Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 7, 2020

• Handle case where input is longer than max_seq_len in padding (#1479)

  Show description for 9e75b15
  

  Michael Gschwind

  authored and

  facebook-github-bot

  committed
  9e75b15

  Copy full SHA for 9e75b15

Commits on Oct 6, 2020

• Config adapter for padding_control => seq_padding_control (#1478)

  Show description for 3c1e1a2
  

  Michael Gschwind

  authored and

  facebook-github-bot

  committed
  3c1e1a2

  Copy full SHA for 3c1e1a2

Commits on Oct 2, 2020

• Add batch length padding for fixed function accelerators (#1457)

  Show description for 033f262
  

  Michael Gschwind

  authored and

  facebook-github-bot

  committed
  033f262

  Copy full SHA for 033f262

• support manifold for caffe2 model loading (#1476)

  Show description for ebd9547
  

  hudeven

  authored and

  facebook-github-bot

  committed
  ebd9547

  Copy full SHA for ebd9547

• Adopt TorchText SPM as default PyText tokenizer (#1477)

  Show description for ce6e262
  

  JiesiZhao077

  authored and

  facebook-github-bot

  committed
  ce6e262

  Copy full SHA for ce6e262

Commits on Oct 1, 2020

• Make right_dense and left_dense attribute Optional since default value is None

  Show description for 7ea2cf7
  

  Xiangge Meng

  authored and

  facebook-github-bot

  committed
  7ea2cf7

  Copy full SHA for 7ea2cf7

• Add SPM file path to URL (#1475)

  Show description for 7357d48
  

  JiesiZhao077

  authored and

  facebook-github-bot

  committed
  7357d48

  Copy full SHA for 7357d48

• Replace control flow with equivalent logic so that it's fx-able

  Show description for e5ace58
  

  dskhudia

  authored and

  facebook-github-bot

  committed
  e5ace58

  Copy full SHA for e5ace58

• Add support for Knowledge Distilling RoBERTa Q&A models.

  Show description for 2670221
  

  Debojeet Chatterjee

  authored and

  facebook-github-bot

  committed
  2670221

  Copy full SHA for 2670221

• Fix the workflow for the default value of `output_columns`.

  Show description for 68f2c6a
  

  Debojeet Chatterjee

  authored and

  facebook-github-bot

  committed
  68f2c6a

  Copy full SHA for 68f2c6a

Commits on Sep 30, 2020

• Back out "Use TorchText SPM in PyText"

  Show description for 8472355
  

  JiesiZhao077

  authored and

  facebook-github-bot

  committed
  8472355

  Copy full SHA for 8472355

Commits on Sep 29, 2020

• Use TorchText SPM in PyText (#1456)

  Show description for 1317d05
  

  JiesiZhao077

  authored and

  facebook-github-bot

  committed
  1317d05

  Copy full SHA for 1317d05

Commits on Sep 28, 2020

• generic doc classification task with hydra (#1473)

  Show description for 3ec01ff
  

  hudeven

  authored and

  facebook-github-bot

  committed
  3ec01ff

  Copy full SHA for 3ec01ff

• pass vocab to DocModel

  Show description for b450665
  

  hudeven

  authored and

  facebook-github-bot

  committed
  b450665

  Copy full SHA for b450665

• add doc classification data module (#1471)

  Show description for e0be17f
  

  hudeven

  authored and

  facebook-github-bot

  committed
  e0be17f

  Copy full SHA for e0be17f

• add DocTransform (#1470)

  Show description for 4576822
  

  hudeven

  authored and

  facebook-github-bot

  committed
  4576822

  Copy full SHA for 4576822

• merge transform and collate_fn (#1469)

  Show description for 020a3eb
  

  hudeven

  authored and

  facebook-github-bot

  committed
  020a3eb

  Copy full SHA for 020a3eb

• Support exporting embedding for two tower document classifier (#1472)

  Show description for 84267ca
  

  HannaMao

  authored and

  facebook-github-bot

  committed
  84267ca

  Copy full SHA for 84267ca

Commits on Sep 25, 2020

• disable dual writing, save models to Manifold (#1455)

  Show description for a5052de
  

  hudeven

  authored and

  facebook-github-bot

  committed
  a5052de

  Copy full SHA for a5052de

• Tests for to_omega_conf (#1465)

  Show description for 81ec375
  

  snisarg

  authored and

  facebook-github-bot

  committed
  81ec375

  Copy full SHA for 81ec375

• replace Transform and ScriptTransform with nn.Module (#1468)

  Show description for ba81ca4
  

  hudeven

  authored and

  facebook-github-bot

  committed
  ba81ca4

  Copy full SHA for ba81ca4

• move datasets under data (#1466)

  Show description for 553c40d
  

  hudeven

  authored and

  facebook-github-bot

  committed
  553c40d

  Copy full SHA for 553c40d

• simplify doc_model with primitive args (#1467)

  Show description for 0f99643
  

  hudeven

  authored and

  facebook-github-bot

  committed
  0f99643

  Copy full SHA for 0f99643

Commits on Sep 24, 2020

• XLMR uses GELU. (#1463)

  Show description for 2f7d2e6
  

  snisarg

  authored and

  facebook-github-bot

  committed
  2f7d2e6

  Copy full SHA for 2f7d2e6

• TruncateTransform (#1462)

  Show description for fad6ea6
  

  snisarg

  authored and

  facebook-github-bot

  committed
  fad6ea6

  Copy full SHA for fad6ea6

Commits on Sep 23, 2020

• PyText's metric reporting (#1464)

  Show description for ac2c8e0
  

  snisarg

  authored and

  facebook-github-bot

  committed
  ac2c8e0

  Copy full SHA for ac2c8e0

Commits on Sep 22, 2020

• Check that random seed is set for pytext distributed training

  Show description for 2b1f6ed
  

  Michael Wu

  authored and

  facebook-github-bot

  committed
  2b1f6ed

  Copy full SHA for 2b1f6ed

Commits on Sep 21, 2020

• Change dependency from pytorch-dp to opacus

  Show description for 8e38443
  

  Darktex

  authored and

  facebook-github-bot

  committed
  8e38443

  Copy full SHA for 8e38443

Commits on Sep 19, 2020

• Fix our forward function

  Show description for 14226fb
  

  snisarg

  authored and

  facebook-github-bot

  committed
  14226fb

  Copy full SHA for 14226fb

Commits on Sep 18, 2020

• Decode for decoupled tokenizer (#1461)

  Show description for 347f453
  

  Akshat Shrivastava

  authored and

  facebook-github-bot

  committed
  347f453

  Copy full SHA for 347f453

Commits on Sep 17, 2020

• Fix Fairseq FP16 Optimizer (#1460)

  Show description for 3a55d55
  

  Akshat Shrivastava

  authored and

  facebook-github-bot

  committed
  3a55d55

  Copy full SHA for 3a55d55

Commits on Sep 16, 2020

• Add AssistantXLMDocModel that Implements non-batched inference

  Show description for 81ab44e
  

  abhinavarora

  authored and

  facebook-github-bot

  committed
  81ab44e

  Copy full SHA for 81ab44e

Commits on Sep 15, 2020

• Enable Graph Mode Quantization For Linformer

  Show description for 38f09f7
  

  Qing Yu

  authored and

  facebook-github-bot

  committed
  38f09f7

  Copy full SHA for 38f09f7

Commits on Sep 14, 2020

• MLPDecoder test-time temperature scaling (#1458)

  Show description for f790ece
  

  shreydesai

  authored and

  facebook-github-bot

  committed
  f790ece

  Copy full SHA for f790ece

Commits on Sep 11, 2020

• manual data sharding in distributed training (#1434)

  Show description for 3586c8b
  

  hudeven

  authored and

  facebook-github-bot

  committed
  3586c8b

  Copy full SHA for 3586c8b

Pagination

Previous

Next