Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 28, 2018

• state_dict checkpoint before tensor storage serialization part

  motus

  committed
  69ac603

  Copy full SHA for 69ac603

• added references to state_dict and libtorch pages from the root README

  motus

  committed
  6869565

  Copy full SHA for 6869565

• done with the Python part of model serialization tutorial

  motus

  committed
  6976b57

  Copy full SHA for 6976b57

• Wrote the state_dict tutorial up to the load_state_dict() part. Did not go into details of tensor serialization in C etc.

  motus

  committed
  fabfeee

  Copy full SHA for fabfeee

• rename the saved mnist model to .pt

  motus

  committed
  a6658bb

  Copy full SHA for a6658bb

Commits on Nov 27, 2018

• not real work on backprop yet - will do serialization first

  motus

  committed
  e7fd534

  Copy full SHA for e7fd534

• started working on state_dict tutorial; added saved model binary for mnist

  motus

  committed
  c2050b5

  Copy full SHA for c2050b5

• started section on Python integration

  motus

  committed
  0e50990

  Copy full SHA for 0e50990

• started a notebook on torch state_dict and serialization

  motus

  committed
  587e307

  Copy full SHA for 587e307

Commits on Nov 26, 2018

• few typos some minor fixes

  motus

  committed
  e1bef0e

  Copy full SHA for e1bef0e

• add section on code generation and metadata

  motus

  committed
  ff24918

  Copy full SHA for ff24918

Commits on Nov 22, 2018

• wrote about Tensor and TensorImpl and Type; need to finish the code

  Show description for 67597d1
  motus

  committed
  67597d1

  Copy full SHA for 67597d1

Commits on Nov 20, 2018

• started documenting at::Torch and underlying classes

  motus

  committed
  f5a2862

  Copy full SHA for f5a2862

Commits on Nov 14, 2018

• fix the file permissions

  motus

  committed
  df5f147

  Copy full SHA for df5f147

• started a document on low-level pytorch api; remove config and license

  Show description for 0717b4f
  motus

  committed
  0717b4f

  Copy full SHA for 0717b4f

Commits on Nov 2, 2018

• Set theme jekyll-theme-minimal

  motus

  committed
  ac8f733

  Copy full SHA for ac8f733

Commits on Nov 1, 2018

• document all files (i.e. *.hpp _and_ *.h headers)

  motus

  committed
  4e674ed

  Copy full SHA for 4e674ed

Commits on Oct 31, 2018

• add more filters for TH_CONCAT_4 macros

  motus

  committed
  c5bc525

  Copy full SHA for c5bc525

• config to generate doxygen docs of (py|lib)torch internals

  motus

  committed
  a1d6754

  Copy full SHA for a1d6754

Commits on Oct 28, 2018

• added some matrix autodiff

  motus

  committed
  aed59ee

  Copy full SHA for aed59ee

• another take on backpropagation - starting with scalar multiplication and .backward()

  motus

  committed
  fe90934

  Copy full SHA for fe90934

Commits on Oct 27, 2018

• roll back to some smaller autodiff examples. before adding text

  motus

  committed
  d10979e

  Copy full SHA for d10979e

Commits on Oct 26, 2018

• no significant chages - playing with autograd

  motus

  committed
  5340e40

  Copy full SHA for 5340e40

Commits on Oct 25, 2018

• minor fixes to autorgrad. need to write about matrix derivatives

  motus

  committed
  1e42359

  Copy full SHA for 1e42359

• istarted the backprop section

  motus

  committed
  3c58ab6

  Copy full SHA for 3c58ab6

Commits on Oct 24, 2018

• started grad_fn

  motus

  committed
  7d63d2b

  Copy full SHA for 7d63d2b

• started a chapter on backpropagation

  motus

  committed
  13e4784

  Copy full SHA for 13e4784

Commits on Oct 23, 2018

• add reference to the little schemer in the intro

  motus

  committed
  6791716

  Copy full SHA for 6791716

• add TOC to section 1 of MNIST example

  motus

  committed
  a0b9452

  Copy full SHA for a0b9452

• reordered the dropout text; added section on softmax the conclusion.

  Show description for 3b2b878
  motus

  committed
  3b2b878

  Copy full SHA for 3b2b878

• finished sections on Linear unit and dropout(); minor fixes in TOC and

  Show description for abdccda
  motus

  committed
  abdccda

  Copy full SHA for abdccda

• much better text on max pooling; finished section on data dimensions; just before writing the chapter on linear unit

  motus

  committed
  47c6ad7

  Copy full SHA for 47c6ad7

• minor fixes to maxpooling - still need to elaborate on kernel size and give better examples

  motus

  committed
  4c0afdd

  Copy full SHA for 4c0afdd

Commits on Oct 22, 2018

• autosave

  motus

  committed
  bb31400

  Copy full SHA for bb31400

• MNIST example: better TOC, started layer 3, fixes to dropout

  motus

  committed
  e8cb7c0

  Copy full SHA for e8cb7c0

Pagination

Previous

Next