Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[WIP] Merge apache/incubator-tvm #87

Merged
merged 267 commits into from
Mar 2, 2020
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
267 commits
Select commit Hold shift + click to select a range
db08bf7
[REFACTOR] Unified IR base types. (#4616)
tqchen Jan 4, 2020
c88830e
[REFACTOR] TVM_REGISTER_API -> TVM_REGISTER_GLOBAL (#4621)
tqchen Jan 4, 2020
47b425d
[REFACTOR] IRPrinter->NodePrinter, move to node/printer.h (#4622)
tqchen Jan 5, 2020
32f58d7
tensor_array split test (#4619)
zhiics Jan 5, 2020
b86a4ed
Added declare of aluBits for TensorAlu (#4624)
kevinyuan Jan 6, 2020
3eb73ee
Get around limitation of g++-4.8 (#4626)
tqchen Jan 6, 2020
41876cc
[REFACTOR] Automatically deduce function type signature in Registry.s…
tqchen Jan 6, 2020
7cceff6
[Topi]Allow empty tensor for reshape, tile and strided_slice (#4618)
kevinthesun Jan 6, 2020
34da389
[CONV] Asymmetric padding (#4511)
optima2005 Jan 6, 2020
9716fa4
[REFACTOR][IR] Introduce SeqStmt to replace ir::Block (#4627)
tqchen Jan 6, 2020
44072f8
Improve comments (#4633)
Jan 6, 2020
9afc152
Pin python pillow to "<7" due to torchvision 1.2.0 dependency issue (…
leandron Jan 6, 2020
41e20cd
Update image version tags in Dockerfile comments (#4631)
leandron Jan 6, 2020
4260dbf
[FRONTEND][Keras] Add support for tf.Keras networks in Relay Keras fr…
leandron Jan 6, 2020
5c958bf
[COMMUNITY] @wweic -> committer (#4636)
tqchen Jan 6, 2020
754ce68
[CI] better deletion script for pycache (#4635)
tqchen Jan 7, 2020
df54994
Resolve constexpr related link error in debug mode (#4641)
tqchen Jan 7, 2020
96b0d77
[QNN] Channel wise quantization - Quantize & Requantize (#4629)
anijain2305 Jan 7, 2020
6b9a76c
[RUNTIME][DSO] Improve TVMBackendPackedCFunc to allow return val (#4637)
tqchen Jan 7, 2020
eaa9a23
[COMMUNITY] @MarisaKirisame -> committer (#4645)
icemelon Jan 8, 2020
ec061d4
reduce input size to fix oom (#4653)
zhiics Jan 8, 2020
546b40d
[REFACTOR][IR] Add Node suffix to low-level IR nodes (#4649)
tqchen Jan 8, 2020
213f8a1
[CONV] Reduce data size of asymmetric padding testcase (#4658)
optima2005 Jan 8, 2020
a0af2d5
[CI] Recover Windows Mac Build CI via Github Actions (#4662)
tqchen Jan 9, 2020
d7319ad
[Autotvm] Use VM compile to extract autotvm tasks (#4328)
icemelon Jan 9, 2020
43a22c5
[Relay/Topi][Op] 1D Pooling (#4663)
jwfromm Jan 9, 2020
23f60be
[REFACTOR] relay::Module Def -> TypeDef (#4665)
tqchen Jan 9, 2020
444ee5e
[Relay][Frontend][TFlite] Add parses support for unary elemwise ops (…
inadob Jan 9, 2020
bb4554f
[REFACTOR][IR] tvm::Expr -> PrimExpr(Primitive Expr) (#4669)
tqchen Jan 9, 2020
bdc4a0e
download fallback config file for search from tophub if it does not e…
Jan 10, 2020
13864e3
Added pool autopadding and simplified parsers. (#4672)
jwfromm Jan 10, 2020
8653ab5
[VTA] Update docker for TSIM based simulation (#4674)
liangfu Jan 10, 2020
d60aaa2
[CodeGen] Generate blob use LLVM directly (#4657)
FrozenGene Jan 10, 2020
7234e9a
[CI] Update deps for chisel (#4675)
tqchen Jan 10, 2020
54897b4
fix topi.nn.global_pool layout="NHWC" (#4656)
qihaitao Jan 10, 2020
a849bf9
[CI] Bump to use the new cpu image (#4677)
tqchen Jan 10, 2020
2289760
Also package core.rly (#4679)
abergeron Jan 10, 2020
d3a53e4
GitHub actions/checkout@v1 --> v2 (#4680)
cclauss Jan 10, 2020
c8e9900
os.path --> osp to match the import (#4681)
cclauss Jan 10, 2020
e7f4128
[Relay][Frontend][TFlite] Add parses support for SLICE (#4502)
inadob Jan 10, 2020
e5206fd
Fix Python syntax error in start_rpc_server_to_tracker.py (#4682)
cclauss Jan 10, 2020
c8cff57
[Bugfix] fskip of EliminateCommonSubexpr cannot always return false (…
yuliujq Jan 11, 2020
e9d0706
[Relay][TOPI]Fix meaning of conv2d_transpose output_padding parameter…
abergeron Jan 11, 2020
d8fb2cf
[REFACTOR] Replace TensorObj and TensorValue with NDArray (#4643)
zhiics Jan 11, 2020
ca8b724
[REFACTOR][IR] Initialize Unified IR Expr Data Structure (#4673)
tqchen Jan 11, 2020
4675a76
[TOPI][RELAY][OP] add op crop_and_resize (#4417)
yongwww Jan 11, 2020
f05c988
Fix Python syntax error AGAIN in start_rpc_server_to_tracker.py (#4685)
cclauss Jan 11, 2020
6e64f53
Use ==/!= to compare str, bytes, and int literals (#4686)
cclauss Jan 11, 2020
151d986
[REFACTOR][IR] Allow Module to store BaseFunc. (#4678)
tqchen Jan 11, 2020
fedbe7b
Update and rename start_rpc_server_to_tracker.py to start_rpc_server_…
cclauss Jan 11, 2020
85a427f
[Tutorial] Deploy Quantized Model on CUDA (#4667)
vinx13 Jan 11, 2020
18dc161
[REFACTOR][IR] Unified IR Primitive Op and Registry (#4687)
tqchen Jan 11, 2020
9b01d20
[Relay/Topi][Op] Conv1D (#4639)
jwfromm Jan 11, 2020
875ceac
GitHub Action lint Python code for syntax errors (#4688)
cclauss Jan 12, 2020
bfce223
[REFACTOR][IR] Unified IR IRModule structure. (#4699)
tqchen Jan 14, 2020
ede10d2
[VTA] Fix an issue in updating uop_idx in the TensorGemm module (#4694)
liangfu Jan 14, 2020
2ac6ada
fix RemoveUnusedFunctions pass
zhiics Jan 13, 2020
db4d9de
[REFACTOR][IR] Move error.h into ir (#4701)
tqchen Jan 14, 2020
9faeaf2
[REFACTOR][IR] Initialize Unified IR Pass Infra. (#4702)
tqchen Jan 14, 2020
45a6ff1
[relay] Relay annotation and partitioning for external compilers (#4570)
zhiics Jan 14, 2020
9eee098
[REFACTOR][IR] Polish ir/type (#4705)
tqchen Jan 14, 2020
2a43201
[REFACTOR][IR] Unify IntImm and UIntImm (#4706)
tqchen Jan 15, 2020
f7250be
use packed func macro for external codegen (#4710)
zhiics Jan 15, 2020
bca5a67
Revert "[Relay][TOPI]Fix meaning of conv2d_transpose output_padding p…
icemelon Jan 15, 2020
32e59ad
link the math library by default (#4713)
zhigaowu Jan 15, 2020
a91303b
[Relay][Frontend][TF] fix _parse_param bug (#4711)
fwd4 Jan 15, 2020
1e6bbb9
[Relay][Frontend][TFLite] Add constant input support for elemwise ops…
wyc-ruiker Jan 15, 2020
f667078
[REFACTOR][IR] attrs.h -> ir (#4709)
tqchen Jan 15, 2020
fa22c2f
[REFACTOR] Move support related code to include/tvm/support (#4716)
tqchen Jan 15, 2020
5d9f4fa
[REFACTOR][FFI] Make more clear naming for C API Type codes. (#4715)
tqchen Jan 16, 2020
9e24150
[VERSION] Update mainline version to 0.7.dev0 (#4720)
tqchen Jan 16, 2020
9b7a1a4
[REFACTOR][IR] Introduce include/tvm/target (#4721)
tqchen Jan 16, 2020
ee21948
[Arith] add SizeVar representing non-neg valued variable in a tensor …
yzhliu Jan 16, 2020
c96e256
[COMMUNITY] @FrozenGene -> committer (#4719)
tqchen Jan 16, 2020
b2e35ab
[Relay][Frontend][TFLite] Add parser support for squared difference (…
wyc-ruiker Jan 16, 2020
aab749e
[CPP RPC] Fix the compile problem of cpp_rpc (#4725)
FrozenGene Jan 16, 2020
2323914
[Relay][Op] Add type check to dense (#4724)
wweic Jan 16, 2020
d61177d
[REFACTOR][ARITH] Unified IR, introduce arith subfolder. (#4722)
tqchen Jan 16, 2020
0cef36d
[Runtime] EdgeTPU runtime for Coral Boards (#4698)
tmoreau89 Jan 16, 2020
aa9fa94
[Docs] Bring Your Own Codegen Guide -- Part 1 (#4602)
comaniac Jan 16, 2020
3c7eebc
[REFACTOR] top - namespace for Tensor Operation DSL (#4727)
tqchen Jan 16, 2020
980da22
[Docs] Convert Layout pass. (#4664)
anijain2305 Jan 16, 2020
6b7f273
[REFACTOR] Polish runtime (#4729)
tqchen Jan 17, 2020
8b4fc0c
[Relay] Invoke tvm::build from relay compile_engine and interpreter (…
hlu1 Jan 17, 2020
3773aab
export builtin_fp16 on Windows (#4731)
vexilligera Jan 17, 2020
809151f
[VTA] Update Jenkinsfile for VTA test with TSIM (#4734)
liangfu Jan 17, 2020
1d7d38e
[QNN] Conv2D type checking for kernel per-channel scales. (#4732)
anijain2305 Jan 17, 2020
1838fcb
[TOOLS] JSON upgrader to upgrade serialized json. (#4730)
tqchen Jan 17, 2020
708e9ad
[x86 schedule] Fallback schedule for Int8 depthwise. (#4733)
anijain2305 Jan 17, 2020
e2c1cbf
[REFACTOR] Get rid of packed_func_ext. (#4735)
tqchen Jan 17, 2020
1c63225
[VTA][TSIM] Enable TSIM CI Testing (#4407)
liangfu Jan 17, 2020
c6848c6
[CodeGen][CUDA] Improve CUDA vectorizer (#4736)
wpan11nv Jan 18, 2020
174ddbe
[runtime][refactor] Unify vm and interpreter objects (#4693)
zhiics Jan 18, 2020
e3953a0
Fix dense (#4728)
icemelon Jan 18, 2020
2e89fa2
[REFACTOR] Establish tir (#4740)
tqchen Jan 19, 2020
643c2b3
Fix demo dockerfile build failed (#4744)
pingsutw Jan 19, 2020
06e16da
[REFACTOR][CODEGEN] codegen->target, build_module->driver (#4742)
tqchen Jan 19, 2020
346a72f
Add CUDA conv2d for NHWC layout (#4737)
alexgl-github Jan 20, 2020
cffdcb1
[REFACTOR][TYPE] Finish move all types to IR. (#4746)
tqchen Jan 20, 2020
9eb1870
Expose relay BindParamsByName to Python (#4751)
masahi Jan 20, 2020
4166672
[REFACTOR] Establish printer in the source folder (#4752)
tqchen Jan 21, 2020
ade80b3
[REFACTOR] top->te (#4759)
tqchen Jan 21, 2020
71bdbee
[INFO] Add .asf.yaml for github info (#4761)
tqchen Jan 21, 2020
0889dad
[Docs] Bring Your Own Codegen Guide -- Part 2 (#4718)
comaniac Jan 21, 2020
76e215f
[REFACTOR] driver.h -> driver_api.h (#4760)
tqchen Jan 22, 2020
ea70f5b
Fix padding in pooling op (#4738)
apivovarov Jan 22, 2020
88e0bc7
Remove run_infer_type duplicates (#4766)
apivovarov Jan 22, 2020
f27ff8b
Improve CUDA conv2d_transpose_nchw (#4762)
alexgl-github Jan 22, 2020
0639694
pooling.cc improvements (#4767)
apivovarov Jan 23, 2020
33746c6
[VTA] Support network which have no unique operator as start/stop nam…
huajsj Jan 23, 2020
4a2043f
[Doc] TVM_REGISTER_API -> TVM_REGISTER_GLOBAL (#4768)
hlu1 Jan 24, 2020
60b253b
Fix Tensorflow conv3d pad bug, add non-cubic data and kernel tests (#…
alexgl-github Jan 24, 2020
46cda65
[TOPI] Remove cpp upsampling and resize op (#4769)
masahi Jan 24, 2020
89e214a
add missing nullptr check (#4773)
masahi Jan 24, 2020
384ab51
[Bugfix][Frontend][TF] Fix incorrect calculations in tf SLICE (#4518)
inadob Jan 24, 2020
5cb62fd
Bump prebuilt-image version in demo dockerfile (#4770)
pingsutw Jan 25, 2020
b154053
Update tune_simple_template.py (#4778)
changkaiyan Jan 27, 2020
6badb7a
[Build] Explicitly link to cublasLt if it exists (#4776)
soiferj Jan 27, 2020
0c6bbe6
properly extract error type from windows error message (#4780)
soiferj Jan 27, 2020
242b5b7
[Relay][Frontend][ONNX] Broadcast condition, x, and y for Where op (#…
soiferj Jan 27, 2020
2086d92
Safe remove tmpdir (#4781)
comaniac Jan 28, 2020
652b21c
[PassManager] Implement pass manager tracing API (#4782)
jroesch Jan 28, 2020
096e1a7
[Python] Replace os.path.exists with try...except...else (#4784)
hlu1 Jan 29, 2020
3086a9d
[AUTOTVM] Fix a bug in generating the search space (#4779)
wpan11nv Jan 29, 2020
cd13e95
Make sure to visit the arguments of inlined functions (#4783)
abergeron Jan 30, 2020
ee40d77
[Relay][Frontend][TFlite] Add add parser support for relational ops (…
inadob Jan 30, 2020
c8e7fa1
Fix parsing of different exception string formats (#4785)
jmorrill Jan 30, 2020
8473663
Dedup BindParamByName function in VM compiler (#4793)
masahi Jan 30, 2020
5024fd5
[Relay][Topi] Use SimplifyInference for L2 Normazlization. (#4795)
anijain2305 Jan 31, 2020
af1cb07
Add schedule for conv3d NDHWC layout (#4775)
alexgl-github Feb 1, 2020
b3b29ff
[Relay] Expose vm OptimizeModule to Python (#4800)
masahi Feb 2, 2020
66a9e9a
fix #4670: add bias for fc layer (#4801)
kshitij12345 Feb 2, 2020
23c2e57
[QNN] Doc fix on convolution and dequantize (#4799)
masahi Feb 3, 2020
078ecb9
[QNN] Conv2D with dilation support. (#4796)
anijain2305 Feb 3, 2020
95b719e
Change color channel from BGR to RGB for darknet preprocessing (#4794)
vizero1 Feb 3, 2020
eee9f26
[ThreadPool] Solve ARM BIG.LITTLE heterogeneous multicores (#4747)
FrozenGene Feb 3, 2020
7311065
[TIR] Create a StringImm reference type (#4806)
mbaret Feb 3, 2020
2fdec4b
[TOPI] upsample operator 'NCHWinic' format support. (#4791)
huajsj Feb 3, 2020
b3d3c0b
[LINT] Fix -Wextra (#4804)
tqchen Feb 4, 2020
d1d6550
[DOCS] Fix vta tutorial (#4809)
tqchen Feb 4, 2020
2d6f6a5
[AutoTVM] Minor bug fixes in AutoTVM for QNN graphs (#4797)
anijain2305 Feb 4, 2020
d353f1b
fix memory leak (#4811)
icemelon Feb 4, 2020
67ba9c7
[TOPI][x86] Injective schedule improvement (#4786)
anijain2305 Feb 4, 2020
96b550e
[REFACTOR][PY] tvm._ffi (#4813)
tqchen Feb 5, 2020
6dec647
allow customize mkldnn library location (#4814)
icemelon Feb 5, 2020
ac53f98
Mxnet parser for Qnn dialect (#4714)
shoubhik Feb 5, 2020
9071cf5
[REFACTOR][PY] Establish tvm.runtime (#4818)
tqchen Feb 5, 2020
a0a2eda
Fixed subprocess creation under windows (#4820)
Coderx7 Feb 5, 2020
a8e5631
[Frontend][TFLite] Dynamically calculate input_stats of any fake_quan…
inadob Feb 5, 2020
5e78e3a
[QNN] Optimize lowering for requantize and FixedPointMultiply. (#4798)
anijain2305 Feb 5, 2020
82f1a79
[Relay][Frontend][TFLite] Add parser support for logical operators (#…
inadob Feb 5, 2020
a17742d
[Relay] Conv2D padding representation (#4787)
Feb 5, 2020
8934bc4
[CONTRIB][CC] Enhance cc.cross_compiler (#4817)
tqchen Feb 6, 2020
38eb5fc
[TOPI][Relay] Add bitwise ops (#4815)
abergeron Feb 6, 2020
09ec230
It's gpu not cpu. (#4832)
clhne Feb 6, 2020
f64be22
[CI][DOCKER] Update ci-gpu to v0.60 (#4827)
tqchen Feb 6, 2020
b9137e6
[CI][DOCKER] Update ci-gpu torch1.4 and onnx1.6 (#4826)
tqchen Feb 6, 2020
9887418
Fix doc after moving to unified IR (#4835)
zhiics Feb 7, 2020
4d1238f
[Doc] Introduction to module serialization (#4564)
FrozenGene Feb 7, 2020
cd9a368
[Frontend][ONNX] LSTM Support (#4825)
jwfromm Feb 7, 2020
94fb61d
Improve tol to resolve flaky case (#4836)
tqchen Feb 7, 2020
3d70230
[Doc] ConvertLayout - Call RemoveUnunsedFunctions.
anijain2305 Feb 6, 2020
97248b4
[Relay][Frontend][TFlite] Add support for quantized LOGISTIC (#4696)
inadob Feb 7, 2020
cbb12fc
[Frontend][TFLite] Add MIRROR_PAD operator (#4822)
wyc-ruiker Feb 7, 2020
c3b9269
[REFACTOR][PY][API-Change] Polish tvm.runtime, tvm.runtime.module API…
tqchen Feb 7, 2020
5d0bbea
[COMMUNITY] comaniac -> reviewer (#4841)
ZihengJiang Feb 7, 2020
268cd7d
[Doc][AutoTVM] Fix bugs that override n_trials (#4842)
comaniac Feb 7, 2020
0ae95db
Fixed process termination routine in windows (#4844)
Coderx7 Feb 8, 2020
15377b7
[TEST] test_cuddn flaky (#4846)
tqchen Feb 8, 2020
4c48e2b
[LINT][PY] Fixes for pylint==2.4.4 (#4849)
tqchen Feb 9, 2020
7608c84
[CI] Update ci-lint to v0.60 (#4850)
tqchen Feb 9, 2020
1d212a2
[CI][DOCKER] Update ci-lint to pylint2.4.4 (#4851)
tqchen Feb 9, 2020
047d793
[Frontend][TFlite] use qnn helper function in softmax (#4840)
wyc-ruiker Feb 10, 2020
0aabc54
Fixed bug in ExprOp that caused bitwise operators to fail when a basi…
dpankratz Feb 10, 2020
0ecffb9
[Relay] Added Merge Composite pass (#4771)
mbaret Feb 10, 2020
7c51d11
reverse changes in pr #4849 (#4853)
Laurawly Feb 10, 2020
7306114
[TFLite] Using real image for QNN testing. (#4816)
anijain2305 Feb 11, 2020
0f86db6
[RUNTIME] Fix memory leakage of TVMByteArray (#4856)
merrymercy Feb 11, 2020
4839c7e
[LLVM] Explicit llvm::StringRef to std::string conversion (#4859)
hlu1 Feb 11, 2020
eb0a70d
[TVM] const auto p -> const auto &p (#4861)
hlu1 Feb 11, 2020
ae79ce8
add resize op converter (#4838)
masahi Feb 11, 2020
0771b02
[Refactor] move vm.py under runtime and adt to runtime.container.py (…
zhiics Feb 11, 2020
e4b8b35
Fix onnx import bugs (#4750)
kice Feb 11, 2020
4e7edbe
[Topi] Missing header (#4865)
hlu1 Feb 11, 2020
2b90113
[REFACTOR][PY][API-CHANGE] establish tvm.ir, migrate corresponding fi…
tqchen Feb 12, 2020
e1ecef6
[DOCS][PY] Sphinx docs about tvm.ir
tqchen Feb 12, 2020
1be9754
Fix optimize
tqchen Feb 12, 2020
75cd6cd
[JVM] Update the runtime PackedFunc for module
tqchen Feb 12, 2020
ec7707b
[REFACTOR][PY][API-CHANGE] Establish tvm.target
tqchen Feb 12, 2020
363bfee
[FRONTEND][TFLITE] Add support for TFLite_Detection_PostProcess (#4543)
mbaret Feb 13, 2020
d9deac7
Optimize x86 conv3d_ndhwc using data packing approach. (#4866)
alexgl-github Feb 13, 2020
8b56d15
fix vm doc
zhiics Feb 11, 2020
388bc5f
Update docs/dev/virtual_machine.rst
zhiics Feb 12, 2020
46dd9f4
Update docs/dev/virtual_machine.rst
zhiics Feb 12, 2020
65e5403
[REFACTOR][PY] Establish tvm.tir
tqchen Feb 12, 2020
b4e5db2
[TOPI][CUDA] Enable vectorization on fp16 type (#4867)
wpan11nv Feb 14, 2020
fc80731
[QNN] More doc fix on quantize and convolution (#4874)
masahi Feb 14, 2020
3808cd1
[QNN] Add support for per channel weight scale in dense op (#4880)
masahi Feb 15, 2020
c9724b5
[AutoTVM] Support range in index based tuners (#4870)
comaniac Feb 15, 2020
c834c5b
improve antlr import error message (#4888)
masahi Feb 15, 2020
afe236b
[CodeGen][CUDA] Fix issues in cuda codegen (#4876)
wpan11nv Feb 16, 2020
4484283
[Relay] Fix VM compiler for while loop with free vars (#4889)
masahi Feb 16, 2020
19d3a09
[CI] Cleanup logfile before tutorial runs (#4896)
tqchen Feb 16, 2020
abfb409
Fix alpha_equal bug (#4897)
zhiics Feb 17, 2020
719a91a
Update faq.md (#4893)
badenh Feb 17, 2020
bca1935
Fast exponent (#4790)
alexgl-github Feb 17, 2020
8176f43
[DOCS] Introduce how to add hardware backend to FAQ (#4898)
tqchen Feb 17, 2020
56bd0ac
[Relay][Pass] Fix bug in re-processing call node in MergeComposite pa…
soiferj Feb 17, 2020
0de6f02
[REFACTOR][PY] Establish tvm.te and tvm.driver (#4900)
tqchen Feb 18, 2020
78ebb38
Fixed bugs that occured when using bitwise operators on floating poin…
dpankratz Feb 18, 2020
4a0dc91
[CI] Update ci docker to add autodocsumm (#4903)
tqchen Feb 18, 2020
40a2342
[CI] Add autodocsum as dep (#4902)
tqchen Feb 18, 2020
6e35155
[REFACTOR][PY] Establish tvm.arith (#4904)
tqchen Feb 18, 2020
27f0c78
[Relay][Frontend][Keras] NHWC import support. (#4899)
jwfromm Feb 18, 2020
c2353e3
[Relay] Expose FunctionGetAttr to Python (#4905)
soiferj Feb 18, 2020
68ea861
[DOCS] Update API docs to reflect the status after the refactor. (#4907)
tqchen Feb 18, 2020
0b6164d
Fix tvm.target.generic_func runtime detection (#4910)
robo-corg Feb 19, 2020
d3e7992
[RELAY][FRONTEND][TF] Fix FuseBatchNorm output cast error if need_cas…
yhcvb Feb 19, 2020
911c3b4
[REFACTOR] Polish ffi convention. (#4912)
tqchen Feb 19, 2020
fd28647
[DOCS] Fix sphinx warnings (#4917)
comaniac Feb 20, 2020
aa6d17f
[Relay] Fix an assertion exposed by loop vectorizer (#4916)
wpan11nv Feb 20, 2020
7c9f3dc
[DOCS] Fix Sphinx Warnings (RST indent, cross-ref, and image scale) (…
comaniac Feb 20, 2020
4f5041c
[CODEGEN] Support cuda tensorcore subbyte int data type in auto tenso…
Orion34-lanbo Feb 21, 2020
e6930fa
Fix tests for tflite unary elemwise operations (#4913)
inadob Feb 21, 2020
6831d50
[COMMUNITY] @anijain2305 -> Committer (#4921)
tqchen Feb 21, 2020
7c0a4e6
[TEST][FLAKY] topi/tests/python/test_topi_sort.py::test_argsort (#4891)
cchung100m Feb 21, 2020
f492525
[Fix] Fix get_valid_count flaky test for cuda (#4901)
Laurawly Feb 21, 2020
194a863
[Relay][AutoTVM] Relay op strategy (#4644)
icemelon Feb 24, 2020
7303b7d
[FRONTEND][KERAS]GaussianDropout/Noise parsing support (#4928)
siju-samuel Feb 25, 2020
9e100d1
Use opencv reisze method for preprocessing of image in darknet (#4883)
vizero1 Feb 25, 2020
60a4ee8
[Relay] Add a PyTorch to Relay Parser (#4497)
alexwong Feb 25, 2020
ccadb53
[Relay][External Codegen] Support data types for CSourceModuleCodegen…
soiferj Feb 25, 2020
1dfb61c
[LLVM] Fix build breaks from StringRef changes (#4923)
wpan11nv Feb 25, 2020
7702e48
[Fix] remove unnecessary spliting in the cached chunk (#4935)
yidawang Feb 25, 2020
59f0bdf
[WIP] Fixing an Infinite Loop case in UnmatchedChecker. (#4881)
MarisaKirisame Feb 26, 2020
146a7b8
Tensor Expression Debug Display (TEDD) (#4651)
yongfeng-nv Feb 26, 2020
ca852c9
[DOCS] Fix Sphinx Warning: the target found for cross-reference (#4925)
cchung100m Feb 26, 2020
b18c0a1
Bump up dev version (#4941)
icemelon Feb 26, 2020
6d6638e
[Tutorial] Add a tutorial for PyTorch (#4936)
alexwong Feb 26, 2020
dab13fb
[Relay][pass] call graph for relay (#4922)
zhiics Feb 26, 2020
5ebbba2
Remove SGX toolchain installation from CI Dockerfile (#4948)
nhynes Feb 26, 2020
104d944
[Frontend][TFLite] Add parser support for 'square' operator (#4915)
inadob Feb 26, 2020
48a38b1
[VTA] YoloV3 Support (#4887)
huajsj Feb 26, 2020
a9f7c95
[TUTORIAL] Fix tedd tutorial after strategy change (#4947)
tqchen Feb 27, 2020
d05c0dc
Update dmlc_tvm_commit_id.txt
alexwong Feb 26, 2020
e7ba1a4
Remove merge artifact
alexwong Feb 26, 2020
3d6275d
Add back empty functions
alexwong Feb 28, 2020
8b25353
Use runtime.module in tensorrt test
alexwong Feb 28, 2020
95f3c65
Rename the custom functions and call them
alexwong Feb 28, 2020
3599939
Rename tvm.module -> tvm.runtime
alexwong Feb 28, 2020
File filter

Filter by extension

Filter by extension


Conversations
Failed to load comments.
Loading
Jump to
The table of contents is too big for display.
Diff view
Diff view
  •  
  •  
  •  
34 changes: 34 additions & 0 deletions .asf.yaml
Original file line number Diff line number Diff line change
@@ -0,0 +1,34 @@
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.

github:
description: "Open deep learning compiler stack for cpu, gpu and specialized accelerators"
homepage: https://tvm.apache.org/
labels:
- tvm
- compiler
- tensor
- deep-learning
- gpu
- opencl
- metal
- performance
- javascript
- rocm
- vulkan
- spirv
- machine-learning
79 changes: 79 additions & 0 deletions .github/workflows/main.yml
Original file line number Diff line number Diff line change
@@ -0,0 +1,79 @@
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.

# GH actions.
# We use it to cover windows and mac builds
# Jenkins is still the primary CI

name: WinMacBuild

on:
push:
branches:
- master

pull_request:
branches:
- master

jobs:
Build:
strategy:
matrix:
os: [windows-latest, macOS-latest]

runs-on: ${{ matrix.os }}

steps:
- uses: actions/checkout@v2
- uses: actions/setup-python@v1
- name: Lint Python
if: matrix.os == 'macOS-latest'
run: |
pip install flake8
flake8 . --count --select=E9,F63,F7 --show-source --statistics
- name: Initialize submodules
run: git submodule update --recursive --init

- name: Make Build Directory
run: cmake -E make_directory build.common

# configuration for Windows
- name: CMake@Win
if: matrix.os == 'windows-latest'
working-directory: build.common
run: >-
cmake
"-DUSE_SORT=ON"
"-DUSE_RPC=ON"
"-DUSE_GRAPH_RUNTIME=ON"
..

# configuration for Mac
- name: CMake@MacOS
if: matrix.os == 'macOS-latest'
working-directory: build.common
run: >-
cmake
"-DUSE_SORT=ON"
"-DUSE_RPC=ON"
"-DUSE_GRAPH_RUNTIME=ON"
"-DUSE_METAL=ON"
..

- name: Build
run: cmake --build build.common -j3
2 changes: 1 addition & 1 deletion .gitignore
Original file line number Diff line number Diff line change
Expand Up @@ -65,7 +65,7 @@ docs/_build/
docs/gen_modules

# PyBuilder
target/
/target/

# IPython Notebook
.ipynb_checkpoints
Expand Down
33 changes: 20 additions & 13 deletions CMakeLists.txt
Original file line number Diff line number Diff line change
Expand Up @@ -101,8 +101,8 @@ if(MSVC)
else(MSVC)
if ("${CMAKE_BUILD_TYPE}" STREQUAL "Debug")
message("Build in Debug mode")
set(CMAKE_C_FLAGS "-O0 -g -Wall -fPIC ${CMAKE_C_FLAGS} -rdynamic")
set(CMAKE_CXX_FLAGS "-O0 -g -Wall -fPIC ${CMAKE_CXX_FLAGS} -rdynamic")
set(CMAKE_C_FLAGS "-O0 -g -Wall -fPIC ${CMAKE_C_FLAGS}")
set(CMAKE_CXX_FLAGS "-O0 -g -Wall -fPIC ${CMAKE_CXX_FLAGS}")
else()
set(CMAKE_C_FLAGS "-O2 -Wall -fPIC ${CMAKE_C_FLAGS}")
set(CMAKE_CXX_FLAGS "-O2 -Wall -fPIC ${CMAKE_CXX_FLAGS}")
Expand All @@ -125,18 +125,25 @@ assign_source_group("Source" ${GROUP_SOURCE})
assign_source_group("Include" ${GROUP_INCLUDE})

# Source file lists
file(GLOB COMPILER_SRCS
src/api/*.cc
src/arithmetic/*.cc
src/autotvm/*.cc
src/codegen/*.cc
src/lang/*.cc
src/pass/*.cc
src/op/*.cc
file(GLOB_RECURSE COMPILER_SRCS
src/node/*.cc
src/schedule/*.cc
src/ir/*.cc
src/arith/*.cc
src/te/*.cc
src/autotvm/*.cc
src/tir/*.cc
src/driver/*.cc
src/printer/*.cc
src/support/*.cc
)

file(GLOB CODEGEN_SRCS
src/target/*.cc
src/target/source/*.cc
)

list(APPEND COMPILER_SRCS ${CODEGEN_SRCS})

file(GLOB_RECURSE RELAY_OP_SRCS
src/relay/op/*.cc
)
Expand Down Expand Up @@ -166,7 +173,7 @@ if(USE_VM_PROFILER)
list(APPEND COMPILER_SRCS ${BACKEND_VM_PROFILER_SRCS})
endif(USE_VM_PROFILER)

file(GLOB DATATYPE_SRCS src/codegen/datatype/*.cc)
file(GLOB DATATYPE_SRCS src/target/datatype/*.cc)
list(APPEND COMPILER_SRCS ${DATATYPE_SRCS})

file(GLOB TOPI_SRCS
Expand All @@ -192,7 +199,7 @@ if(USE_RPC)
endif(USE_RPC)

file(GLOB STACKVM_RUNTIME_SRCS src/runtime/stackvm/*.cc)
file(GLOB STACKVM_CODEGEN_SRCS src/codegen/stackvm/*.cc)
file(GLOB STACKVM_CODEGEN_SRCS src/target/stackvm/*.cc)
list(APPEND COMPILER_SRCS ${STACKVM_CODEGEN_SRCS})
if(USE_STACKVM_RUNTIME)
message(STATUS "Build with stackvm support in runtime...")
Expand Down
6 changes: 5 additions & 1 deletion CONTRIBUTORS.md
Original file line number Diff line number Diff line change
Expand Up @@ -41,10 +41,13 @@ We do encourage everyone to work anything they are interested in.

- [Aditya Atluri](https://github.com/adityaatluri): @adityaatluri - rocm
- [Tianqi Chen](https://github.com/tqchen) (PPMC): @tqchen - topi, compiler, relay, docs
- [Wei Chen](https://github.com/wweic): @wweic - runtime, relay, vm
- [Zhi Chen](https://github.com/zhiics): @zhiics - relay, quantization, pass manager
- [Yuwei Hu](https://github.com/Huyuwei): @Huyuwei - topi, frontends
- [Nick Hynes](https://github.com/nhynes): @nhynes: - sgx, rust
- [Animesh Jain](https://github.com/anijain2305): @anijain2305 - quantization, relay
- [Ziheng Jiang](https://github.com/ZihengJiang) (PPMC): @ZihengJiang - relay, compiler
- [Marisa Kirisame](https://github.com/MarisaKirisame): @MarisaKirisame - relay
- [Wuwei Lin](https://github.com/vinx13): @vinx13 - relay, topi
- [Yizhi Liu](https://github.com/yzhliu) (PPMC): @yzhliu - jvm, topi, relay
- [Hao Lu](https://github.com/hlu1): @hlu1 - nnpack, frontends
Expand All @@ -59,6 +62,7 @@ We do encourage everyone to work anything they are interested in.
- [Leyuan Wang](https://github.com/Laurawly): @Laurawly: - topi
- [Yao Wang](https://github.com/kevinthesun): @kevinthesun: - topi, vision
- [Jian Weng](https://github.com/were): @were: - hybrid script
- [Zhao Wu](https://github.com/FrozenGene): @FrozenGene - runtime, topi, frontends
- [Eddie Yan](https://github.com/eqy) (PPMC): @eqy - runtime, autotvm, rpc, topi
- [Lianmin Zheng](https://github.com/merrymercy) (PPMC): @merrymercy - autotvm, topi, relay

Expand All @@ -67,7 +71,6 @@ We do encourage everyone to work anything they are interested in.
- [Aditya Atluri](https://github.com/adityaatluri): @adityaatluri
- [Tianqi Chen](https://github.com/tqchen): @tqchen
- [Liangfu Chen](https://github.com/liangfu): @liangfu
- [Wei Chen](https://github.com/wweic): @wweic
- [Zhi Chen](https://github.com/zhiics): @zhiics
- [Neo Chien](https://github.com/cchung100m): @cchung100m
- [Meghan Cowan](https://github.com/cowanmeg): @cowanmeg
Expand Down Expand Up @@ -110,6 +113,7 @@ We do encourage everyone to work anything they are interested in.
- [Zhao Wu](https://github.com/FrozenGene): @FrozenGene
- [Bing Xu](https://github.com/antinucleon): @antinucleon
- [Eddie Yan](https://github.com/eqy): @eqy
- [Hao Yu](https://github.com/comaniac): @comaniac
- [Joshua Z. Zhang](https://github.com/zhreshold): @zhreshold
- [Lianmin Zheng](https://github.com/merrymercy): @merrymercy

Expand Down
13 changes: 7 additions & 6 deletions Jenkinsfile
Original file line number Diff line number Diff line change
Expand Up @@ -41,12 +41,12 @@

// Hashtag in the source to build current CI docker builds
//
// - ci-cpu:v0.54: e7c88a99f830de30814df14eaa980547ecbd61c1
// - ci-cpu:v0.55: 07b45d958d4af91ec1bab66f6cf391d1ce12ddaf
//

ci_lint = "tvmai/ci-lint:v0.51"
ci_gpu = "tvmai/ci-gpu:v0.56"
ci_cpu = "tvmai/ci-cpu:v0.54"
ci_lint = "tvmai/ci-lint:v0.60"
ci_gpu = "tvmai/ci-gpu:v0.61"
ci_cpu = "tvmai/ci-cpu:v0.60"
ci_i386 = "tvmai/ci-i386:v0.52"

// tvm libraries
Expand Down Expand Up @@ -216,7 +216,8 @@ stage('Build') {
timeout(time: max_time, unit: 'MINUTES') {
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_unittest.sh"
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_integration.sh"
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_vta.sh"
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_vta_fsim.sh"
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_python_vta_tsim.sh"
sh "${docker_run} ${ci_cpu} ./tests/scripts/task_golang.sh"
}
}
Expand Down Expand Up @@ -268,7 +269,7 @@ stage('Unit Test') {
timeout(time: max_time, unit: 'MINUTES') {
sh "${docker_run} ${ci_i386} ./tests/scripts/task_python_unittest.sh"
sh "${docker_run} ${ci_i386} ./tests/scripts/task_python_integration.sh"
sh "${docker_run} ${ci_i386} ./tests/scripts/task_python_vta.sh"
sh "${docker_run} ${ci_i386} ./tests/scripts/task_python_vta_fsim.sh"
}
}
}
Expand Down
12 changes: 6 additions & 6 deletions Makefile
Original file line number Diff line number Diff line change
Expand Up @@ -17,7 +17,7 @@

ROOTDIR = $(CURDIR)

.PHONY: clean all test doc pylint cpplint lint\
.PHONY: clean all test doc pylint cpplint scalalint lint\
cython cython2 cython3 web runtime vta

ifndef DMLC_CORE_PATH
Expand Down Expand Up @@ -80,7 +80,10 @@ pylint:
jnilint:
python3 3rdparty/dmlc-core/scripts/lint.py tvm4j-jni cpp jvm/native/src

lint: cpplint pylint jnilint
scalalint:
make -C vta/hardware/chisel lint

lint: cpplint pylint jnilint scalalint

doc:
doxygen docs/Doxyfile
Expand All @@ -91,10 +94,7 @@ javadoc:

# Cython build
cython:
cd python; python setup.py build_ext --inplace

cython2:
cd python; python2 setup.py build_ext --inplace
cd python; python3 setup.py build_ext --inplace

cython3:
cd python; python3 setup.py build_ext --inplace
Expand Down
4 changes: 2 additions & 2 deletions apps/benchmark/util.py
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ def get_network(name, batch_size, dtype='float32'):

Returns
-------
net: relay.Module
net: tvm.IRModule
The relay function of network definition
params: dict
The random parameters for benchmark
Expand Down Expand Up @@ -70,7 +70,7 @@ def get_network(name, batch_size, dtype='float32'):
net, params = relay.frontend.from_mxnet(block, shape={'data': input_shape}, dtype=dtype)
net = net["main"]
net = relay.Function(net.params, relay.nn.softmax(net.body), None, net.type_params, net.attrs)
net = relay.Module.from_expr(net)
net = tvm.IRModule.from_expr(net)
else:
raise ValueError("Unsupported network: " + name)

Expand Down
2 changes: 1 addition & 1 deletion apps/bundle_deploy/bundle.cc
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@ TVM_BUNDLE_FUNCTION void *tvm_runtime_create() {
const std::string json_data(&build_graph_json[0],
&build_graph_json[0] + build_graph_json_len);
tvm::runtime::Module mod_syslib =
(*tvm::runtime::Registry::Get("module._GetSystemLib"))();
(*tvm::runtime::Registry::Get("runtime.SystemLib"))();
int device_type = kDLCPU;
int device_id = 0;
tvm::runtime::Module mod =
Expand Down
6 changes: 3 additions & 3 deletions apps/cpp_rpc/main.cc
Original file line number Diff line number Diff line change
Expand Up @@ -31,13 +31,13 @@
#include <vector>
#include <sstream>

#include "../../src/common/util.h"
#include "../../src/common/socket.h"
#include "../../src/support/util.h"
#include "../../src/support/socket.h"
#include "rpc_server.h"

using namespace std;
using namespace tvm::runtime;
using namespace tvm::common;
using namespace tvm::support;

static const string kUSAGE = \
"Command line usage\n" \
Expand Down
12 changes: 6 additions & 6 deletions apps/cpp_rpc/rpc_env.cc
Original file line number Diff line number Diff line change
Expand Up @@ -36,7 +36,7 @@
#include <cstring>

#include "rpc_env.h"
#include "../../src/common/util.h"
#include "../../src/support/util.h"
#include "../../src/runtime/file_util.h"

namespace tvm {
Expand Down Expand Up @@ -164,7 +164,7 @@ void LinuxShared(const std::string output,
}
cmd += " " + options;
std::string err_msg;
auto executed_status = common::Execute(cmd, &err_msg);
auto executed_status = support::Execute(cmd, &err_msg);
if (executed_status) {
LOG(FATAL) << err_msg;
}
Expand Down Expand Up @@ -195,22 +195,22 @@ void CreateShared(const std::string output, const std::vector<std::string> &file
*/
Module Load(std::string *fileIn, const std::string fmt) {
std::string file = *fileIn;
if (common::EndsWith(file, ".so")) {
if (support::EndsWith(file, ".so")) {
return Module::LoadFromFile(file, fmt);
}

#if defined(__linux__) || defined(__ANDROID__)
std::string file_name = file + ".so";
if (common::EndsWith(file, ".o")) {
if (support::EndsWith(file, ".o")) {
std::vector<std::string> files;
files.push_back(file);
CreateShared(file_name, files);
} else if (common::EndsWith(file, ".tar")) {
} else if (support::EndsWith(file, ".tar")) {
std::string tmp_dir = "./rpc/tmp/";
mkdir(&tmp_dir[0], 0777);
std::string cmd = "tar -C " + tmp_dir + " -zxf " + file;
std::string err_msg;
int executed_status = common::Execute(cmd, &err_msg);
int executed_status = support::Execute(cmd, &err_msg);
if (executed_status) {
LOG(FATAL) << err_msg;
}
Expand Down
Loading