twitter-archive/torch-dataset

Issues

• module 'libdataset' not found

  #18 opened 5 years ago by willfrey
  5

• walk through index.csv file in order?

  #13 opened 5 years ago by willfrey
  2

• not found:No LuaRocks module found for dataset.StreamedDataset

  #29 opened 8 years ago by chienlinhuang1116
  4

• IndexCSV and IndexDirectory should have the ability to pre-shuffle and do deterministic linear sample with partitions

  #9 opened 8 years ago by zakattacktwitter
  1

• Cannot handle inputDims when using IndexDirectory.lua, it will load the whole file

  #35 opened 8 years ago by chienlinhuang1116
  0

• How to use SlowFS?

  #30 opened 8 years ago by chienlinhuang1116
  6

• How can I get the FloatTensor values instead of the binary content when using IndexDirectory.lua ?

  #34 opened 8 years ago by chienlinhuang1116
  2

• What is the difference between 'viewfs://' and 'slowfs://' ?

  #32 opened 8 years ago by chienlinhuang1116
  1

• Can we open files by "FloatTensor" instead of "io.popen" when testing "test_StreamedDataset.lua" ?

  #33 opened 8 years ago by chienlinhuang1116
  1

• Why "SlowFS.find(URL)" returns "nil"?

  #31 opened 8 years ago by chienlinhuang1116
  1

• How to read data using torch-dataset?

  #25 opened 8 years ago by chienlinhuang1116
  2

• The number of batchSize cannot be defined larger than 256 in “torch-dataset”.

  #24 opened 8 years ago by chienlinhuang1116
  2

• Best way to load dataset with large batchsize

  #27 opened 8 years ago by mkvb01
  4

• Clean up the README.

  #2 opened 8 years ago by zakattacktwitter
  0

• Provide & document a clean way to integrate a SlowFS.

  #3 opened 8 years ago by zakattacktwitter
  0

• Provide an S3 implementation of SlowFS and an example.

  #4 opened 8 years ago by zakattacktwitter
  0

• Cache does not handle partial imports well

  #11 opened 8 years ago by zakattacktwitter
  0

• Cache needs a proper eviction implementation.

  #1 opened 8 years ago by zakattacktwitter
  0

• Got stuck in getBatch with larger batch size

  #21 opened 8 years ago by joeyhng
  7

• Calling Torch library within Processor generates Seg Fault

  #23 opened 8 years ago by JimmyWhitaker
  3

• Is there any limitation in the memory when using torch-dataset?

  #26 opened 8 years ago by chienlinhuang1116
  1

• A bug may exist when tensors is the input to Dataset

  #20 opened 8 years ago by jundengdeng
  1

• MNIST example error

  #19 opened 8 years ago by vcodreanu
  2

• IndexCSV Datasets are not partitioned correctly

  #16 opened 8 years ago by willfrey
  2

• IndexCSV Datasets are not partitioned correctly

  #17 opened 8 years ago by willfrey
  0

• Support for require 'audio'

  #12 opened 8 years ago by willfrey
  22

• `opt.inputDims` is a required argument but is not enforced

  #14 opened 8 years ago by vzhong
  1

• Grouped partitions

  #10 opened 8 years ago by clementfarabet
  2

• Shuffle cifar10/training.csv and use a linear sampler

  #8 opened 8 years ago by zakattacktwitter
  0

• HTTP getter is really slow, even from CloudFront

  #7 opened 8 years ago by zakattacktwitter
  0

• need to bring luasocket fixes out into the user supplied create

  #6 opened 8 years ago by zakattacktwitter
  0

• Fix the test_Getters.lua to use public URLs.

  #5 opened 8 years ago by zakattacktwitter
  0