Commits

Breadcrumbs

History for

onmaster

User selector

All users

Datepicker

All time

Commit History

Commits on Jan 4, 2018

• Fix a bug in webgl storage and fix demos (tensorflow#505)

  Show description for 4b66a21
  dsmilkov

  authored
  4b66a21

  Copy full SHA for 4b66a21

Commits on Jan 2, 2018

• Make NDArray<_, R> assignable to Scalar/ArrayRD (tensorflow#501)

  dsmilkov

  authored
  bd3fa1a

  Copy full SHA for bd3fa1a

Commits on Dec 31, 2017

• Force matching dtypes for ops that take 2 inputs, to be compatible with TensorFlow (tensorflow#492)

  Show description for d539fd5
  dsmilkov

  authored
  d539fd5

  Copy full SHA for d539fd5

• Add math.vjp, custom gradients, softmax gradients, softmaxCrossEntropyWithLogits + gradients. (tensorflow#486)

  Show description for 8fff915
  

  Nikhil Thorat

  authored
  8fff915

  Copy full SHA for 8fff915

Commits on Dec 30, 2017

• update version

  dsmilkov

  committed
  14808c4

  Copy full SHA for 14808c4

• Add math.maximum and math.minimum (tensorflow#487)

  Show description for 42f2b99
  dsmilkov

  authored
  42f2b99

  Copy full SHA for 42f2b99

Commits on Dec 28, 2017

• Support negative axis indexing for axis param in reduction ops (tensorflow#485)

  Show description for 4b157c7
  dsmilkov

  authored
  4b157c7

  Copy full SHA for 4b157c7

• Add backprop for Softmax (Graph API) (tensorflow#427)

  

  vuoristo

  authored and

  Nikhil Thorat

  committed
  3f8cd66

  Copy full SHA for 3f8cd66

• Add rank as generic type and introduce Variable (updateble NDArray) (tensorflow#478)

  Show description for 403d48d
  dsmilkov

  authored
  403d48d

  Copy full SHA for 403d48d

• Cleanup usage of getValues() and pass NDArray to test_util.expectArraysClose (tensorflow#474)

  Show description for 32b190a
  dsmilkov

  authored
  32b190a

  Copy full SHA for 32b190a

• Add batchNormalization4D to NDArrayMath (tensorflow#465)

  

  caisq

  authored and

  dsmilkov

  committed
  af05e39

  Copy full SHA for af05e39

Commits on Dec 27, 2017

• Update versions of core library and models/* (tensorflow#473)

  dsmilkov

  authored
  10a76c8

  Copy full SHA for 10a76c8

Commits on Dec 26, 2017

• Add delayed upload mode in MathBackendWebGL (tensorflow#466)

  Show description for 228e0ee
  dsmilkov

  authored
  228e0ee

  Copy full SHA for 228e0ee

• Add gradients and valueAndGradients. Remove gradientWrt on the public math API. (tensorflow#462)

  Show description for ca9f610
  

  Nikhil Thorat

  authored
  ca9f610

  Copy full SHA for ca9f610

Commits on Dec 23, 2017

• Fix demos and models (tensorflow#461)

  Show description for 0555416
  dsmilkov

  authored
  0555416

  Copy full SHA for 0555416

Commits on Dec 21, 2017

• Consolidate Tape logic and scope logic. (tensorflow#448)

  Show description for a9003c5
  

  Nikhil Thorat

  authored
  a9003c5

  Copy full SHA for a9003c5

• update version

  dsmilkov

  committed
  561e5e5

  Copy full SHA for 561e5e5

• Fix bugs introduced by global tracking (tensorflow#460)

  Show description for ae6dd14
  dsmilkov

  authored
  ae6dd14

  Copy full SHA for ae6dd14

Commits on Dec 20, 2017

• Add PReLU math backend function and differentiable graph operation. (tensorflow#453)

  

  wang2bo2

  authored and

  Nikhil Thorat

  committed
  ba7d597

  Copy full SHA for ba7d597

• Add the high level norm op (tensorflow#442)

  Show description for 673afff
  pyu10055

  authored
  673afff

  Copy full SHA for 673afff

Commits on Dec 19, 2017

• Add generic broadcasting support for CPU backend binary ops. (tensorflow#458)

  

  Nikhil Thorat

  authored
  cd65b71

  Copy full SHA for cd65b71

Commits on Dec 15, 2017

• Improve backend registration and move it as part of the global ENV (tensorflow#444)

  dsmilkov

  authored
  462beaf

  Copy full SHA for 462beaf

• Add a Tape class and a gradientWrt method. (tensorflow#436)

  Show description for 8d1b41d
  

  Nikhil Thorat

  authored
  8d1b41d

  Copy full SHA for 8d1b41d

Commits on Dec 12, 2017

• Update to 0.3.14 (tensorflow#435)

  

  Nikhil Thorat

  authored
  10de689

  Copy full SHA for 10de689

• 2 standard deviations from the mean is (std * 2 +- mean). (tensorflow#434)

  Show description for 0c00d5e
  nkreeger

  authored
  0c00d5e

  Copy full SHA for 0c00d5e

• When calling NDArray.fill, write to the NDArrayStorage backend. (tensorflow#433)

  

  Nikhil Thorat

  authored
  b711ac8

  Copy full SHA for b711ac8

• Handle error in case of 404 not found variable (tensorflow#431)

  Show description for bb43201
  

  Lewuathe

  authored and

  Nikhil Thorat

  committed
  bb43201

  Copy full SHA for bb43201

• Decouple NDArray storage from NDArray and make math global (tensorflow#404)

  Show description for d055e92
  dsmilkov

  authored
  d055e92

  Copy full SHA for d055e92

Commits on Dec 11, 2017

• Add kernel input configurations as the interface between Math and Backends via a BackendEngine. (tensorflow#426)

  Show description for b934f06
  

  Nikhil Thorat

  authored
  b934f06

  Copy full SHA for b934f06

• Add the pow operator (tensorflow#425)

  Show description for 35034fa
  pyu10055

  authored
  35034fa

  Copy full SHA for 35034fa

Commits on Dec 7, 2017

• Add math.conv1d (tensorflow#368)

  

  easadler

  authored and

  dsmilkov

  committed
  22b2003

  Copy full SHA for 22b2003

Commits on Dec 6, 2017

• Lint cleanup + README Update (tensorflow#422)

  Show description for 4b0708a
  nkreeger

  authored
  4b0708a

  Copy full SHA for 4b0708a

• Use correct Array type instead of casting. (tensorflow#421)

  nkreeger

  authored
  1a9b611

  Copy full SHA for 1a9b611

• Add ones() and onesLike() to NDArray* APIs (tensorflow#420)

  nkreeger

  authored
  163f12e

  Copy full SHA for 163f12e

Commits on Dec 4, 2017

• Remove unnecessary scope in activation functions (tensorflow#369)

  

  Lewuathe

  authored and

  Nikhil Thorat

  committed
  d513f5f

  Copy full SHA for d513f5f

Pagination

Previous

Next