anthropics / triton
File Size

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

Intro
  • File size measurements show the distribution of size of files.
  • Files are classified in four categories based on their size (lines of code): 1-100 (very small files), 101-200 (small files), 201-500 (medium size files), 501-1000 (long files), 1001+(very long files).
  • It is a good practice to keep files small. Long files may become "bloaters", code that have increased to such gargantuan proportions that they are hard to work with.
Learn more...
File Size Overall
  • There are 239 files with 46,406 lines of code.
    • 7 very long files (9,612 lines of code)
    • 19 long files (14,298 lines of code)
    • 43 medium size files (13,172 lines of code)
    • 31 small files (4,714 lines of code)
    • 139 very small files (4,610 lines of code)
20% | 30% | 28% | 10% | 9%
Legend:
1001+
501-1000
201-500
101-200
1-100


explore: grouped by folders | grouped by size | sunburst | 3D view
File Size per Extension
1001+
501-1000
201-500
101-200
1-100
cpp16% | 45% | 25% | 7% | 4%
py32% | 8% | 36% | 14% | 7%
cc77% | 0% | 0% | 15% | 6%
h0% | 18% | 31% | 13% | 37%
c0% | 0% | 76% | 0% | 23%
hpp0% | 0% | 0% | 0% | 100%
yaml0% | 0% | 0% | 0% | 100%
in0% | 0% | 0% | 0% | 100%
File Size per Logical Decomposition
primary
1001+
501-1000
201-500
101-200
1-100
python32% | 6% | 33% | 14% | 13%
lib16% | 46% | 24% | 6% | 5%
unittest0% | 51% | 29% | 11% | 8%
include0% | 0% | 36% | 17% | 45%
.github0% | 0% | 0% | 0% | 100%
ROOT0% | 0% | 0% | 0% | 100%
Longest Files (Top 50)
File# lines# units
Dialect.cpp
in lib/Dialect/TritonGPU/IR
 1632 108
LoadStoreOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1500 14
ElementwiseOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1465 39
ir.cc
in python/src
 1443 13
math.py
in python/triton/language
 1292 200
core.py
in python/triton/language
 1185 204
semantic.py
in python/triton/language
 1095 95
TritonGPUToLLVMBase.h
in lib/Conversion/TritonGPUToLLVM
 991 37
NVGPUToLLVMPass.cpp
in lib/Conversion/NVGPUToLLVM
 982 31
code_generator.py
in python/triton/compiler
 963 81
ConvertLayoutOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 915 4
AxisInfo.cpp
in lib/Analysis
 905 47
RemoveLayoutConversions.cpp
in lib/Dialect/TritonGPU/Transforms
 853 26
Ops.cpp
in lib/Dialect/Triton/IR
 843 51
PlanCTA.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 821 45
WSPipeline.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 811 15
TritonGPUToLLVMPass.cpp
in lib/Conversion/TritonGPUToLLVM
 748 14
TritonGPUToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 695 7
MatmulLoopPipeline.cpp
in lib/Dialect/TritonGPU/Transforms/Pipeliner
 691 16
RewriteTensorPointer.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 645 19
Utility.cpp
in lib/Analysis
 609 53
TritonToTritonGPUPass.cpp
in lib/Conversion/TritonToTritonGPU
 608 10
Utility.cpp
in lib/Dialect/TritonGPU/Transforms
 599 33
PipelineExpander.cpp
in lib/Dialect/TritonGPU/Transforms/Pipeliner
 555 6
WSMaterialization.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 555 18
EmitIndicesTest.cpp
in unittest/Conversion/TritonGPUToLLVM
 509 19
WGMMA.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 481 16
06-fused-attention.py
in python/tutorials
 481 10
Allocation.cpp
in lib/Analysis
 467 19
cuda.c
in python/triton/runtime/backends
 458 5
SharedToDotOperandMMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM
 455 7
jit.py
in python/triton/runtime
 429 44
ReduceOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 425 12
Utility.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 396 17
flash_attention.py
in python/triton/ops
 381 6
OptimizeThreadLocality.cpp
in lib/Dialect/TritonGPU/Transforms
 375 7
ScanOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 373 7
interpreter.py
in python/triton/runtime
 371 56
SharedToDotOperandMMAv1.cpp
in lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM
 369 4
RewriteTensorPointer.cpp
in lib/Dialect/Triton/Transforms
 364 13
MMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 350 11
Utility.cpp
in lib/Conversion/TritonGPUToLLVM
 315 25
build_extern.py
in python/triton/tools
 307 26
matmul.py
in python/triton/ops/blocksparse
 306 11
AccelerateMatmul.cpp
in lib/Dialect/TritonGPU/Transforms
 301 5
DumpLayout.cpp
in unittest/Conversion/TritonGPUToLLVM
 296 7
Prefetch.cpp
in lib/Dialect/TritonGPU/Transforms
 280 7
WSMutex.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 277 5
Utility.h
in lib/Conversion/TritonGPUToLLVM
 272 8
OptimizeDotOperands.cpp
in lib/Dialect/TritonGPU/Transforms
 250 3
Files With Most Units (Top 50)
File# lines# units
core.py
in python/triton/language
 1185 204
math.py
in python/triton/language
 1292 200
Dialect.cpp
in lib/Dialect/TritonGPU/IR
 1632 108
semantic.py
in python/triton/language
 1095 95
code_generator.py
in python/triton/compiler
 963 81
interpreter.py
in python/triton/runtime
 371 56
Utility.cpp
in lib/Analysis
 609 53
Ops.cpp
in lib/Dialect/Triton/IR
 843 51
AxisInfo.cpp
in lib/Analysis
 905 47
PlanCTA.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 821 45
jit.py
in python/triton/runtime
 429 44
ElementwiseOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1465 39
TritonGPUToLLVMBase.h
in lib/Conversion/TritonGPUToLLVM
 991 37
standard.py
in python/triton/language
 240 36
Utility.cpp
in lib/Dialect/TritonGPU/Transforms
 599 33
NVGPUToLLVMPass.cpp
in lib/Conversion/NVGPUToLLVM
 982 31
driver.py
in python/triton/runtime
 190 28
utils.py
in python/triton/compiler
 210 28
RemoveLayoutConversions.cpp
in lib/Dialect/TritonGPU/Transforms
 853 26
build_extern.py
in python/triton/tools
 307 26
Utility.cpp
in lib/Conversion/TritonGPUToLLVM
 315 25
Utility.h
in include/triton/Analysis
 249 24
Utility.h
in python/triton/_C/include/triton/Analysis
 249 24
Allocation.h
in include/triton/Analysis
 177 21
backend.py
in python/triton/common
 115 21
Allocation.h
in python/triton/_C/include/triton/Analysis
 177 21
EmitIndicesTest.cpp
in unittest/Conversion/TritonGPUToLLVM
 509 19
RewriteTensorPointer.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 645 19
Allocation.cpp
in lib/Analysis
 467 19
AxisInfo.h
in include/triton/Analysis
 247 19
cache.py
in python/triton/runtime
 116 19
AxisInfo.h
in python/triton/_C/include/triton/Analysis
 247 19
compiler.py
in python/triton/compiler
 198 19
WSMaterialization.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 555 18
link.py
in python/triton/tools
 235 18
Utility.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 396 17
MatmulLoopPipeline.cpp
in lib/Dialect/TritonGPU/Transforms/Pipeliner
 691 16
WGMMA.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 481 16
WSPipeline.cpp
in lib/Dialect/TritonNvidiaGPU/Transforms
 811 15
autotuner.py
in python/triton/runtime
 207 15
cuda.py
in python/triton/compiler/backends
 248 15
LoadStoreOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 1500 14
TritonGPUToLLVMPass.cpp
in lib/Conversion/TritonGPUToLLVM
 748 14
RewriteTensorPointer.cpp
in lib/Dialect/Triton/Transforms
 364 13
ir.cc
in python/src
 1443 13
Types.cpp
in lib/Dialect/Triton/IR
 93 12
ReduceOpToLLVM.cpp
in lib/Conversion/TritonGPUToLLVM
 425 12
MMAv2.cpp
in lib/Conversion/TritonGPUToLLVM/DotOpToLLVM
 350 11
MLIRTypes.h
in include/triton/Conversion
 28 11
MLIRTypes.h
in python/triton/_C/include/triton/Conversion
 28 11
Files With Long Lines (Top 12)

There are 12 files with lines longer than 120 characters. In total, there are 45 long lines.

File# lines# units# long lines
EmitIndicesTest.cpp
in unittest/Conversion/TritonGPUToLLVM
 509 19 17
semantic.py
in python/triton/language
 1095 95 5
GCNAsmFormat.h
in include/triton/Conversion/TritonGPUToLLVM
 241 7 3
PTXAsmFormat.h
in include/triton/Conversion/TritonGPUToLLVM
 180 7 3
GCNAsmFormat.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 241 7 3
PTXAsmFormat.h
in python/triton/_C/include/triton/Conversion/TritonGPUToLLVM
 180 7 3
code_generator.py
in python/triton/compiler
 963 81 3
core.py
in python/triton/language
 1185 204 2
jit.py
in python/triton/runtime
 429 44 2
link.py
in python/triton/tools
 235 18 2
build_extern.py
in python/triton/tools
 307 26 1
compiler.py
in python/triton/compiler
 198 19 1
Correlations

File Size vs. Commits (all time): 201 points

python/tutorials/04-low-memory-dropout.py x: 8 commits (all time) y: 73 lines of code python/src/ir.cc x: 3 commits (all time) y: 1443 lines of code python/src/llvm.cc x: 2 commits (all time) y: 172 lines of code python/src/nvidia.cc x: 2 commits (all time) y: 120 lines of code python/triton/compiler/backends/cuda.py x: 10 commits (all time) y: 248 lines of code python/triton/compiler/code_generator.py x: 43 commits (all time) y: 963 lines of code python/triton/compiler/compiler.py x: 52 commits (all time) y: 198 lines of code python/tutorials/01-vector-add.py x: 35 commits (all time) y: 59 lines of code python/tutorials/05-layer-norm.py x: 19 commits (all time) y: 237 lines of code python/src/main.cc x: 8 commits (all time) y: 17 lines of code python/src/passes.cc x: 2 commits (all time) y: 62 lines of code python/src/passes.h x: 1 commits (all time) y: 16 lines of code python/triton/runtime/jit.py x: 73 commits (all time) y: 429 lines of code python/triton/language/semantic.py x: 102 commits (all time) y: 1095 lines of code lib/Dialect/TritonGPU/Transforms/Coalesce.cpp x: 17 commits (all time) y: 242 lines of code lib/Dialect/TritonGPU/Transforms/ReorderInstructions.cpp x: 10 commits (all time) y: 127 lines of code python/triton/common/backend.py x: 11 commits (all time) y: 115 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/RewriteTensorPointer.cpp x: 9 commits (all time) y: 645 lines of code include/triton/Dialect/TritonGPU/Transforms/Passes.h x: 9 commits (all time) y: 27 lines of code lib/Dialect/TritonGPU/Transforms/AccelerateMatmul.cpp x: 27 commits (all time) y: 301 lines of code lib/Dialect/TritonGPU/Transforms/DecomposeConversions.cpp x: 6 commits (all time) y: 65 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeDotOperands.cpp x: 22 commits (all time) y: 250 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeEpilogue.cpp x: 4 commits (all time) y: 82 lines of code lib/Dialect/TritonGPU/Transforms/OptimizeThreadLocality.cpp x: 6 commits (all time) y: 375 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/SoftwarePipeliner.cpp x: 3 commits (all time) y: 79 lines of code lib/Dialect/TritonGPU/Transforms/Prefetch.cpp x: 17 commits (all time) y: 280 lines of code lib/Dialect/TritonGPU/Transforms/RemoveLayoutConversions.cpp x: 53 commits (all time) y: 853 lines of code python/triton/ops/matmul_perf_model.py x: 16 commits (all time) y: 117 lines of code python/triton/compiler/backends/__init__.py x: 2 commits (all time) y: 3 lines of code python/triton/runtime/backends/hip.c x: 3 commits (all time) y: 92 lines of code python/triton/runtime/driver.py x: 12 commits (all time) y: 190 lines of code python/src/interpreter.cc x: 1 commits (all time) y: 40 lines of code python/triton/language/core.py x: 142 commits (all time) y: 1185 lines of code python/triton/runtime/interpreter.py x: 7 commits (all time) y: 371 lines of code lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.cpp x: 62 commits (all time) y: 1465 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.cpp x: 42 commits (all time) y: 748 lines of code lib/Dialect/Triton/IR/Ops.cpp x: 31 commits (all time) y: 843 lines of code lib/Analysis/Utility.cpp x: 50 commits (all time) y: 609 lines of code python/triton/language/math.py x: 13 commits (all time) y: 1292 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.cpp x: 5 commits (all time) y: 555 lines of code python/triton/language/standard.py x: 10 commits (all time) y: 240 lines of code include/triton/Dialect/TritonNvidiaGPU/Transforms/Utility.h x: 2 commits (all time) y: 53 lines of code lib/Analysis/Alias.cpp x: 9 commits (all time) y: 55 lines of code lib/Analysis/Membar.cpp x: 17 commits (all time) y: 157 lines of code lib/Conversion/TritonGPUToLLVM/LoadStoreOpToLLVM.cpp x: 43 commits (all time) y: 1500 lines of code lib/Dialect/TritonGPU/IR/Dialect.cpp x: 55 commits (all time) y: 1632 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/MatmulLoopPipeline.cpp x: 8 commits (all time) y: 691 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/MaterializeLoadStore.cpp x: 5 commits (all time) y: 225 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/Utility.cpp x: 5 commits (all time) y: 396 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSMaterialization.cpp x: 9 commits (all time) y: 555 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSMutex.cpp x: 6 commits (all time) y: 277 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSPipeline.cpp x: 12 commits (all time) y: 811 lines of code python/triton/ops/matmul.py x: 40 commits (all time) y: 174 lines of code lib/Analysis/Allocation.cpp x: 25 commits (all time) y: 467 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM.cpp x: 43 commits (all time) y: 915 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv1.cpp x: 11 commits (all time) y: 369 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandMMAv2.cpp x: 23 commits (all time) y: 455 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM.cpp x: 17 commits (all time) y: 165 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/FMA.cpp x: 5 commits (all time) y: 83 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv1.cpp x: 6 commits (all time) y: 124 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/MMAv2.cpp x: 10 commits (all time) y: 350 lines of code lib/Conversion/TritonGPUToLLVM/DotOpToLLVM/WGMMA.cpp x: 17 commits (all time) y: 481 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVMBase.h x: 48 commits (all time) y: 991 lines of code lib/Conversion/TritonGPUToLLVM/TypeConverter.cpp x: 19 commits (all time) y: 138 lines of code lib/Dialect/TritonGPU/Transforms/Utility.cpp x: 46 commits (all time) y: 599 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/FenceInsertion.cpp x: 7 commits (all time) y: 96 lines of code unittest/Conversion/TritonGPUToLLVM/EmitIndicesTest.cpp x: 4 commits (all time) y: 509 lines of code unittest/Dialect/TritonGPU/SwizzleTest.cpp x: 5 commits (all time) y: 45 lines of code lib/Conversion/TritonGPUToLLVM/ViewOpToLLVM.cpp x: 18 commits (all time) y: 242 lines of code lib/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.cpp x: 41 commits (all time) y: 608 lines of code python/triton/language/__init__.py x: 38 commits (all time) y: 217 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/PlanCTA.cpp x: 8 commits (all time) y: 821 lines of code include/triton/Dialect/TritonGPU/IR/Traits.h x: 2 commits (all time) y: 30 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp x: 33 commits (all time) y: 695 lines of code lib/Dialect/TritonGPU/IR/Traits.cpp x: 5 commits (all time) y: 18 lines of code python/triton/runtime/cache.py x: 9 commits (all time) y: 116 lines of code python/triton/runtime/backends/cuda.c x: 10 commits (all time) y: 458 lines of code python/triton/compiler/__init__.py x: 6 commits (all time) y: 3 lines of code python/triton/ops/__init__.py x: 11 commits (all time) y: 5 lines of code python/triton/tools/compile.py x: 9 commits (all time) y: 104 lines of code lib/Analysis/AxisInfo.cpp x: 28 commits (all time) y: 905 lines of code python/triton/runtime/autotuner.py x: 26 commits (all time) y: 207 lines of code python/triton/__init__.py x: 41 commits (all time) y: 51 lines of code python/triton/compiler/utils.py x: 6 commits (all time) y: 210 lines of code python/triton/language/extra/cuda.py x: 6 commits (all time) y: 12 lines of code lib/Conversion/NVGPUToLLVM/NVGPUToLLVMPass.cpp x: 9 commits (all time) y: 982 lines of code lib/Conversion/TritonGPUToLLVM/BarrierOpToLLVM.cpp x: 3 commits (all time) y: 200 lines of code lib/Conversion/TritonGPUToLLVM/ConvertLayoutOpToLLVM/SharedToDotOperandFMA.cpp x: 7 commits (all time) y: 195 lines of code lib/Conversion/TritonGPUToLLVM/ReduceOpToLLVM.cpp x: 27 commits (all time) y: 425 lines of code lib/Conversion/TritonGPUToLLVM/ScanOpToLLVM.cpp x: 9 commits (all time) y: 373 lines of code lib/Conversion/TritonGPUToLLVM/Utility.cpp x: 19 commits (all time) y: 315 lines of code lib/Conversion/TritonGPUToLLVM/Utility.h x: 22 commits (all time) y: 272 lines of code lib/Dialect/NVGPU/IR/Dialect.cpp x: 2 commits (all time) y: 71 lines of code lib/Dialect/Triton/IR/Dialect.cpp x: 8 commits (all time) y: 63 lines of code unittest/Conversion/TritonGPUToLLVM/DumpLayout.cpp x: 4 commits (all time) y: 296 lines of code include/triton/Dialect/TritonGPU/IR/Dialect.h x: 21 commits (all time) y: 55 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSFeasibilityChecking.cpp x: 2 commits (all time) y: 39 lines of code lib/Dialect/TritonGPU/Transforms/TritonGPUConversion.cpp x: 11 commits (all time) y: 83 lines of code include/triton/Dialect/Triton/IR/Traits.h x: 3 commits (all time) y: 89 lines of code lib/Dialect/Triton/IR/Traits.cpp x: 7 commits (all time) y: 211 lines of code lib/Dialect/TritonNvidiaGPU/IR/Traits.cpp x: 2 commits (all time) y: 10 lines of code include/triton/Dialect/TritonGPU/IR/TritonGPUInterfaces.h x: 1 commits (all time) y: 4 lines of code python/triton/language/random.py x: 27 commits (all time) y: 95 lines of code python/tutorials/06-fused-attention.py x: 42 commits (all time) y: 481 lines of code .pre-commit-config.yaml x: 11 commits (all time) y: 52 lines of code python/triton/common/build.py x: 6 commits (all time) y: 109 lines of code lib/Dialect/Triton/Transforms/RewriteTensorPointer.cpp x: 7 commits (all time) y: 364 lines of code .github/workflows/torch-inductor/scripts/check_perf.py x: 3 commits (all time) y: 57 lines of code python/triton/compiler/make_launcher.py x: 13 commits (all time) y: 85 lines of code python/triton/ops/blocksparse/matmul.py x: 32 commits (all time) y: 306 lines of code python/triton/ops/blocksparse/softmax.py x: 24 commits (all time) y: 183 lines of code python/triton/ops/cross_entropy.py x: 16 commits (all time) y: 62 lines of code python/triton/ops/flash_attention.py x: 20 commits (all time) y: 381 lines of code python/triton/runtime/__init__.py x: 9 commits (all time) y: 17 lines of code python/triton/runtime/errors.py x: 3 commits (all time) y: 10 lines of code python/triton/tools/build_extern.py x: 17 commits (all time) y: 307 lines of code python/triton/tools/link.py x: 7 commits (all time) y: 235 lines of code python/tutorials/02-fused-softmax.py x: 29 commits (all time) y: 81 lines of code python/tutorials/03-matrix-multiplication.py x: 39 commits (all time) y: 119 lines of code python/tutorials/07-math-functions.py x: 3 commits (all time) y: 37 lines of code python/tutorials/09-experimental-tma-matrix-multiplication.py x: 5 commits (all time) y: 138 lines of code python/tutorials/10-experimental-tma-store-matrix-multiplication.py x: 4 commits (all time) y: 104 lines of code include/triton/Dialect/TritonGPU/Transforms/Utility.h x: 11 commits (all time) y: 72 lines of code lib/Dialect/TritonGPU/Transforms/Pipeliner/PipelineExpander.h x: 1 commits (all time) y: 37 lines of code lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.h x: 6 commits (all time) y: 14 lines of code .github/workflows/torch-inductor/scripts/check_acc.py x: 1 commits (all time) y: 10 lines of code include/triton/Analysis/Alias.h x: 4 commits (all time) y: 48 lines of code include/triton/Dialect/Triton/IR/Dialect.h x: 8 commits (all time) y: 45 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSDecomposing.cpp x: 2 commits (all time) y: 179 lines of code include/triton/Analysis/Utility.h x: 26 commits (all time) y: 249 lines of code include/triton/Tools/Sys/GetEnv.hpp x: 5 commits (all time) y: 32 lines of code lib/Dialect/Triton/Transforms/ReorderBroadcast.cpp x: 4 commits (all time) y: 201 lines of code lib/Dialect/Triton/Transforms/Combine.cpp x: 11 commits (all time) y: 178 lines of code python/triton/tools/disasm.py x: 10 commits (all time) y: 89 lines of code include/triton/Dialect/TritonNvidiaGPU/IR/Dialect.h x: 3 commits (all time) y: 21 lines of code lib/Target/LLVMIR/LLVMIRBreakPhiStruct.cpp x: 1 commits (all time) y: 50 lines of code include/triton/Analysis/Allocation.h x: 9 commits (all time) y: 177 lines of code include/triton/Conversion/TritonGPUToLLVM/TritonGPUToLLVMPass.h x: 6 commits (all time) y: 20 lines of code lib/Conversion/TritonGPUToLLVM/ElementwiseOpToLLVM.h x: 8 commits (all time) y: 12 lines of code python/triton/tools/compile.c x: 3 commits (all time) y: 50 lines of code python/triton/tools/compile.h x: 4 commits (all time) y: 10 lines of code include/triton/Conversion/TritonGPUToLLVM/PTXAsmFormat.h x: 5 commits (all time) y: 180 lines of code lib/Conversion/TritonGPUToLLVM/PTXAsmFormat.cpp x: 5 commits (all time) y: 192 lines of code lib/Dialect/TritonNvidiaGPU/Transforms/WSFixupMissingAttrs.cpp x: 1 commits (all time) y: 54 lines of code include/triton/Conversion/TritonToTritonGPU/TritonToTritonGPUPass.h x: 3 commits (all time) y: 19 lines of code include/triton/Dialect/Triton/IR/Types.h x: 5 commits (all time) y: 21 lines of code include/triton/Dialect/Triton/Transforms/Passes.h x: 4 commits (all time) y: 14 lines of code include/triton/Dialect/TritonNvidiaGPU/IR/Traits.h x: 1 commits (all time) y: 21 lines of code lib/Conversion/TritonGPUToLLVM/TensorPtrOpsToLLVM.cpp x: 1 commits (all time) y: 63 lines of code lib/Dialect/Triton/IR/Types.cpp x: 6 commits (all time) y: 93 lines of code lib/Dialect/TritonGPU/IR/Types.cpp x: 1 commits (all time) y: 27 lines of code lib/Target/LLVMIR/LLVMDIScope.cpp x: 1 commits (all time) y: 115 lines of code python/MANIFEST.in x: 7 commits (all time) y: 5 lines of code include/triton/Conversion/TritonGPUToLLVM/AsmFormat.h x: 2 commits (all time) y: 21 lines of code include/triton/Conversion/TritonGPUToLLVM/GCNAsmFormat.h x: 2 commits (all time) y: 241 lines of code lib/Conversion/TritonGPUToLLVM/GCNAsmFormat.cpp x: 1 commits (all time) y: 161 lines of code lib/Conversion/TritonGPUToLLVM/TypeConverter.h x: 7 commits (all time) y: 22 lines of code unittest/Conversion/TritonGPUToLLVM/PTXAsmFormatTest.cpp x: 2 commits (all time) 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
0 142.0
commits (all time)
min: 1.0 | average: 12.4 | 25th percentile: 2.0 | median: 6.0 | 75th percentile: 14.5 | max: 142.0

File Size vs. Contributors (all time): 201 points

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

File Size vs. Commits (30 days): 0 points

No data for "commits (30d)" vs. "lines of code".

File Size vs. Contributors (30 days): 0 points

No data for "contributors (30d)" vs. "lines of code".

File Size vs. Commits (90 days): 0 points

No data for "commits (90d)" vs. "lines of code".

File Size vs. Contributors (90 days): 0 points

No data for "contributors (90d)" vs. "lines of code".