Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Mar 8, 2024

• [KTO] fix various bugs (#1402)

  Show description for 4d862da
  

  3 peopleauthored
  4d862daView commit details

  Copy full SHA for 4d862da

  Browse repository at this point

Commits on Mar 7, 2024

• fix RM script (#1393)

  younesbelkada

  authored
  22b4f54View commit details

  Copy full SHA for 22b4f54

  Browse repository at this point

Commits on Mar 5, 2024

• Fix the pad_token_id error (#1394)

  Show description for 4219cbf
  yuanwu2017

  authored
  4219cbfView commit details

  Copy full SHA for 4219cbf

  Browse repository at this point

Commits on Mar 4, 2024

• Log ddpo reward as float to fix numpy conversion during bf16 training (#1391)

  skavulya

  authored
  3bd0238View commit details

  Copy full SHA for 3bd0238

  Browse repository at this point

Commits on Mar 1, 2024

• [KTO] prevent nans from appearing in metrics (#1386)

  Show description for 067db75
  

  3 peopleauthored
  067db75View commit details

  Copy full SHA for 067db75

  Browse repository at this point

• [KTO] merge eval dataset only if it exists (#1383)

  Show description for 93e85ed
  kashif

  authored
  93e85edView commit details

  Copy full SHA for 93e85ed

  Browse repository at this point

Commits on Feb 29, 2024

• fix bugs in KTO implementation (#1380)

  Show description for 14e0d78
  

  3 peopleauthored
  14e0d78View commit details

  Copy full SHA for 14e0d78

  Browse repository at this point

Commits on Feb 27, 2024

• FIX: Fix the CI again .. (#1374)

  Show description for b32656f
  younesbelkada

  authored
  b32656fView commit details

  Copy full SHA for b32656f

  Browse repository at this point

• Update tests-main.yml (#1373)

  younesbelkada

  authored
  9399bc1View commit details

  Copy full SHA for 9399bc1

  Browse repository at this point

• Update tests-main.yml (#1372)

  younesbelkada

  authored
  11f122aView commit details

  Copy full SHA for 11f122a

  Browse repository at this point

• feature request add `force_use_ref_model` (#1367)

  younesbelkada

  authored
  009c9a6View commit details

  Copy full SHA for 009c9a6

  Browse repository at this point

• add `eval_packing` (#1369)

  younesbelkada

  authored
  7712d42View commit details

  Copy full SHA for 7712d42

  Browse repository at this point

• add ci message sending on TRL (#1370)

  younesbelkada

  authored
  7c2213bView commit details

  Copy full SHA for 7c2213b

  Browse repository at this point

• Add some arguments for support XPU (#1366)

  Show description for ddeebce
  yuanwu2017

  authored
  ddeebceView commit details

  Copy full SHA for ddeebce

  Browse repository at this point

• Fix version for Python<3.8 (#1363)

  samuki

  authored
  cf68d87View commit details

  Copy full SHA for cf68d87

  Browse repository at this point

Commits on Feb 23, 2024

• set seed in sft/dpo/reward_modeling to make result reproducable (#1357)

  Show description for 2a2676e
  sywangyi

  authored
  2a2676eView commit details

  Copy full SHA for 2a2676e

  Browse repository at this point

• fix 8-bit multi-gpu training bug (#1353)

  Show description for ca90cba
  

  fancyerii

  and

  ubuntu

  authored
  ca90cbaView commit details

  Copy full SHA for ca90cba

  Browse repository at this point

Commits on Feb 22, 2024

• more userfriendly (#1350)

  younesbelkada

  authored
  4f97fb4View commit details

  Copy full SHA for 4f97fb4

  Browse repository at this point

Commits on Feb 19, 2024

• Kto trainer (#1181)

  Show description for a46cd84
  

  4 peopleauthored
  a46cd84View commit details

  Copy full SHA for a46cd84

  Browse repository at this point

Commits on Feb 18, 2024

• Update Example to reflect #aa35fec (#1333)

  nautsimon

  authored
  1f56bffView commit details

  Copy full SHA for 1f56bff

  Browse repository at this point

Commits on Feb 16, 2024

• set dev version (#1332)

  younesbelkada

  authored
  1bfe0b8View commit details

  Copy full SHA for 1bfe0b8

  Browse repository at this point

• Release: v0.7.11 (#1331)

  younesbelkada

  authored
  0f13e51View commit details

  Copy full SHA for 0f13e51

  Browse repository at this point

Commits on Feb 15, 2024

• [`core` / `xxxTrainer`] Automatic tagging (#1329)

  Show description for 1e77d8a
  younesbelkada

  authored
  1e77d8aView commit details

  Copy full SHA for 1e77d8a

  Browse repository at this point

• add tests on transformers peft main (#1328)

  younesbelkada

  authored
  3b1911cView commit details

  Copy full SHA for 3b1911c

  Browse repository at this point

• [`core` / `DDPO`] Fix diffusers import issue (#1314)

  Show description for 851e7fe
  younesbelkada

  authored
  851e7feView commit details

  Copy full SHA for 851e7fe

  Browse repository at this point

• Update README.md to clarify model requirement (#1315)

  Show description for 31b02d0
  markstur

  authored
  31b02d0View commit details

  Copy full SHA for 31b02d0

  Browse repository at this point

• pre-commit: replace linters + formatters with Ruff; fix some issues (#1300)

  Show description for 9bc478e
  akx

  authored
  9bc478eView commit details

  Copy full SHA for 9bc478e

  Browse repository at this point

Commits on Feb 14, 2024

• Best practice recommendation update for dpo_trainer.mdx (#1325)

  Show description for 29f162b
  R-seny

  authored
  29f162bView commit details

  Copy full SHA for 29f162b

  Browse repository at this point

Commits on Feb 6, 2024

• Fix PPOTrainer argument train_dataset -> dataset (#1321)

  Show description for 6852097
  j-cb

  authored
  6852097View commit details

  Copy full SHA for 6852097

  Browse repository at this point

Commits on Feb 2, 2024

• Fix AttributeError in dpo_trainer for reference_free case in dpo_loss function (#1313)

  Show description for f12a1da
  maliozer

  authored
  f12a1daView commit details

  Copy full SHA for f12a1da

  Browse repository at this point

• Fix typos in docs for Multi Adapter RL (MARL). (#1312)

  Show description for ae87b3a
  elhusseiniali

  authored
  ae87b3aView commit details

  Copy full SHA for ae87b3a

  Browse repository at this point

Commits on Feb 1, 2024

• ENH: Run CI only if relevant files are modified (#1309)

  Show description for 3f7cee7
  younesbelkada

  authored
  3f7cee7View commit details

  Copy full SHA for 3f7cee7

  Browse repository at this point

• Codemod Unittest assertions to bare asserts (#1301)

  Show description for ae8431b
  akx

  authored
  ae8431bView commit details

  Copy full SHA for ae8431b

  Browse repository at this point

• Update sft_trainer.mdx to add note on launching DDP training (#1308)

  Show description for 66a976c
  johnowhitaker

  authored
  66a976cView commit details

  Copy full SHA for 66a976c

  Browse repository at this point

• Add num_proc arg to the eval_dataset processing (#1307)

  esceptico

  authored
  8149303View commit details

  Copy full SHA for 8149303

  Browse repository at this point