Commits

Branch selector

 6260c9f

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 10, 2018

• adding elmo naively as a replacement

  bmccann

  committed
  6260c9f

  Copy full SHA for 6260c9f

• better to have tensorboard logging on as default

  bmccann

  committed
  3aecae0

  Copy full SHA for 3aecae0

Commits on Nov 28, 2018

• Remove CUDA from the command to build docker image

  

  daminisatya

  authored and

  bmccann

  committed
  6b31d8f

  Copy full SHA for 6b31d8f

Commits on Nov 17, 2018

• Update README.md

  bmccann

  authored
  39720d0

  Copy full SHA for 39720d0

• Update README.md

  bmccann

  authored
  42f1bad

  Copy full SHA for 42f1bad

• CoVe results

  bmccann

  authored
  ae8d64b

  Copy full SHA for ae8d64b

• better predict; better MQAN+CoVe

  bmccann

  committed
  fe58004

  Copy full SHA for fe58004

Commits on Oct 26, 2018

• repairing docker commands in readme

  bmccann

  authored
  9447f41

  Copy full SHA for 9447f41

Commits on Oct 25, 2018

• Update README.md

  bmccann

  authored
  72aeb19

  Copy full SHA for 72aeb19

• Fix embedding layer implementation

  Show description for c0e4f70
  

  todpole3

  authored and

  bmccann

  committed
  c0e4f70

  Copy full SHA for c0e4f70

• getting rid of .data in models/common.py

  bmccann

  committed
  4523bca

  Copy full SHA for 4523bca

• ptr-gen mem efficient dist; only trainable params

  bmccann

  committed
  2c837ea

  Copy full SHA for 2c837ea

• pg self_attended_context in greedy

  bmccann

  committed
  754d76f

  Copy full SHA for 754d76f

• Reducing train_batch_tokens; 10k OOMs on 0.4.1

  bmccann

  committed
  f835558

  Copy full SHA for f835558

• mv save out of val; add arg for no tensorboard logging

  bmccann

  committed
  38ce8ac

  Copy full SHA for 38ce8ac

• val loss item

  bmccann

  committed
  80408b9

  Copy full SHA for 80408b9

• updating torch03 images with cove

  bmccann

  committed
  741fb6f

  Copy full SHA for 741fb6f

• updated README re: docker images for devices and pytorch versions

  bmccann

  committed
  c5719e7

  Copy full SHA for c5719e7

• update devices

  bmccann

  committed
  2bcd810

  Copy full SHA for 2bcd810

• updating dockerfiles

  bmccann

  committed
  b120715

  Copy full SHA for b120715

• touching up predict

  bmccann

  committed
  eab80b6

  Copy full SHA for eab80b6

• mv to cpu for save; multinli.in.out regression

  bmccann

  committed
  fb0dcaa

  Copy full SHA for fb0dcaa

• bwds compat for old CoVe params; print metrics after preds

  bmccann

  committed
  a465784

  Copy full SHA for a465784

• rm Variable from sapg

  bmccann

  committed
  4a962fa

  Copy full SHA for 4a962fa

• rm Variable from pg

  bmccann

  committed
  ee3ba6c

  Copy full SHA for ee3ba6c

• rm Variable from cpg; move CoattentiveLayer to models/common

  bmccann

  committed
  59a1dcd

  Copy full SHA for 59a1dcd

• using torch 0.4 style device

  bmccann

  committed
  e958d2d

  Copy full SHA for e958d2d

• rm requires_grad from sapg

  bmccann

  committed
  d1416e6

  Copy full SHA for d1416e6

• rm requires_grad from pg

  bmccann

  committed
  b503c69

  Copy full SHA for b503c69

• rm requires_grad from cpg

  bmccann

  committed
  e18ea29

  Copy full SHA for e18ea29

• rm requires_grad from mqan

  bmccann

  committed
  a6abfc7

  Copy full SHA for a6abfc7

• update field for torch 0.4

  bmccann

  committed
  2838e23

  Copy full SHA for 2838e23

• adding optimizers to pretrained models; mv pretrained models to cpu

  bmccann

  committed
  b82b510

  Copy full SHA for b82b510

• torch.no_grad replacing volatile

  bmccann

  committed
  4372d44

  Copy full SHA for 4372d44

• attempt at migrating w/o Variables

  bmccann

  committed
  7630473

  Copy full SHA for 7630473

Pagination

Previous

Next