Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 11, 2023

• OnnxExporter (ludwig-ai#3761)

  Show description for 8c47c3c
  skanjila

  authored
  8c47c3c

  Copy full SHA for 8c47c3c

Commits on Nov 9, 2023

• [BUGFIX] Un-pin "transformers" version and update file names for saved merged model directory, because issue with "transformers" impacting "peft" has been fixed. (ludwig-ai#3775)

  alexsherstinsky

  authored
  dc3ec73

  Copy full SHA for dc3ec73

Commits on Nov 8, 2023

• enh: Add on_checkpoint callback that triggers after every checkpoint even if the model was not evaluated (ludwig-ai#3763)

  Show description for 9389b16
  

  3 peopleauthored
  9389b16

  Copy full SHA for 9389b16

• fix LudwigProgressBar rendering on model resume (ludwig-ai#3770)

  geoffreyangus

  authored
  266a36c

  Copy full SHA for 266a36c

• fix: adapter checkpoint loading on resume (ludwig-ai#3769)

  Show description for 7458885
  

  geoffreyangus

  and

  pre-commit-ci[bot]

  authored
  7458885

  Copy full SHA for 7458885

Commits on Nov 6, 2023

• [MAINTENANCE] Update contributor instructions with additional steps that may be needed for installation to succeed. (ludwig-ai#3764)

  alexsherstinsky

  authored
  9992f32

  Copy full SHA for 9992f32

Commits on Nov 4, 2023

• Revert to "pytorch_model.bin", for Transformers version 4.34.1 -- it will change to "model.safetensors" for versions 4.35.0 and greater.

  alexsherstinsky

  committed
  cd47fd2

  Copy full SHA for cd47fd2

• Revert to "pytorch_model.bin", for Transformers version 4.34.1 -- it will change to "model.safetensors" for versions 4.35.0 and greater.

  alexsherstinsky

  committed
  e87aeae

  Copy full SHA for e87aeae

• Pinning Transformers to not include version 4.35.0 because it breaks a method in PEFT.

  alexsherstinsky

  committed
  bead1f1

  Copy full SHA for bead1f1

• WIP

  alexsherstinsky

  committed
  ec2bfa9

  Copy full SHA for ec2bfa9

Commits on Nov 3, 2023

• fix: Check device type in _initialize_llm before moving to cuda. (ludwig-ai#3762)

  justinxzhao

  authored
  4bf7c23

  Copy full SHA for 4bf7c23

Commits on Oct 30, 2023

• Update ludwig version to 0.9.dev. (ludwig-ai#3638)

  justinxzhao

  authored
  96c2653

  Copy full SHA for 96c2653

Commits on Oct 28, 2023

• Define initialize_llm() and generate() methods. Remove extra logging in llm.py (ludwig-ai#3711)

  justinxzhao

  authored
  cd20458

  Copy full SHA for cd20458

Commits on Oct 27, 2023

• QOL: Fail during preprocessing if max sequence lengths are shorter than the prompt template. (ludwig-ai#3719)

  justinxzhao

  authored
  a0c42d8

  Copy full SHA for a0c42d8

Commits on Oct 26, 2023

• Pin getdaft<0.2 in requirements (ludwig-ai#3757)

  Show description for ed34128
  

  jaychia

  and

  Jay Chia

  authored
  ed34128

  Copy full SHA for ed34128

Commits on Oct 25, 2023

• Update base.py (ludwig-ai#3750)

  jetaudio

  authored
  6b18a8e

  Copy full SHA for 6b18a8e

• Fix: Prevent memory from ballooning during post-training evaluation (ludwig-ai#3756)

  arnavgarg1

  authored
  a5628f3

  Copy full SHA for a5628f3

• [MAINTENACE] Fixing method overload violations; providing typehints and method argument documentation (ludwig-ai#3753)

  alexsherstinsky

  authored
  31db203

  Copy full SHA for 31db203

Commits on Oct 23, 2023

• [MAINTENANCE] Restrict "fsspec" to not include the latest version "2023.10.0" (ludwig-ai#3752)

  alexsherstinsky

  authored
  8b423f1

  Copy full SHA for 8b423f1

Commits on Oct 20, 2023

• Fix Consumer Complaints Gen Dataset Config (ludwig-ai#3746)

  connor-mccorm

  authored
  0620b00

  Copy full SHA for 0620b00

• Uncap the torchmetrics version, so we can get fixes in 1.0+. (ludwig-ai#3741)

  Show description for 77259c7
  

  3 peopleauthored
  77259c7

  Copy full SHA for 77259c7

• Increment the epoch before running evaluation (ludwig-ai#3729)

  Show description for e6c60ab
  

  jeffkinnison

  and

  justinxzhao

  authored
  e6c60ab

  Copy full SHA for e6c60ab

• Simplify how we set pad token and pad token ID for huggingface tokenizers. (ludwig-ai#3735)

  justinxzhao

  authored
  e52f346

  Copy full SHA for e52f346

• Retry LLM model downloads from HF Hub with exponential backoff when there is a Read timeout (ludwig-ai#3742)

  arnavgarg1

  authored
  cd39185

  Copy full SHA for cd39185

• Raise error if output feature name is in the prompt template (ludwig-ai#3743)

  arnavgarg1

  authored
  50ad21c

  Copy full SHA for 50ad21c

Commits on Oct 19, 2023

• [CI] Fix test broken by transformers v.4.34.1 (ludwig-ai#3739)

  justinxzhao

  authored
  f98b8f6

  Copy full SHA for f98b8f6

Commits on Oct 18, 2023

• [BUGFIX] Making integration tests in the group "integration_tests_e" pass in Azure CI (ludwig-ai#3737)

  alexsherstinsky

  authored
  fb6d866

  Copy full SHA for fb6d866

• [BUGFIX] Make sure that the "hyperopt" directory in the local repository path is deleted after test run is completed (ludwig-ai#3736)

  alexsherstinsky

  authored
  f7010f3

  Copy full SHA for f7010f3

Commits on Oct 17, 2023

• Fix calculation of dataset size (ludwig-ai#3723)

  Infernaught

  authored
  c711d0a

  Copy full SHA for c711d0a

• DRAFT: Call .finish() for WandB callback on trainer end. (ludwig-ai#3710)

  justinxzhao

  authored
  92d0e0c

  Copy full SHA for 92d0e0c

Commits on Oct 16, 2023

• [MAINTENACE] Improve typehints and use constants instead of common hardcoded string values. (ludwig-ai#3732)

  alexsherstinsky

  authored
  305fdd2

  Copy full SHA for 305fdd2

• Add functionality to allow direct importing of datasets from Hugging Face (ludwig-ai#3664)

  Infernaught

  authored
  d6c8e09

  Copy full SHA for d6c8e09

• Set default text preprocessing lowercase behavior to False (ludwig-ai#3721)

  connor-mccorm

  authored
  fa94f4d

  Copy full SHA for fa94f4d

Commits on Oct 13, 2023

• Limit ray versions to keep 2.4 out of the picture. (ludwig-ai#3731)

  Show description for 4491d65
  

  trevenrawr

  and

  Trevor DiMartino

  authored
  4491d65

  Copy full SHA for 4491d65

• Pin PEFT to PEFT master for Mistral-7B support (ludwig-ai#3725)

  arnavgarg1

  authored
  b6bdf9c

  Copy full SHA for b6bdf9c

Pagination

Previous

Next