.0
0 44.0
# contributors
min: 1.0 | average: 5.96 | 25th percentile: 2.0 | median: 4.0 | 75th percentile: 8.0 | max: 44.0

Number of Contributors vs. File Size: 201 points

python/tutorials/04-low-memory-dropout.py x: 7 # contributors y: 73 lines of code python/src/ir.cc x: 1 # contributors y: 1443 lines of code python/src/llvm.cc x: 1 # contributors y: 172 lines of code python/src/nvidia.cc x: 1 # contributors y: 120 lines of code python/triton/compiler/backends/cuda.py x: 3 # contributors y: 248 lines of code python/triton/compiler/code_generator.py x: 14 # contributors y: 963 lines of code python/triton/compiler/compiler.py x: 25 # contributors y: 198 lines of code python/tutorials/01-vector-add.py x: 13 # contributors y: 59 lines of code python/tutorials/05-layer-norm.py x: 10 # contributors y: 237 lines of code python/src/main.cc x: 3 # contributors y: 17 lines of code python/src/passes.cc x: 1 # contributors y: 62 lines of code python/src/passes.h x: 1 # contributors y: 16 lines of code python/triton/runtime/jit.py x: 32 # contributors y: 429 lines of code python/triton/language/semantic.py x: 36 # contributors y: 1095 lines of code lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 9 # contributors y: 242 lines of code lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 4 # contributors y: 127 lines of code python/triton/common/backend.py x: 9 # contributors y: 115 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 7 # contributors y: 645 lines of code include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 5 # contributors y: 27 lines of code lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 11 # contributors y: 301 lines of code lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 3 # contributors y: 65 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 9 # contributors y: 250 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 3 # contributors y: 82 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 5 # contributors y: 375 lines of code lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 9 # contributors y: 280 lines of code lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 14 # contributors y: 853 lines of code python/triton/ops/matmul_perf_model.py x: 9 # contributors y: 117 lines of code python/triton/compiler/backends/__init__.py x: 1 # contributors y: 3 lines of code python/triton/runtime/backends/hip.c x: 2 # contributors y: 92 lines of code python/triton/runtime/driver.py x: 5 # contributors y: 190 lines of code python/src/interpreter.cc x: 1 # contributors y: 40 lines of code python/triton/language/core.py x: 44 # contributors y: 1185 lines of code python/triton/runtime/interpreter.py x: 4 # contributors y: 371 lines of code lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 19 # contributors y: 1465 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 16 # contributors y: 748 lines of code lib/Dialect/Triton/IR/Ops.cpp x: 12 # contributors y: 843 lines of code lib/Analysis/Utility.cpp x: 14 # contributors y: 609 lines of code python/triton/language/math.py x: 7 # contributors y: 1292 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 4 # contributors y: 555 lines of code python/triton/language/standard.py x: 8 # contributors y: 240 lines of code include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 2 # contributors y: 53 lines of code lib/Analysis/Alias.cpp x: 5 # contributors y: 55 lines of code lib/Analysis/Membar.cpp x: 8 # contributors y: 157 lines of code lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 15 # contributors y: 1500 lines of code lib/Dialect/TritonGPU/IR/Dialect.cpp x: 19 # contributors y: 1632 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 4 # contributors y: 691 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 5 # contributors y: 225 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 4 # contributors y: 396 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 9 # contributors y: 555 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSMutex.cpp x: 5 # contributors y: 277 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 7 # contributors y: 811 lines of code python/triton/ops/matmul.py x: 18 # contributors y: 174 lines of code lib/Analysis/Allocation.cpp x: 14 # contributors y: 467 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 14 # contributors y: 915 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 6 # contributors y: 369 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 10 # contributors y: 455 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.cpp x: 9 # contributors y: 165 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/FMA.cpp x: 4 # contributors y: 83 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 5 # contributors y: 124 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 7 # contributors y: 350 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 7 # contributors y: 481 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 17 # contributors y: 991 lines of code lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 9 # contributors y: 138 lines of code lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 13 # contributors y: 599 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 6 # contributors y: 96 lines of code unittest/Conversion/TritonGPUToLLVM/EmitIndicesTest.cpp x: 4 # contributors y: 509 lines of code unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 2 # contributors y: 45 lines of code lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 17 # contributors y: 608 lines of code python/triton/language/__init__.py x: 15 # contributors y: 217 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 7 # contributors y: 821 lines of code include/triton/Dialect/TritonGPU/IR/Traits.h x: 2 # contributors y: 30 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 14 # contributors y: 695 lines of code lib/Dialect/TritonGPU/IR/Traits.cpp x: 4 # contributors y: 18 lines of code python/triton/runtime/cache.py x: 8 # contributors y: 116 lines of code python/triton/runtime/backends/cuda.c x: 8 # contributors y: 458 lines of code python/triton/compiler/__init__.py x: 4 # contributors y: 3 lines of code python/triton/ops/__init__.py x: 5 # contributors y: 5 lines of code python/triton/tools/compile.py x: 5 # contributors y: 104 lines of code lib/Analysis/AxisInfo.cpp x: 10 # contributors y: 905 lines of code python/triton/runtime/autotuner.py x: 15 # contributors y: 207 lines of code python/triton/__init__.py x: 11 # contributors y: 51 lines of code python/triton/compiler/utils.py x: 6 # contributors y: 210 lines of code python/triton/language/extra/cuda.py x: 3 # contributors y: 12 lines of code lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 4 # contributors y: 982 lines of code lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 3 # contributors y: 200 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 6 # contributors y: 195 lines of code lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 11 # contributors y: 425 lines of code lib/Conversion/TritonGPUToLLVM/Utility.cpp x: 9 # contributors y: 315 lines of code lib/Conversion/TritonGPUToLLVM/Utility.h x: 8 # contributors y: 272 lines of code lib/Dialect/NVGPU/IR/Dialect.cpp x: 2 # contributors y: 71 lines of code lib/Dialect/Triton/IR/Dialect.cpp x: 7 # contributors y: 63 lines of code unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 4 # contributors y: 296 lines of code include/triton/Dialect/TritonGPU/IR/Dialect.h x: 11 # contributors y: 55 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSFeasibilityChecking.cpp x: 2 # contributors y: 39 lines of code lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 7 # contributors y: 83 lines of code include/triton/Dialect/Triton/IR/Traits.h x: 3 # contributors y: 89 lines of code lib/Dialect/Triton/IR/Traits.cpp x: 5 # contributors y: 211 lines of code lib/Dialect/TritonNvidiaGPU/IR/Traits.cpp x: 2 # contributors y: 10 lines of code include/triton/Dialect/TritonGPU/IR/Attributes.h x: 2 # contributors y: 6 lines of code python/triton/language/random.py x: 9 # contributors y: 95 lines of code include/triton/Analysis/AxisInfo.h x: 6 # contributors y: 247 lines of code python/tutorials/06-fused-attention.py x: 14 # contributors y: 481 lines of code .pre-commit-config.yaml x: 4 # contributors y: 52 lines of code lib/Dialect/Triton/Transforms/RewriteTensorPointer.cpp x: 7 # contributors y: 364 lines of code .github/workflows/torch-inductor/scripts/check_perf.py x: 3 # contributors y: 57 lines of code python/triton/compiler/make_launcher.py x: 9 # contributors y: 85 lines of code python/triton/ops/blocksparse/matmul.py x: 15 # contributors y: 306 lines of code python/triton/ops/blocksparse/softmax.py x: 10 # contributors y: 183 lines of code python/triton/ops/cross_entropy.py x: 10 # contributors y: 62 lines of code python/triton/ops/flash_attention.py x: 12 # contributors y: 381 lines of code python/triton/tools/build_extern.py x: 8 # contributors y: 307 lines of code python/triton/tools/link.py x: 5 # contributors y: 235 lines of code python/tutorials/02-fused-softmax.py x: 12 # contributors y: 81 lines of code python/tutorials/03-matrix-multiplication.py x: 16 # contributors y: 119 lines of code python/tutorials/07-math-functions.py x: 2 # contributors y: 37 lines of code python/tutorials/08-experimental-block-pointer.py x: 2 # contributors y: 85 lines of code python/tutorials/09-experimental-tma-matrix-multiplication.py x: 5 # contributors y: 138 lines of code python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 4 # contributors y: 104 lines of code include/triton/Dialect/TritonGPU/Transforms/Utility.h x: 6 # contributors y: 72 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.h x: 1 # contributors y: 37 lines of code .github/workflows/torch-inductor/scripts/check_acc.py x: 1 # contributors y: 10 lines of code include/triton/Analysis/Alias.h x: 3 # contributors y: 48 lines of code include/triton/Dialect/Triton/IR/Dialect.h x: 5 # contributors y: 45 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSDecomposing.cpp x: 2 # contributors y: 179 lines of code include/triton/Analysis/Utility.h x: 8 # contributors y: 249 lines of code lib/Dialect/Triton/Transforms/Combine.cpp x: 7 # contributors y: 178 lines of code python/triton/tools/disasm.py x: 6 # contributors y: 89 lines of code include/triton/Dialect/TritonNvidiaGPU/IR/Dialect.h x: 2 # contributors y: 21 lines of code lib/Target/LLVMIR/LLVMIRBreakPhiStruct.cpp x: 1 # contributors y: 50 lines of code include/triton/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.h x: 5 # contributors y: 20 lines of code lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.h x: 5 # contributors y: 12 lines of code include/triton/Conversion/TritonGPUToLLVM/PTXAsmFormat.h x: 4 # contributors y: 180 lines of code lib/Conversion/TritonGPUToLLVM/PTXAsmFormat.cpp x: 4 # contributors y: 192 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSFixupMissingAttrs.cpp x: 1 # contributors y: 54 lines of code include/triton/Dialect/Triton/IR/Types.h x: 4 # contributors y: 21 lines of code include/triton/Dialect/TritonGPU/Transforms/TritonGPUConversion.h x: 3 # contributors y: 24 lines of code include/triton/Dialect/TritonNvidiaGPU/IR/Traits.h x: 1 # contributors y: 21 lines of code lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.h x: 4 # contributors y: 12 lines of code lib/Dialect/Triton/IR/Types.cpp x: 4 # contributors y: 93 lines of code lib/Dialect/TritonGPU/IR/Types.cpp x: 1 # contributors y: 27 lines of code lib/Target/LLVMIR/LLVMDIScope.cpp x: 1 # contributors y: 115 lines of code include/triton/Conversion/TritonGPUToLLVM/GCNAsmFormat.h x: 1 # contributors y: 241 lines of code lib/Conversion/TritonGPUToLLVM/GCNAsmFormat.cpp x: 1 # contributors y: 161 lines of code unittest/Conversion/TritonGPUToLLVM/PTXAsmFormatTest.cpp x: 2 # contributors y: 110 lines of code
1632.0
lines of code
  min: 1.0
  average: 220.81
  25th percentile: 21.0
  median: 89.0
  75th percentile: 249.5
  max: 1632.0
0 44.0
# contributors
min: 1.0 | average: 5.96 | 25th percentile: 2.0 | median: 4.0 | 75th percentile: 8.0 | max: 44.0

Number of Changes vs. Line Churn: 844 points

python/tutorials/04-low-memory-dropout.py x: 8 # changes y: 221 line churn python/src/ir.cc x: 3 # changes y: 1643 line churn python/src/llvm.cc x: 2 # changes y: 444 line churn python/src/nvidia.cc x: 2 # changes y: 144 line churn python/triton/compiler/backends/cuda.py x: 9 # changes y: 550 line churn python/triton/compiler/code_generator.py x: 42 # changes y: 1824 line churn python/triton/compiler/compiler.py x: 45 # changes y: 2421 line churn python/tutorials/01-vector-add.py x: 31 # changes y: 948 line churn python/tutorials/05-layer-norm.py x: 17 # changes y: 1247 line churn CMakeLists.txt x: 53 # changes y: 783 line churn lib/Target/LLVMIR/CMakeLists.txt x: 11 # changes y: 43 line churn lib/Target/LLVMIR/LLVMIRTranslation.cpp x: 31 # changes y: 1610 line churn python/src/main.cc x: 8 # changes y: 51 line churn python/src/passes.h x: 1 # changes y: 20 line churn python/src/translation.cc x: 3 # changes y: 482 line churn python/triton/runtime/jit.py x: 65 # changes y: 2349 line churn python/test/unit/language/test_core.py x: 193 # changes y: 10567 line churn python/triton/language/semantic.py x: 87 # changes y: 6072 line churn lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 15 # changes y: 873 line churn lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 10 # changes y: 181 line churn python/triton/common/backend.py x: 11 # changes y: 222 line churn test/TritonGPU/reorder-instructions.mlir x: 5 # changes y: 104 line churn lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 9 # changes y: 1039 line churn test/TritonGPU/rewrite-tensor-pointer.mlir x: 5 # changes y: 148 line churn python/test/unit/language/test_block_pointer.py x: 4 # changes y: 178 line churn include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 9 # changes y: 85 line churn include/triton/Dialect/TritonGPU/Transforms/Passes.td x: 11 # changes y: 278 line churn lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 24 # changes y: 926 line churn lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 6 # changes y: 82 line churn lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 21 # changes y: 1092 line churn lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 6 # changes y: 442 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 3 # changes y: 115 line churn lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 17 # changes y: 513 line churn lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 42 # changes y: 3200 line churn python/src/runtime.cc x: 2 # changes y: 48 line churn python/src/triton.cc x: 132 # changes y: 10105 line churn python/triton/ops/matmul_perf_model.py x: 16 # changes y: 436 line churn python/triton/testing.py x: 59 # changes y: 1540 line churn bin/CMakeLists.txt x: 12 # changes y: 219 line churn bin/triton-translate.cpp x: 12 # changes y: 366 line churn lib/Target/CMakeLists.txt x: 4 # changes y: 5 line churn lib/Target/PTX/PTXTranslation.cpp x: 17 # changes y: 424 line churn python/triton/runtime/backends/hip.c x: 3 # changes y: 148 line churn python/triton/runtime/driver.py x: 12 # changes y: 343 line churn python/triton/language/core.py x: 121 # changes y: 6324 line churn python/triton/runtime/interpreter.py x: 6 # changes y: 622 line churn python/test/unit/language/test_conversions.py x: 1 # changes y: 305 line churn include/triton/Dialect/Triton/IR/TritonAttrDefs.td x: 9 # changes y: 153 line churn include/triton/Dialect/Triton/IR/TritonOps.td x: 40 # changes y: 1166 line churn lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 55 # changes y: 6903 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 40 # changes y: 1733 line churn lib/Dialect/Triton/IR/Ops.cpp x: 30 # changes y: 1436 line churn test/Conversion/tritongpu_to_llvm_hopper.mlir x: 15 # changes y: 364 line churn lib/Analysis/Utility.cpp x: 47 # changes y: 1401 line churn python/triton/language/math.py x: 12 # changes y: 4649 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 5 # changes y: 863 line churn include/triton/Dialect/TritonGPU/IR/TritonGPUOps.td x: 16 # changes y: 602 line churn include/triton/Dialect/TritonNvidiaGPU/IR/TritonNvidiaGPUOps.td x: 9 # changes y: 432 line churn lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 38 # changes y: 2688 line churn lib/Dialect/TritonGPU/IR/Dialect.cpp x: 53 # changes y: 3623 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 7 # changes y: 1082 line churn lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 5 # changes y: 348 line churn lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 4 # changes y: 561 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 8 # changes y: 980 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 12 # changes y: 1258 line churn test/TritonGPU/fence-inserstion.mlir x: 6 # changes y: 285 line churn test/TritonGPU/loop-pipeline-hopper.mlir x: 6 # changes y: 546 line churn test/TritonGPU/wsmaterialization.mlir x: 5 # changes y: 1088 line churn python/test/unit/operators/test_matmul.py x: 28 # changes y: 913 line churn python/triton/ops/matmul.py x: 37 # changes y: 1328 line churn include/triton/Dialect/TritonGPU/IR/TritonGPUAttrDefs.td x: 22 # changes y: 1229 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 40 # changes y: 2112 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 11 # changes y: 568 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 21 # changes y: 1770 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.cpp x: 16 # changes y: 881 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 6 # changes y: 179 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 10 # changes y: 556 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 39 # changes y: 2461 line churn lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 19 # changes y: 428 line churn lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 41 # changes y: 1838 line churn lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 7 # changes y: 164 line churn test/Analysis/test-allocation.mlir x: 18 # changes y: 847 line churn test/Analysis/test-membar.mlir x: 16 # changes y: 1050 line churn test/Conversion/tritongpu_to_llvm.mlir x: 37 # changes y: 2519 line churn test/TritonGPU/combine.mlir x: 42 # changes y: 3303 line churn test/TritonGPU/dot-operands.mlir x: 14 # changes y: 504 line churn test/TritonGPU/loop-pipeline.mlir x: 27 # changes y: 1443 line churn test/TritonGPU/optimize-locality.mlir x: 4 # changes y: 623 line churn test/TritonGPU/wsdecomposing.mlir x: 4 # changes y: 796 line churn test/TritonNvidiaGPU/ws-feasibility-checking.mlir x: 6 # changes y: 1121 line churn lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 38 # changes y: 1908 line churn python/triton/language/__init__.py x: 35 # changes y: 318 line churn lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 8 # changes y: 1088 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 30 # changes y: 1506 line churn python/test/unit/hopper/test_persistent_warp_specialized_gemm.py x: 15 # changes y: 1791 line churn python/triton/runtime/backends/cuda.c x: 10 # changes y: 663 line churn python/triton/ops/__init__.py x: 10 # changes y: 42 line churn docs/python-api/triton.language.rst x: 17 # changes y: 269 line churn lib/Analysis/AxisInfo.cpp x: 26 # changes y: 1785 line churn test/Analysis/test-alignment.mlir x: 15 # changes y: 1213 line churn .github/workflows/llvm-build.yml x: 10 # changes y: 964 line churn python/triton/runtime/autotuner.py x: 26 # changes y: 652 line churn .github/workflows/wheels.yml x: 16 # changes y: 198 line churn python/setup.py x: 118 # changes y: 1467 line churn .github/workflows/integration-tests.yml x: 54 # changes y: 1186 line churn python/test/unit/tools/test_aot.py x: 10 # changes y: 726 line churn python/triton/__init__.py x: 38 # changes y: 195 line churn include/triton/Dialect/NVGPU/IR/NVGPUOps.td x: 7 # changes y: 586 line churn lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 9 # changes y: 3021 line churn lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 3 # changes y: 267 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 7 # changes y: 318 line churn lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 26 # changes y: 2213 line churn lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.cpp x: 9 # changes y: 662 line churn lib/Conversion/TritonGPUToLLVM/Utility.h x: 21 # changes y: 683 line churn unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 4 # changes y: 401 line churn include/triton/Dialect/TritonGPU/IR/Dialect.h x: 21 # changes y: 179 line churn .github/workflows/documentation.yml x: 15 # changes y: 244 line churn README.md x: 46 # changes y: 480 line churn test/Triton/combine.mlir x: 10 # changes y: 351 line churn python/test/unit/language/test_random.py x: 13 # changes y: 461 line churn python/triton/language/random.py x: 25 # changes y: 885 line churn python/tutorials/06-fused-attention.py x: 36 # changes y: 2436 line churn python/test/unit/operators/test_flash_attention.py x: 14 # changes y: 246 line churn python/test/unit/runtime/test_cache.py x: 27 # changes y: 765 line churn python/test/regression/test_performance.py x: 36 # changes y: 1247 line churn docs/conf.py x: 18 # changes y: 540 line churn python/test/unit/hopper/test_flashattention.py x: 3 # changes y: 729 line churn python/test/unit/hopper/test_gemm.py x: 6 # changes y: 860 line churn python/test/unit/hopper/test_gemm_fusion.py x: 2 # changes y: 263 line churn python/test/unit/operators/test_blocksparse.py x: 15 # changes y: 624 line churn python/triton/compiler/make_launcher.py x: 13 # changes y: 683 line churn python/triton/ops/blocksparse/matmul.py x: 29 # changes y: 3472 line churn python/triton/ops/blocksparse/softmax.py x: 23 # changes y: 1266 line churn python/triton/ops/flash_attention.py x: 18 # changes y: 1336 line churn python/triton/tools/build_extern.py x: 15 # changes y: 728 line churn python/tutorials/02-fused-softmax.py x: 26 # changes y: 894 line churn python/tutorials/03-matrix-multiplication.py x: 36 # changes y: 1955 line churn python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 4 # changes y: 265 line churn lib/Dialect/TritonGPU/Transforms/Pipeline.cpp x: 39 # changes y: 5622 line churn lib/Conversion/TritonGPUToLLVM/CMakeLists.txt x: 13 # changes y: 120 line churn include/triton/Analysis/Utility.h x: 25 # changes y: 446 line churn bin/RegisterTritonDialects.h x: 7 # changes y: 57 line churn python/triton/interpreter/tl_lang.py x: 6 # changes y: 1284 line churn lib/Dialect/Triton/Transforms/Combine.cpp x: 11 # changes y: 443 line churn include/triton/Analysis/Allocation.h x: 9 # changes y: 322 line churn lib/Dialect/NVGPU/ToLLVMIR/NVGPUToLLVMIR.cpp x: 2 # changes y: 1610 line churn lib/Hopper/HopperHelpers.c x: 3 # changes y: 1024 line churn python/triton/third_party/cuda/include/cuda.h x: 2 # changes y: 23373 line churn lib/Target/LLVMIR/LLVMDIScope.cpp x: 1 # changes y: 148 line churn docs/getting-started/installation.rst x: 15 # changes y: 179 line churn python/triton/tools/aot.py x: 7 # changes y: 444 line churn python/triton/debugger/tl_lang.py x: 4 # changes y: 1260 line churn python/triton/language/extern.py x: 8 # changes y: 276 line churn python/triton/compiler.py x: 51 # changes y: 7138 line churn lib/Conversion/TritonGPUToLLVM/DotOpHelpers.cpp x: 7 # changes y: 3160 line churn lib/Conversion/TritonGPUToLLVM/DotOpHelpers.h x: 15 # changes y: 4372 line churn python/triton/language/mathlib.py x: 1 # changes y: 3052 line churn python/triton/language/libdevice.py x: 7 # changes y: 5370 line churn bin/FileCheck/FileCheck.cpp x: 3 # changes y: 1770 line churn lib/Dialect/TritonGPU/Transforms/Combine.cpp x: 18 # changes y: 3734 line churn include/triton/codegen/analysis/layout.h x: 19 # changes y: 668 line churn include/triton/codegen/selection/generator.h x: 27 # changes y: 531 line churn include/triton/driver/dispatch.h x: 8 # changes y: 691 line churn include/triton/external/CUDA/cuda.h x: 5 # changes y: 30556 line churn include/triton/external/CUDA/nvml.h x: 2 # changes y: 6327 line churn include/triton/external/half.hpp x: 2 # changes y: 3085 line churn include/triton/external/hip.h x: 2 # changes y: 576 line churn include/triton/ir/enums.h x: 14 # changes y: 235 line churn include/triton/ir/instructions.h x: 37 # changes y: 1887 line churn include/triton/ir/type.h x: 15 # changes y: 495 line churn include/triton/ir/visitor.h x: 16 # changes y: 256 line churn include/triton/tools/sha1.hpp x: 3 # changes y: 378 line churn lib/codegen/selection/generator.cc x: 87 # changes y: 12742 line churn lib/codegen/transform/peephole.cc x: 20 # changes y: 594 line churn lib/driver/dispatch.cc x: 13 # changes y: 779 line churn lib/ir/builder.cc x: 29 # changes y: 1075 line churn lib/ir/instructions.cc x: 33 # changes y: 1730 line churn lib/ir/print.cc x: 6 # changes y: 829 line churn python/src/functions.h x: 8 # changes y: 1408 line churn python/src/pybind11/attr.h x: 2 # changes y: 2381 line churn python/src/pybind11/buffer_info.h x: 2 # changes y: 736 line churn python/src/pybind11/cast.h x: 2 # changes y: 7088 line churn python/src/pybind11/chrono.h x: 2 # changes y: 846 line churn python/src/pybind11/detail/class.h x: 2 # changes y: 2583 line churn python/src/pybind11/detail/common.h x: 2 # changes y: 4377 line churn python/src/pybind11/detail/init.h x: 2 # changes y: 1537 line churn python/src/pybind11/detail/internals.h x: 4 # changes y: 2131 line churn python/src/pybind11/detail/type_caster_base.h x: 2 # changes y: 4040 line churn python/src/pybind11/embed.h x: 2 # changes y: 994 line churn python/src/pybind11/eval.h x: 2 # changes y: 603 line churn python/src/pybind11/numpy.h x: 2 # changes y: 7242 line churn python/src/pybind11/pybind11.h x: 2 # changes y: 10884 line churn python/src/pybind11/pytypes.h x: 2 # changes y: 8807 line churn python/src/pybind11/stl_bind.h x: 2 # changes y: 2944 line churn python/triton/code_gen.py x: 79 # changes y: 6706 line churn include/triton/ir/dispatch.h x: 20 # changes y: 476 line churn lib/ir/dispatch.cc x: 33 # changes y: 3988 line churn include/triton/external/CL/cl.h x: 1 # changes y: 1468 line churn include/triton/external/CL/cl.hpp x: 1 # changes y: 12947 line churn include/triton/external/CL/cl2.hpp x: 1 # changes y: 9677 line churn include/triton/external/CL/cl_ext.h x: 1 # changes y: 670 line churn include/triton/external/CL/cl_ext_intel.h x: 1 # changes y: 429 line churn lib/driver/module.cc x: 31 # changes y: 1178 line churn python/test/test_blocksparse.py x: 8 # changes y: 502 line churn python/triton/language.py x: 14 # changes y: 1838 line churn python/triton/core.py x: 3 # changes y: 1306 line churn include/triton/lang/token.h x: 5 # changes y: 439 line churn include/triton/runtime/function.h x: 13 # changes y: 574 line churn lib/lang/ast.cc x: 9 # changes y: 1167 line churn lib/lang/cpp.cc x: 1 # changes y: 884 line churn lib/lang/parser.cc x: 10 # changes y: 2857 line churn lib/runtime/function.cc x: 41 # changes y: 2454 line churn python/triton/kernel.py x: 33 # changes y: 1383 line churn python/src/bindings.cc x: 20 # changes y: 1325 line churn python/examples/blocksparse.py x: 1 # changes y: 728 line churn docs/tutorials/pytriton.md x: 1 # changes y: 486 line churn
30556.0
line churn
  min: 1.0
  average: 602.51
  25th percentile: 44.0
  median: 139.0
  75th percentile: 443.75
  max: 30556.0
0 193.0
# changes
min: 1.0 | average: 7.6 | 25th percentile: 2.0 | median: 3.0 | 75th percentile: 7.0 | max: 193.0

Number of Contributors vs. Line Churn: 200 points

python/tutorials/04-low-memory-dropout.py x: 7 # contributors y: 221 line churn python/src/ir.cc x: 1 # contributors y: 1643 line churn python/src/llvm.cc x: 1 # contributors y: 444 line churn python/src/nvidia.cc x: 1 # contributors y: 144 line churn python/triton/compiler/backends/cuda.py x: 3 # contributors y: 550 line churn python/triton/compiler/code_generator.py x: 14 # contributors y: 1824 line churn python/triton/compiler/compiler.py x: 25 # contributors y: 2421 line churn python/tutorials/01-vector-add.py x: 13 # contributors y: 948 line churn python/tutorials/05-layer-norm.py x: 10 # contributors y: 1247 line churn python/src/main.cc x: 3 # contributors y: 51 line churn python/src/passes.cc x: 1 # contributors y: 181 line churn python/src/passes.h x: 1 # contributors y: 20 line churn python/triton/runtime/jit.py x: 32 # contributors y: 2349 line churn python/triton/language/semantic.py x: 36 # contributors y: 6072 line churn lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 9 # contributors y: 873 line churn lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 4 # contributors y: 181 line churn python/triton/common/backend.py x: 9 # contributors y: 222 line churn lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 7 # contributors y: 1039 line churn include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 5 # contributors y: 85 line churn lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 11 # contributors y: 926 line churn lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 3 # contributors y: 82 line churn lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 9 # contributors y: 1092 line churn lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 3 # contributors y: 145 line churn lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 5 # contributors y: 442 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 3 # contributors y: 115 line churn lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 9 # contributors y: 513 line churn lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 14 # contributors y: 3200 line churn python/triton/ops/matmul_perf_model.py x: 9 # contributors y: 436 line churn python/triton/runtime/backends/hip.c x: 2 # contributors y: 148 line churn python/triton/runtime/driver.py x: 5 # contributors y: 343 line churn python/src/interpreter.cc x: 1 # contributors y: 44 line churn python/triton/language/core.py x: 44 # contributors y: 6324 line churn python/triton/runtime/interpreter.py x: 4 # contributors y: 622 line churn lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 19 # contributors y: 6903 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 16 # contributors y: 1733 line churn lib/Dialect/Triton/IR/Ops.cpp x: 12 # contributors y: 1436 line churn lib/Analysis/Utility.cpp x: 14 # contributors y: 1401 line churn python/triton/language/math.py x: 7 # contributors y: 4649 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 4 # contributors y: 863 line churn python/triton/language/standard.py x: 8 # contributors y: 522 line churn include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 2 # contributors y: 97 line churn lib/Analysis/Alias.cpp x: 5 # contributors y: 109 line churn lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 15 # contributors y: 2688 line churn lib/Dialect/TritonGPU/IR/Dialect.cpp x: 19 # contributors y: 3623 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 4 # contributors y: 1082 line churn lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 4 # contributors y: 561 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 9 # contributors y: 980 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSMutex.cpp x: 5 # contributors y: 476 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 7 # contributors y: 1258 line churn python/triton/ops/matmul.py x: 18 # contributors y: 1328 line churn lib/Analysis/Allocation.cpp x: 14 # contributors y: 845 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 14 # contributors y: 2112 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 6 # contributors y: 568 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 10 # contributors y: 1770 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/FMA.cpp x: 4 # contributors y: 148 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 5 # contributors y: 179 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 7 # contributors y: 556 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 7 # contributors y: 847 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 17 # contributors y: 2461 line churn lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 9 # contributors y: 428 line churn lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 13 # contributors y: 1838 line churn lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 6 # contributors y: 164 line churn unittest/Conversion/TritonGPUToLLVM/EmitIndicesTest.cpp x: 4 # contributors y: 693 line churn unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 2 # contributors y: 74 line churn lib/Conversion/TritonGPUToLLVM/ViewOpToLLVM.cpp x: 9 # contributors y: 663 line churn lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 17 # contributors y: 1908 line churn python/triton/language/__init__.py x: 15 # contributors y: 318 line churn lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 7 # contributors y: 1088 line churn include/triton/Dialect/TritonGPU/IR/Traits.h x: 2 # contributors y: 42 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 14 # contributors y: 1506 line churn lib/Dialect/TritonGPU/IR/Traits.cpp x: 4 # contributors y: 29 line churn python/triton/runtime/cache.py x: 8 # contributors y: 205 line churn python/triton/runtime/backends/cuda.c x: 8 # contributors y: 663 line churn python/triton/ops/__init__.py x: 5 # contributors y: 42 line churn python/triton/tools/compile.py x: 5 # contributors y: 206 line churn python/triton/runtime/autotuner.py x: 15 # contributors y: 652 line churn python/triton/__init__.py x: 11 # contributors y: 195 line churn python/triton/compiler/utils.py x: 6 # contributors y: 358 line churn lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 4 # contributors y: 3021 line churn lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 3 # contributors y: 267 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 6 # contributors y: 318 line churn lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 11 # contributors y: 2213 line churn lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.cpp x: 5 # contributors y: 662 line churn lib/Conversion/TritonGPUToLLVM/Utility.cpp x: 9 # contributors y: 570 line churn lib/Conversion/TritonGPUToLLVM/Utility.h x: 8 # contributors y: 683 line churn lib/Dialect/NVGPU/IR/Dialect.cpp x: 2 # contributors y: 112 line churn lib/Dialect/Triton/IR/Dialect.cpp x: 7 # contributors y: 105 line churn unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 4 # contributors y: 401 line churn include/triton/Dialect/TritonGPU/IR/Dialect.h x: 11 # contributors y: 179 line churn lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 7 # contributors y: 206 line churn include/triton/Analysis/Membar.h x: 3 # contributors y: 326 line churn include/triton/Dialect/TritonGPU/IR/Attributes.h x: 2 # contributors y: 9 line churn include/triton/Analysis/AxisInfo.h x: 6 # contributors y: 482 line churn python/tutorials/06-fused-attention.py x: 14 # contributors y: 2436 line churn .pre-commit-config.yaml x: 4 # contributors y: 232 line churn python/triton/compiler/make_launcher.py x: 9 # contributors y: 683 line churn python/triton/ops/blocksparse/matmul.py x: 15 # contributors y: 3472 line churn python/triton/ops/blocksparse/softmax.py x: 10 # contributors y: 1266 line churn python/triton/ops/cross_entropy.py x: 10 # contributors y: 340 line churn python/triton/ops/flash_attention.py x: 12 # contributors y: 1336 line churn python/triton/tools/build_extern.py x: 8 # contributors y: 728 line churn python/triton/tools/link.py x: 5 # contributors y: 496 line churn python/tutorials/02-fused-softmax.py x: 12 # contributors y: 894 line churn python/tutorials/03-matrix-multiplication.py x: 16 # contributors y: 1955 line churn python/tutorials/08-experimental-block-pointer.py x: 2 # contributors y: 301 line churn python/tutorials/09-experimental-tma-matrix-multiplication.py x: 5 # contributors y: 268 line churn python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 4 # contributors y: 265 line churn python/tutorials/11-grouped-gemm.py x: 3 # contributors y: 355 line churn include/triton/Dialect/TritonGPU/Transforms/Utility.h x: 6 # contributors y: 228 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.h x: 1 # contributors y: 101 line churn include/triton/Dialect/Triton/IR/Dialect.h x: 5 # contributors y: 76 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSDecomposing.cpp x: 2 # contributors y: 264 line churn include/triton/Analysis/Utility.h x: 8 # contributors y: 446 line churn lib/Dialect/Triton/Transforms/ReorderBroadcast.cpp x: 3 # contributors y: 272 line churn lib/Dialect/Triton/Transforms/Combine.cpp x: 7 # contributors y: 443 line churn lib/Target/LLVMIR/LLVMIRBreakPhiStruct.cpp x: 1 # contributors y: 60 line churn include/triton/Analysis/Allocation.h x: 7 # contributors y: 322 line churn python/triton/tools/compile.c x: 3 # contributors y: 75 line churn include/triton/Conversion/TritonGPUToLLVM/PTXAsmFormat.h x: 4 # contributors y: 350 line churn include/triton/Dialect/Triton/Transforms/Passes.h x: 4 # contributors y: 24 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.h x: 3 # contributors y: 25 line churn python/MANIFEST.in x: 5 # contributors y: 6 line churn include/triton/Conversion/TritonGPUToLLVM/GCNAsmFormat.h x: 1 # contributors y: 381 line churn lib/Conversion/TritonGPUToLLVM/GCNAsmFormat.cpp x: 1 # contributors y: 191 line churn
6903.0
line churn
  min: 3.0
  average: 612.45
  25th percentile: 47.0
  median: 200.0
  75th percentile: 663.0
  max: 6903.0
0 44.0
# contributors
min: 1.0 | average: 5.98 | 25th percentile: 2.0 | median: 4.0 | 75th percentile: 8.0 | max: 44.0

File Size vs. Line Churn: 200 points

python/tutorials/04-low-memory-dropout.py x: 73 lines of code y: 221 line churn python/src/ir.cc x: 1443 lines of code y: 1643 line churn python/src/llvm.cc x: 172 lines of code y: 444 line churn python/src/nvidia.cc x: 120 lines of code y: 144 line churn python/triton/compiler/backends/cuda.py x: 248 lines of code y: 550 line churn python/triton/compiler/code_generator.py x: 963 lines of code y: 1824 line churn python/triton/compiler/compiler.py x: 198 lines of code y: 2421 line churn python/tutorials/01-vector-add.py x: 59 lines of code y: 948 line churn python/tutorials/05-layer-norm.py x: 237 lines of code y: 1247 line churn python/src/main.cc x: 17 lines of code y: 51 line churn python/src/passes.cc x: 62 lines of code y: 181 line churn python/src/passes.h x: 16 lines of code y: 20 line churn python/triton/runtime/jit.py x: 429 lines of code y: 2349 line churn python/triton/language/semantic.py x: 1095 lines of code y: 6072 line churn lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 242 lines of code y: 873 line churn lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 127 lines of code y: 181 line churn python/triton/common/backend.py x: 115 lines of code y: 222 line churn lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 645 lines of code y: 1039 line churn include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 27 lines of code y: 85 line churn lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 301 lines of code y: 926 line churn lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 65 lines of code y: 82 line churn lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 250 lines of code y: 1092 line churn lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 82 lines of code y: 145 line churn lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 375 lines of code y: 442 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 79 lines of code y: 115 line churn lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 280 lines of code y: 513 line churn lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 853 lines of code y: 3200 line churn python/triton/ops/matmul_perf_model.py x: 117 lines of code y: 436 line churn python/triton/compiler/backends/__init__.py x: 3 lines of code y: 5 line churn python/triton/runtime/backends/hip.c x: 92 lines of code y: 148 line churn python/triton/runtime/driver.py x: 190 lines of code y: 343 line churn python/src/interpreter.cc x: 40 lines of code y: 44 line churn python/triton/language/core.py x: 1185 lines of code y: 6324 line churn python/triton/runtime/interpreter.py x: 371 lines of code y: 622 line churn lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 1465 lines of code y: 6903 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 748 lines of code y: 1733 line churn lib/Dialect/Triton/IR/Ops.cpp x: 843 lines of code y: 1436 line churn lib/Analysis/Utility.cpp x: 609 lines of code y: 1401 line churn python/triton/language/math.py x: 1292 lines of code y: 4649 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 555 lines of code y: 863 line churn python/triton/language/standard.py x: 240 lines of code y: 522 line churn include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 53 lines of code y: 97 line churn lib/Analysis/Alias.cpp x: 55 lines of code y: 109 line churn lib/Analysis/Membar.cpp x: 157 lines of code y: 528 line churn lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 1500 lines of code y: 2688 line churn lib/Dialect/TritonGPU/IR/Dialect.cpp x: 1632 lines of code y: 3623 line churn lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 691 lines of code y: 1082 line churn lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 225 lines of code y: 348 line churn lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 396 lines of code y: 561 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 555 lines of code y: 980 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSMutex.cpp x: 277 lines of code y: 476 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 811 lines of code y: 1258 line churn python/triton/ops/matmul.py x: 174 lines of code y: 1328 line churn lib/Analysis/Allocation.cpp x: 467 lines of code y: 845 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 915 lines of code y: 2112 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 369 lines of code y: 568 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 455 lines of code y: 1770 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.cpp x: 165 lines of code y: 881 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/FMA.cpp x: 83 lines of code y: 148 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 124 lines of code y: 179 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 350 lines of code y: 556 line churn lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 481 lines of code y: 847 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 991 lines of code y: 2461 line churn lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 138 lines of code y: 428 line churn lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 599 lines of code y: 1838 line churn lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 96 lines of code y: 164 line churn unittest/Conversion/TritonGPUToLLVM/EmitIndicesTest.cpp x: 509 lines of code y: 693 line churn unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 45 lines of code y: 74 line churn lib/Conversion/TritonGPUToLLVM/ViewOpToLLVM.cpp x: 242 lines of code y: 663 line churn lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 608 lines of code y: 1908 line churn python/triton/language/__init__.py x: 217 lines of code y: 318 line churn lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 821 lines of code y: 1088 line churn include/triton/Dialect/TritonGPU/IR/Traits.h x: 30 lines of code y: 42 line churn lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 695 lines of code y: 1506 line churn lib/Dialect/TritonGPU/IR/Traits.cpp x: 18 lines of code y: 29 line churn python/triton/runtime/cache.py x: 116 lines of code y: 205 line churn python/triton/runtime/backends/cuda.c x: 458 lines of code y: 663 line churn python/triton/compiler/__init__.py x: 3 lines of code y: 32 line churn python/triton/tools/compile.py x: 104 lines of code y: 206 line churn lib/Analysis/AxisInfo.cpp x: 905 lines of code y: 1785 line churn python/triton/runtime/autotuner.py x: 207 lines of code y: 652 line churn python/triton/__init__.py x: 51 lines of code y: 195 line churn python/triton/compiler/utils.py x: 210 lines of code y: 358 line churn python/triton/language/extra/cuda.py x: 12 lines of code y: 52 line churn lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 982 lines of code y: 3021 line churn lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 200 lines of code y: 267 line churn lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 195 lines of code y: 318 line churn lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 425 lines of code y: 2213 line churn lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.cpp x: 373 lines of code y: 662 line churn lib/Conversion/TritonGPUToLLVM/Utility.cpp x: 315 lines of code y: 570 line churn lib/Conversion/TritonGPUToLLVM/Utility.h x: 272 lines of code y: 683 line churn lib/Dialect/NVGPU/IR/Dialect.cpp x: 71 lines of code y: 112 line churn lib/Dialect/Triton/IR/Dialect.cpp x: 63 lines of code y: 105 line churn unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 296 lines of code y: 401 line churn include/triton/Dialect/TritonGPU/IR/Dialect.h x: 55 lines of code y: 179 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSFeasibilityChecking.cpp x: 39 lines of code y: 71 line churn lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 83 lines of code y: 206 line churn include/triton/Analysis/Membar.h x: 82 lines of code y: 326 line churn include/triton/Dialect/Triton/IR/Traits.h x: 89 lines of code y: 130 line churn lib/Dialect/Triton/IR/Traits.cpp x: 211 lines of code y: 333 line churn lib/Dialect/TritonNvidiaGPU/IR/Traits.cpp x: 10 lines of code y: 38 line churn include/triton/Dialect/TritonGPU/IR/Attributes.h x: 6 lines of code y: 9 line churn python/triton/language/random.py x: 95 lines of code y: 885 line churn include/triton/Analysis/AxisInfo.h x: 247 lines of code y: 482 line churn python/tutorials/06-fused-attention.py x: 481 lines of code y: 2436 line churn .pre-commit-config.yaml x: 52 lines of code y: 232 line churn python/triton/common/build.py x: 109 lines of code y: 165 line churn lib/Dialect/Triton/Transforms/RewriteTensorPointer.cpp x: 364 lines of code y: 554 line churn lib/Dialect/TritonNvidiaGPU/IR/Ops.cpp x: 57 lines of code y: 95 line churn python/triton/compiler/make_launcher.py x: 85 lines of code y: 683 line churn python/triton/ops/blocksparse/matmul.py x: 306 lines of code y: 3472 line churn python/triton/ops/blocksparse/softmax.py x: 183 lines of code y: 1266 line churn python/triton/ops/cross_entropy.py x: 62 lines of code y: 340 line churn python/triton/ops/flash_attention.py x: 381 lines of code y: 1336 line churn python/triton/runtime/errors.py x: 10 lines of code y: 23 line churn python/triton/tools/build_extern.py x: 307 lines of code y: 728 line churn python/triton/tools/link.py x: 235 lines of code y: 496 line churn python/tutorials/02-fused-softmax.py x: 81 lines of code y: 894 line churn python/tutorials/03-matrix-multiplication.py x: 119 lines of code y: 1955 line churn python/tutorials/07-math-functions.py x: 37 lines of code y: 95 line churn python/tutorials/08-experimental-block-pointer.py x: 85 lines of code y: 301 line churn python/tutorials/09-experimental-tma-matrix-multiplication.py x: 138 lines of code y: 268 line churn python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 104 lines of code y: 265 line churn python/tutorials/11-grouped-gemm.py x: 204 lines of code y: 355 line churn include/triton/Analysis/Alias.h x: 48 lines of code y: 116 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSDecomposing.cpp x: 179 lines of code y: 264 line churn include/triton/Analysis/Utility.h x: 249 lines of code y: 446 line churn include/triton/Tools/Sys/GetEnv.hpp x: 32 lines of code y: 68 line churn lib/Dialect/Triton/Transforms/ReorderBroadcast.cpp x: 201 lines of code y: 272 line churn lib/Dialect/Triton/Transforms/Combine.cpp x: 178 lines of code y: 443 line churn python/triton/tools/disasm.py x: 89 lines of code y: 181 line churn include/triton/Dialect/TritonNvidiaGPU/IR/Dialect.h x: 21 lines of code y: 58 line churn lib/Target/LLVMIR/LLVMIRBreakPhiStruct.cpp x: 50 lines of code y: 60 line churn include/triton/Analysis/Allocation.h x: 177 lines of code y: 322 line churn lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.h x: 12 lines of code y: 60 line churn include/triton/Conversion/TritonGPUToLLVM/PTXAsmFormat.h x: 180 lines of code y: 350 line churn lib/Conversion/TritonGPUToLLVM/PTXAsmFormat.cpp x: 192 lines of code y: 271 line churn include/triton/Dialect/TritonNvidiaGPU/Transforms/Passes.h x: 41 lines of code y: 83 line churn lib/Dialect/TritonNvidiaGPU/Transforms/WSFixupMissingAttrs.cpp x: 54 lines of code y: 69 line churn include/triton/Dialect/Triton/IR/Types.h x: 21 lines of code y: 41 line churn include/triton/Dialect/Triton/Transforms/Passes.h x: 14 lines of code y: 24 line churn include/triton/Dialect/TritonGPU/Transforms/TritonGPUConversion.h x: 24 lines of code y: 44 line churn include/triton/Dialect/TritonNvidiaGPU/IR/Types.h x: 7 lines of code y: 33 line churn include/triton/Target/PTX/TmaMetadata.h x: 55 lines of code y: 107 line churn lib/Conversion/TritonGPUToLLVM/ClusterOpsToLLVM.cpp x: 34 lines of code y: 62 line churn lib/Conversion/TritonGPUToLLVM/RegReallocOpToLLVM.cpp x: 37 lines of code y: 43 line churn lib/Dialect/TritonGPU/IR/Types.cpp x: 27 lines of code y: 38 line churn lib/Target/LLVMIR/LLVMDIScope.cpp x: 115 lines of code y: 148 line churn python/triton/common/__init__.py x: 2 lines of code y: 7 line churn include/triton/Conversion/TritonGPUToLLVM/GCNAsmFormat.h x: 241 lines of code y: 381 line churn lib/Conversion/TritonGPUToLLVM/GCNAsmFormat.cpp x: 161 lines of code y: 191 line churn lib/Conversion/TritonGPUToLLVM/TypeConverter.h x: 22 lines of code y: 333 line churn unittest/Conversion/TritonGPUToLLVM/PTXAsmFormatTest.cpp x: 110 lines of code y: 151 line churn include/triton/Conversion/MLIRTypes.h x: 28 lines of code y: 69 line churn
6903.0
line churn
  min: 3.0
  average: 612.45
  25th percentile: 47.0
  median: 200.0
  75th percentile: 663.0
  max: 6903.0
0 1632.0
lines of code
min: 2.0 | average: 221.91 | 25th percentile: 21.0 | median: 89.0 | 75th percentile: 249.75 | max: 1632.0