Code

Pull#5

Code

Merged

sergei-mironov merged 77 commits intosergei-mironov:mastersergei-mironov/tvm:masterfrom apache:masterapache/tvm:masterCopy head branch name to clipboardAug 30, 2018

+11,316-3,501

Commits

Commits on Aug 8, 2018

• [NNVM] Enhance operator fusion for more element wise patterns (#1548)

  

  masahi

  authored and

  tqchen

  committed
  1ed28ae

  Copy full SHA for 1ed28ae

• add tvm external registry entrypoint (#1562)

  

  souptc

  authored and

  tqchen

  committed
  60da470

  Copy full SHA for 60da470

• [ONNX]onnx gather bug fix (#1543)

  

  siju-samuel

  authored and

  tqchen

  committed
  7ea06e6

  Copy full SHA for 7ea06e6

• l2normalization operator support for tensorflow (#1528)

  

  siju-samuel

  authored and

  tqchen

  committed
  c9f9a3f

  Copy full SHA for c9f9a3f

• Separate fusion and Compilation (#1564)

  Show description for 53d2431
  

  zhiics

  authored and

  yzhliu

  committed
  53d2431

  Copy full SHA for 53d2431

Commits on Aug 9, 2018

• [TOPI] add injective scheduler for HLS backends (#1553)

  Show description for 85483c3
  

  kazum

  authored and

  tmoreau89

  committed
  85483c3

  Copy full SHA for 85483c3

• Use int for int8x4 due to performance overhead of char4 (#1569)

  Show description for 41d4dd6
  

  vinx13

  authored and

  tqchen

  committed
  41d4dd6

  Copy full SHA for 41d4dd6

• [DOCS] Neural network Deployment Guide with System Module Mode #1523 (#1533)

  

  srkreddy1238

  authored and

  tqchen

  committed
  6d4cf44

  Copy full SHA for 6d4cf44

• add conv2d transpose and fix bugs (#1566)

  

  merrymercy

  authored and

  tqchen

  committed
  672147c

  Copy full SHA for 672147c

• [TEST] force openblas threads to be 1 (#1580)

  tqchen

  authored
  764516a

  Copy full SHA for 764516a

Commits on Aug 10, 2018

• Vulkan TVM Android Support (#1571)

  

  Dayananda-V

  authored and

  tqchen

  committed
  2afe024

  Copy full SHA for 2afe024

• [TEAM] merrymercy->code owner (#1581)

  tqchen

  authored
  48fc410

  Copy full SHA for 48fc410

• [AUTOTVM] API change (#1583)

  

  merrymercy

  authored and

  tqchen

  committed
  48ff777

  Copy full SHA for 48ff777

• update dmlc-core for security reason (#1584)

  

  haojin2

  authored and

  tqchen

  committed
  b0368e3

  Copy full SHA for b0368e3

• DLPack Conversion API (#1573)

  

  eqy

  authored and

  tqchen

  committed
  f52255b

  Copy full SHA for f52255b

Commits on Aug 11, 2018

• use phone EditText for numerical fields (#1587)

  

  eqy

  authored and

  tqchen

  committed
  edda6cc

  Copy full SHA for edda6cc

• [RUNTIME] Refactor to enable stackvm in runtime. (#1588)

  tqchen

  authored
  ec3f09b

  Copy full SHA for ec3f09b

• [DLPACK] Enable cython support (#1589)

  tqchen

  authored
  19cf5c6

  Copy full SHA for 19cf5c6

Commits on Aug 13, 2018

• Fixed bugs for SSD sorting and multbox detection (#1578)

  

  Laurawly

  authored and

  tqchen

  committed
  60769b7

  Copy full SHA for 60769b7

• Split_indices negative axis added (#1595)

  

  PariksheetPinjari909

  authored and

  yzhliu

  committed
  81b42bc

  Copy full SHA for 81b42bc

• [FRONTEND][TENSORFLOW] Optimized tensorflow testcases (#1546)

  Show description for 84296f9
  

  Albin Joy

  authored and

  yzhliu

  committed
  84296f9

  Copy full SHA for 84296f9

Commits on Aug 14, 2018

• [NNVM][DOC] Update NNVM symbol documentation to latest. Ref. 1591 (#1599)

  

  srkreddy1238

  authored and

  tqchen

  committed
  daa2569

  Copy full SHA for daa2569

• [NNVM][POOL] bug fix. Remove the hardcode. (#1600)

  

  srkreddy1238

  authored and

  tqchen

  committed
  bc7431b

  Copy full SHA for bc7431b

• [FRONTEND][DARKNET]LSTM and GRU support (#1576)

  

  siju-samuel

  authored and

  tqchen

  committed
  54a115e

  Copy full SHA for 54a115e

Commits on Aug 15, 2018

• [AUTOTVM] Fix GATuner and improve error message (#1605)

  

  merrymercy

  authored and

  tqchen

  committed
  7751a6b

  Copy full SHA for 7751a6b

• [NNVM] Add symbol for inception v3 (#1604)

  

  merrymercy

  authored and

  tqchen

  committed
  a80356b

  Copy full SHA for a80356b

Commits on Aug 16, 2018

• [TEAM] New reviewer: kevinthesun (#1606)

  tqchen

  authored
  9bcc317

  Copy full SHA for 9bcc317

• [RUNTIME] Enable return NDArray in RPC (#1610)

  tqchen

  authored
  11dd933

  Copy full SHA for 11dd933

• [NNVM] Add ONNX upsample converter (#1591)

  

  Jokeren

  authored and

  tqchen

  committed
  093dc74

  Copy full SHA for 093dc74

• add -mattr=+neon for all arm cpu target (#1612)

  

  merrymercy

  authored and

  tqchen

  committed
  e282915

  Copy full SHA for e282915

Commits on Aug 17, 2018

• fix output_shape in conv2d_nchw (#1613)

  

  ashokei

  authored and

  tqchen

  committed
  acc2151

  Copy full SHA for acc2151

• [NNVM] Bug fix Prevent fusing convolution with injective op (#1608)

  

  masahi

  authored and

  tqchen

  committed
  6cd5a8f

  Copy full SHA for 6cd5a8f

Commits on Aug 18, 2018

• [NNVM] TF: Add Pack operation (#1570)

  

  sergei-mironov

  authored and

  tqchen

  committed
  9b0e499

  Copy full SHA for 9b0e499

• #1592 [PASS] Fix missing mem CHECK in storage_rewrite (#1616)

  

  xqdan

  authored and

  tqchen

  committed
  38d0835

  Copy full SHA for 38d0835

• [FRONTEND][COREML]MultiplyLayerParams L2NormalizeLayerParams and UpsampleLayerParams support … (#1511)

  

  siju-samuel

  authored and

  tqchen

  committed
  a5f2759

  Copy full SHA for a5f2759

Commits on Aug 20, 2018

• fix import (#1621)

  

  merrymercy

  authored and

  tqchen

  committed
  1c66012

  Copy full SHA for 1c66012

• Add missing check when deciding conv op and injective op are in the same group (#1622)

  

  masahi

  authored and

  tqchen

  committed
  b00aabc

  Copy full SHA for b00aabc

• [NODEREF] Introduce named attribute system. (#1618)

  tqchen

  authored
  20c495e

  Copy full SHA for 20c495e

• [NNVM][DARKNET]Yolo and Upsample frontend support (#1501)

  Show description for 4dc21bd
  

  PariksheetPinjari909

  authored and

  tqchen

  committed
  4dc21bd

  Copy full SHA for 4dc21bd

• Improve x86 Inception (#1506)

  Show description for 5d533ec
  

  kevinthesun

  authored and

  yzhliu

  committed
  5d533ec

  Copy full SHA for 5d533ec

• [VERSION] Update to 0.5.dev (#1623)

  Show description for 7cb85d8
  tqchen

  authored
  7cb85d8

  Copy full SHA for 7cb85d8

Commits on Aug 21, 2018

• Add int8 gemm recipe (#1614)

  

  vinx13

  authored and

  tqchen

  committed
  21e1301

  Copy full SHA for 21e1301

• [RUNTIME] Add TypedPackedFunc (#1626)

  tqchen

  authored
  16d3c1f

  Copy full SHA for 16d3c1f

Commits on Aug 22, 2018

• check in (#1629)

  

  eqy

  authored and

  tqchen

  committed
  729224b

  Copy full SHA for 729224b

• [AUTOTVM] Allow fallback for template & Fix bugs in tuners (#1615)

  Show description for b7beb1e
  

  merrymercy

  authored and

  tqchen

  committed
  b7beb1e

  Copy full SHA for b7beb1e

• [TOPI][ARM CPU] fuse bias to depthwise conv2d (#1631)

  

  merrymercy

  authored and

  tqchen

  committed
  ca76810

  Copy full SHA for ca76810

• [FRONTEND]minor bug fixes (#1632)

  

  siju-samuel

  authored and

  tqchen

  committed
  d060e91

  Copy full SHA for d060e91

• [ATTRS] change AttrFiledInfo->Node (#1634)

  tqchen

  authored
  b11f2a0

  Copy full SHA for b11f2a0

• [NNVM][KERAS] Add cropping support (#1636)

  

  kazum

  authored and

  tqchen

  committed
  d90c1e4

  Copy full SHA for d90c1e4

• [NNVM][KERAS] Fixed padding in pooling (#1635)

  

  kazum

  authored and

  tqchen

  committed
  58b2395

  Copy full SHA for 58b2395

Commits on Aug 23, 2018

• [TEAM] New reviewer: nishi-t (#1637)

  tqchen

  authored
  0003856

  Copy full SHA for 0003856

• [FRONTEND][COREML]More ops are added (#1619)

  

  siju-samuel

  authored and

  tqchen

  committed
  c98ba60

  Copy full SHA for c98ba60

• trigger ci (#1620)

  

  merrymercy

  authored and

  tqchen

  committed
  56c50d2

  Copy full SHA for 56c50d2

• [AUTOTVM] Simplify TopHub (#1630)

  

  merrymercy

  authored and

  tqchen

  committed
  cfafd21

  Copy full SHA for cfafd21

• Remove leading "./" from include paths (#1640)

  

  kazum

  authored and

  tqchen

  committed
  b95b595

  Copy full SHA for b95b595

• [RUNTIME][PYTHON] Switch to use __new__ for constructing node. (#1644)

  tqchen

  authored
  56ab0ad

  Copy full SHA for 56ab0ad

• fix CO CI problem (#1641)

  

  merrymercy

  authored and

  tqchen

  committed
  7afe6ba

  Copy full SHA for 7afe6ba

• [RUNTIME] [OPENCL] Fix access modifiers (#1643)

  

  imorinaga

  authored and

  tqchen

  committed
  3dbc0d6

  Copy full SHA for 3dbc0d6

• [FIX] Fix issue with TypedPackedFunc template instatition (#1649)

  

  jroesch

  authored and

  tqchen

  committed
  e336544

  Copy full SHA for e336544

• [NNVM][KERAS] Support multiple outputs (#1648)

  

  kazum

  authored and

  tqchen

  committed
  62d34ca

  Copy full SHA for 62d34ca

• [PYTHON] Enable constructors in Node (#1647)

  tqchen

  authored
  6eecec9

  Copy full SHA for 6eecec9

• [NNVM][TEST] Test against numerical grad (#1505)

  Show description for 0edf87e
  

  sgrechanik-h

  authored and

  yzhliu

  committed
  0edf87e

  Copy full SHA for 0edf87e

Commits on Aug 24, 2018

• [AUTOTVM] Fix local executor (#1651)

  Show description for 771d895
  

  merrymercy

  authored and

  tqchen

  committed
  771d895

  Copy full SHA for 771d895

• [NODE] Enable global singleton object, allow set_body_typed in function registry, default fallback of IRPrinter. (#1652)

  tqchen

  authored
  531efd6

  Copy full SHA for 531efd6

• improve text summary (#1655)

  

  merrymercy

  authored and

  tqchen

  committed
  a168899

  Copy full SHA for a168899

Commits on Aug 25, 2018

• [LANG] Improve serializer (#1658)

  tqchen

  authored
  cfc848a

  Copy full SHA for cfc848a

• [RUNTIME][OPENCL] delay device check (#1657)

  

  kazum

  authored and

  tqchen

  committed
  37988ad

  Copy full SHA for 37988ad

• [CODEGEN][AOCL] Add math intrinsic rules (#1653)

  Show description for a03c60b
  

  kazum

  authored and

  tmoreau89

  committed
  a03c60b

  Copy full SHA for a03c60b

Commits on Aug 26, 2018

• [FRONTEND][ONNX]HardSigmoid, min, max, mean ops support (#1645)

  

  siju-samuel

  authored and

  yzhliu

  committed
  a9e0567

  Copy full SHA for a9e0567

Commits on Aug 27, 2018

• [FRONTEND][TENSORFLOW] fix the convertion of sum and add testcase for it (#1654)

  Show description for 463e5c3
  

  nishi-t

  authored and

  yzhliu

  committed
  463e5c3

  Copy full SHA for 463e5c3

• add docstring skip in hybrid script (#1668)

  Show description for 1d7c52f
  

  were

  authored and

  yzhliu

  committed
  1d7c52f

  Copy full SHA for 1d7c52f

Commits on Aug 28, 2018

• [DOCS][NNVM] Delete duplicated tensor operators from list (#1669)

  

  imorinaga

  authored and

  tqchen

  committed
  ea01417

  Copy full SHA for ea01417

• Fix incorrect stride in conv2d_nhwc_python (#1670)

  

  vinx13

  authored and

  yzhliu

  committed
  38203a8

  Copy full SHA for 38203a8

Commits on Aug 29, 2018

• [AUTOTVM] Decouple build and run in measurement (#1661)

  

  merrymercy

  authored and

  tqchen

  committed
  12839e6

  Copy full SHA for 12839e6

• [TOPI] add nn schedulers for HLS backends (#1663)

  Show description for 7c4dd0b
  

  kazum

  authored and

  tmoreau89

  committed
  7c4dd0b

  Copy full SHA for 7c4dd0b

Commits on Aug 30, 2018

• Fix incorrect doc in conv2d_nhwc_python (#1677)

  

  vinx13

  authored and

  yzhliu

  committed
  9f99a4f

  Copy full SHA for 9f99a4f

• [PASS] Enhance gpu verify pass (#1660)

  

  merrymercy

  authored and

  tqchen

  committed
  0c52378

  Copy full SHA for 0c52378