Commit History

  1. Woops

    thomasw21
    0199675
  2. Victor suggestion to send the last batch

    thomasw21
    e4b750b

  1. Get PMD running

    thomasw21
    44f5a1f

  1. Debug

    thomasw21
    20f4f97
  2. I'm stupid

    thomasw21
    fe105ca
  3. More debug

    thomasw21
    702ea76
  4. Woops

    thomasw21
    b5cd8c9
  5. debug

    thomasw21
    ced55ee
  6. remove default logs

    thomasw21
    6434d34
  7. Woops

    thomasw21
    8b87149
  8. Run everything in parallel except sqlite3

    thomasw21
    cbdd0ca

  1. Logs

    thomasw21
    8c55fa4
  2. Woops

    thomasw21
    66052c6
  3. Woops

    thomasw21
    4d17ec4
  4. Woops

    thomasw21
    191bcd5
  5. Test out arrowBuilder

    thomasw21
    622d8d0
  6. Debug proc

    thomasw21
    da3e34c
  7. Debug low throughput

    thomasw21
    1d9efb0
  8. Maybe I need to load everything into memory first

    thomasw21
    1a6bd7c
  9. Use a lot bigger batch size

    thomasw21
    69b5d09
  10. Woops

    thomasw21
    227ad5d
  11. Perhaps that's better

    thomasw21
    ab01e28
  12. Define a batch_size_per_proc to lower memory footprint, there's going to be a natural tradeoff

    thomasw21
    f6d6453
  13. Reduce chunk size

    thomasw21
    e6583b9
  14. Maybe fix

    thomasw21
    004daf7
  15. Add support for multiprocessing for visual genome

    thomasw21
    59af408

  1. Add back all configs

    thomasw21
    3802b20
  2. Remove prints

    thomasw21
    af91e86
  3. missing

    thomasw21
    113d6ab
  4. Debug the filtering of yfcc100m

    thomasw21
    8bf249a
  5. Test yfcc100m

    thomasw21
    a43f5e6
  6. Being able to seriablize

    thomasw21
    3e03859
  7. Debug conceptual_12m

    thomasw21
    09b5b06
  8. PMD dataset

    thomasw21
    1d83a66
  9. initial commit

    Thomas Wang
    cb26dfd