Commits

Branch selector

 master

User selector

zdevito

Datepicker

All time

Commit History

Commits on Aug 30, 2017

• Add gloo submodule.

  Show description for 7310ebb
  

  zdevito

  authored and

  apaszke

  committed
  7310ebb

  Copy full SHA for 7310ebb

• Remove gloo subtree, in preparation for gloo submodule addition.

  

  zdevito

  authored and

  apaszke

  committed
  aaef3a3

  Copy full SHA for aaef3a3

Commits on Jul 29, 2017

• Remove dead THPP code that has been replaced with ATen objects. (pytorch#2235)

  Show description for 43c944a
  

  zdevito

  authored and

  soumith

  committed
  43c944a

  Copy full SHA for 43c944a

• fix a bug where an uninitialized at::Tensor was passed to createPyObject (pytorch#2239)

  

  zdevito

  authored and

  soumith

  committed
  bf26a51

  Copy full SHA for bf26a51

Commits on Jul 26, 2017

• allow retain to be specified for unsafeTensorFromTH

  zdevito

  committed
  fd97d92

  Copy full SHA for fd97d92

• Merge pull request pytorch#35 from ezyang/pr/undefined-dim-doc

  Show description for be9ef92
  zdevito

  authored
  be9ef92

  Copy full SHA for be9ef92

• fix osx build errors related to long/int64_t

  zdevito

  committed
  9c0d52a

  Copy full SHA for 9c0d52a

Commits on Jul 25, 2017

• fix a bug where some scalars were getting truncated to integers incorrectly.

  zdevito

  committed
  ef3b09f

  Copy full SHA for ef3b09f

Commits on Jul 20, 2017

• return a sentinel value when THTensor has undefined dimensions.

  zdevito

  committed
  3719b42

  Copy full SHA for 3719b42

Commits on Jul 18, 2017

• Merge pull request pytorch#20 from killeent/type-equality

  Show description for 7914d67
  zdevito

  authored
  7914d67

  Copy full SHA for 7914d67

Commits on Jul 14, 2017

• produce a Declarations.yaml file that describes Functions/Type/Tensor methods that framework produced.

  zdevito

  committed
  83596bd

  Copy full SHA for 83596bd

• allow type inference to work on TensorList

  zdevito

  committed
  128e02d

  Copy full SHA for 128e02d

Commits on Jul 13, 2017

• fix bug in method declarations

  zdevito

  committed
  9e720f1

  Copy full SHA for 9e720f1

• write generated_cpp. to a file rather than as output to make error reporting clearer.

  zdevito

  committed
  007d6ad

  Copy full SHA for 007d6ad

• Merge pull request pytorch#11 from soumith/master

  Show description for b900a49
  zdevito

  authored
  b900a49

  Copy full SHA for b900a49

Commits on Jul 11, 2017

• add DynamicType variants for ATen functions.

  

  zdevito

  authored and

  soumith

  committed
  2ecb188

  Copy full SHA for 2ecb188

• initialize aten and pytorch to share the same THCState

  

  zdevito

  authored and

  soumith

  committed
  9d8cff9

  Copy full SHA for 9d8cff9

• add build commands for ATen

  

  zdevito

  authored and

  soumith

  committed
  ab3d85c

  Copy full SHA for ab3d85c

Commits on Jul 10, 2017

• always use a custom default float

  zdevito

  committed
  9d0c674

  Copy full SHA for 9d0c674

• python style fixes

  zdevito

  committed
  bff762c

  Copy full SHA for bff762c

• remove inaccessible median variant (pytorch#2015)

  Show description for 8b00356
  

  zdevito

  authored and

  soumith

  committed
  8b00356

  Copy full SHA for 8b00356

• support unsafe functions for getting/constructor tensors from TH objects for backward compat.

  zdevito

  committed
  581921f

  Copy full SHA for 581921f

Commits on Jul 8, 2017

• lazily initialize cuda so that we behave similar to PyTorch

  zdevito

  committed
  f0788af

  Copy full SHA for f0788af

Commits on Jul 7, 2017

• osx build issues and clang warnings

  zdevito

  committed
  a4dc7dc

  Copy full SHA for a4dc7dc

• remove Sparse from dispatch for now, will add dispatch variants later

  zdevito

  committed
  5dd05ed

  Copy full SHA for 5dd05ed

• fix build issue when cuda does not exist

  zdevito

  committed
  4fda678

  Copy full SHA for 4fda678

Commits on Jul 3, 2017

• bind THS THCS, leaving all operators unimplemented. This is required because THPP can represent Sparse tensors even though the wrapper doesn't implement any operators.

  zdevito

  committed
  ec8da55

  Copy full SHA for ec8da55

Commits on Jun 23, 2017

• enable warnings in build and fix warnings

  zdevito

  committed
  a9ea975

  Copy full SHA for a9ea975

• update readme and add assign_(Scalar) variant

  zdevito

  committed
  b1a84e3

  Copy full SHA for b1a84e3

• fix a bug with scalar handling by simplifiying the maybeScalar check.

  zdevito

  committed
  66d93b6

  Copy full SHA for 66d93b6

• handle select and operator[] style operations

  zdevito

  committed
  2af6ba3

  Copy full SHA for 2af6ba3

• add checks for scalars on output

  zdevito

  committed
  b59b44f

  Copy full SHA for b59b44f

• start adding rules to propagate scalar to results

  zdevito

  committed
  a10a1c9

  Copy full SHA for a10a1c9

• Scalar objects can now be backed by 0-dim Tensors.

  zdevito

  committed
  bb6908e

  Copy full SHA for bb6908e

• missing fixed allocator files

  zdevito

  committed
  c555cd8

  Copy full SHA for c555cd8

Pagination

Previous

Next