Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 16, 2024

• TorchAudio installation workaround for incorrect PYTORCH_VERSION env variable (NVIDIA#9736) (NVIDIA#9750)

  Show description for b39a487
  

  github-actions[bot]

  and

  artbataev

  authored
  b39a487

  Copy full SHA for b39a487

• make 'load_directly_on_device' configurable (NVIDIA#9657) (NVIDIA#9674)

  Show description for 2d87359
  

  3 peopleauthored
  2d87359

  Copy full SHA for 2d87359

• [NeMo-UX] Minor bug fix when TE/Apex not installed (NVIDIA#9749)

  Show description for 34bfe1b
  ashors1

  authored
  34bfe1b

  Copy full SHA for 34bfe1b

Commits on Jul 15, 2024

• [NeMo-UX] Make TE and Apex dependencies optional (NVIDIA#9732)

  Show description for f2e3232
  

  3 peopleauthored
  f2e3232

  Copy full SHA for f2e3232

• [Nemo-UX] Including all trainable-params in a PEFT-checkpoint (NVIDIA#9650) (NVIDIA#9691)

  Show description for 8128122
  

  6 peopleauthored
  8128122

  Copy full SHA for 8128122

• update pretrained model text (NVIDIA#9724) (NVIDIA#9745)

  Show description for c21e011
  

  github-actions[bot]

  and

  erastorgueva-nv

  authored
  c21e011

  Copy full SHA for c21e011

• add contianer (NVIDIA#9731)

  Show description for a9dbe37
  

  JRD971000

  and

  Ali Taghibakhshi

  authored
  a9dbe37

  Copy full SHA for a9dbe37

• [NeMo-UX] Fix imports so local configuration of runs works again (NVIDIA#9690) (NVIDIA#9694)

  Show description for b166a8f
  

  github-actions[bot]

  and

  marcromeyn

  authored
  b166a8f

  Copy full SHA for b166a8f

• Remove mask if use fusion mask (NVIDIA#9723)

  Show description for 27b5c47
  hsiehjackson

  authored
  27b5c47

  Copy full SHA for 27b5c47

• refactor: README (NVIDIA#9712)

  Show description for 02ff85b
  ko3n1g

  authored
  02ff85b

  Copy full SHA for 02ff85b

• [NeMo-UX] Fix when optimizers are setup for PEFT (NVIDIA#9619) (NVIDIA#9647)

  Show description for 8e9ef94
  

  4 peopleauthored
  8e9ef94

  Copy full SHA for 8e9ef94

• enables default data step in megatron parallel to operate on a wider variety of tensors - second try (NVIDIA#9671)

  Show description for f477ed1
  

  jomitchellnv

  and

  jstjohn

  authored
  f477ed1

  Copy full SHA for f477ed1

• fix legacy ds padding bug (NVIDIA#9716)

  Show description for 56f024d
  dimapihtar

  authored
  56f024d

  Copy full SHA for 56f024d

Commits on Jul 12, 2024

• refactor: Uniform BRANCH for notebooks (NVIDIA#9710)

  Show description for 599b60f
  ko3n1g

  authored
  599b60f

  Copy full SHA for 599b60f

• Allows non-strict load with distributed checkpoints (NVIDIA#9613) (NVIDIA#9715)

  Show description for 081a163
  

  github-actions[bot]

  and

  mikolajblaz

  authored
  081a163

  Copy full SHA for 081a163

• Adding support for mcore T5 Eval - SFT - PEFT (NVIDIA#9679)

  Show description for 3e2bb21
  

  huvunvidia

  and

  Huy Vu2

  authored
  3e2bb21

  Copy full SHA for 3e2bb21

Commits on Jul 11, 2024

• [TTS] Add fullband mel codec checkpoints (NVIDIA#9704)

  Show description for 472ff9f
  rlangman

  authored
  472ff9f

  Copy full SHA for 472ff9f

• NeMo performance feature documentation (NVIDIA#9482)

  erhoo82

  authored
  6f91dcc

  Copy full SHA for 6f91dcc

• Alit/mamba (NVIDIA#9696)

  Show description for 3482dc1
  

  3 peopleauthored
  3482dc1

  Copy full SHA for 3482dc1

• add a bit more for timeout (NVIDIA#9702)

  Show description for 693c55f
  pablo-garay

  authored
  693c55f

  Copy full SHA for 693c55f

• chore: Version bump NeMo (NVIDIA#9631)

  Show description for 3cf5a1d
  ko3n1g

  authored
  3cf5a1d

  Copy full SHA for 3cf5a1d

• Huvu/mcore t5 (NVIDIA#9677)

  Show description for d4f1d3c
  

  huvunvidia

  and

  Huy Vu2

  authored
  d4f1d3c

  Copy full SHA for d4f1d3c

• Parametrize FPS group (NVIDIA#9648) (NVIDIA#9669)

  Show description for 08cc515
  

  3 peopleauthored
  08cc515

  Copy full SHA for 08cc515

• LITA integration (NVIDIA#9578)

  Show description for 900ca0b
  

  3 peopleauthored
  900ca0b

  Copy full SHA for 900ca0b

Commits on Jul 10, 2024

• fix pipeline parallel dtype bug (NVIDIA#9637) (NVIDIA#9661)

  Show description for 4e5174b
  

  3 peopleauthored
  4e5174b

  Copy full SHA for 4e5174b

• Update llama-3 PEFT notebook to download model from NGC (NVIDIA#9667)

  Show description for 3ab0a2a
  shashank3959

  authored
  3ab0a2a

  Copy full SHA for 3ab0a2a

• Added CPU offloading docs (NVIDIA#9479)

  Show description for 14d42dc
  

  3 peopleauthored
  14d42dc

  Copy full SHA for 14d42dc

• unpin transformers version (NVIDIA#9606)

  Show description for b4821e1
  

  3 peopleauthored
  b4821e1

  Copy full SHA for b4821e1

• Contrastive Reranker/Reward model (NVIDIA#9171)

  Show description for 74e32c8
  arendu

  authored
  74e32c8

  Copy full SHA for 74e32c8

• Revert "enables default data step in megatron parallel to operate on a wider …" (NVIDIA#9666)

  marcromeyn

  authored
  355d3c5

  Copy full SHA for 355d3c5

• enables default data step in megatron parallel to operate on a wider variety of tensors (NVIDIA#9641)

  Show description for f5d5221
  

  jomitchellnv

  and

  marcromeyn

  authored
  f5d5221

  Copy full SHA for f5d5221

Commits on Jul 9, 2024

• Use FP8 in GPT TP2 test (NVIDIA#9451)

  Show description for 2ee8646
  

  jbaczek

  and

  pablo-garay

  authored
  2ee8646

  Copy full SHA for 2ee8646

• Triton deployment improvements for in-framework models (NVIDIA#9600)

  Show description for 8898b76
  

  jukim-nv

  and

  oyilmaz-nvidia

  authored
  8898b76

  Copy full SHA for 8898b76

• SDXL improvements (and support for Draft+) [DRAFT PR] (NVIDIA#9543)

  Show description for 1c73e1b
  

  3 peopleauthored
  1c73e1b

  Copy full SHA for 1c73e1b

• Update Dockerfile.ci (NVIDIA#9651)

  Show description for b97da9c
  huvunvidia

  authored
  b97da9c

  Copy full SHA for b97da9c

Pagination

Previous

Next