Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 15, 2023

• [Feature] Add Ascend NPU accelerator support (huggingface#1096)

  Show description for d708ec2
  statelesshz

  authored
  d708ec2

• Updated documentation for docs/source/reward_trainer.mdx to import the correct Enum for the reward modelling LoRA config (huggingface#1092)

  cm2435

  authored
  8140129

Commits on Dec 12, 2023

• [DPO] use ref model logprobs if it exists in the data (huggingface#885)

  Show description for 48b3ef0
  

  4 peopleauthored
  48b3ef0

• consistency on log (huggingface#1084)

  tcapelle

  authored
  c0ce52a

Commits on Dec 11, 2023

• Removing tyro in sft_llama2.py (huggingface#1081)

  Show description for 393dbf6
  vwxyzjn

  authored
  393dbf6

• Make CI happy (huggingface#1080)

  Show description for 94fa4b0
  younesbelkada

  authored
  94fa4b0

• add local folder support as input for rl_training. (huggingface#1078)

  Show description for cb7819e
  sywangyi

  authored
  cb7819e

• Add args to SFT example (huggingface#1079)

  lewtun

  authored
  8f0fc4c

• [DPO] add KTO loss (huggingface#1075)

  Show description for d275cb4
  

  kashif

  and

  lewtun

  authored
  d275cb4

Commits on Dec 7, 2023

• Add missing loss_type in ValueError message (huggingface#1067)

  alvarobartt

  authored
  7d0a8ee

Commits on Dec 6, 2023

• enable multiple eval datasets (huggingface#1052)

  Show description for 5a23354
  peter-sk

  authored
  5a23354

• [SFTTrainer] Fix Trainer when args is None (huggingface#1064)

  Show description for 9fb00cf
  younesbelkada

  authored
  9fb00cf

• [core] Fix failing tests on main (huggingface#1065)

  Show description for ee44946
  younesbelkada

  authored
  ee44946

• update doc for the computer_metrics argument of SFTTrainer (huggingface#1062)

  albertauyeung

  authored
  7f2401b

Commits on Dec 5, 2023

• Improve PreTrainedModelWrapper._get_current_device (huggingface#1048)

  Show description for 23bf9d4
  billvsme

  authored
  23bf9d4

• Update doc CI (huggingface#1060)

  lewtun

  authored
  501c347

Commits on Dec 4, 2023

• [SFT Trainer] precompute packed iterable into a dataset (huggingface#979)

  Show description for f06f357
  

  3 peopleauthored
  f06f357

• Fixing accelerator version function call. (huggingface#1056)

  Show description for 4cdc03a
  

  ParthaEth

  and

  Partha Ghosh

  authored
  4cdc03a

Commits on Dec 1, 2023

• Update dpo_trainer.py (huggingface#1049)

  rdk31

  authored
  a60ceef

• Revert "[DPO] Refactor eval logging of dpo trainer (huggingface#954)" (huggingface#1047)

  Show description for baa8f09
  lvwerra

  authored
  baa8f09

• remove spurious optimize_cuda_cache deprecation warning on init (huggingface#1045)

  Show description for c859f5f
  ChanderG

  authored
  c859f5f

Commits on Nov 30, 2023

• Fixes reward and text gathering in distributed training (huggingface#850)

  Show description for 481ef96
  

  3 peopleauthored
  481ef96

• [DPO] Refactor eval logging of dpo trainer (huggingface#954)

  Show description for 6d9ea38
  

  mnoukhov

  and

  lvwerra

  authored
  6d9ea38

• spelling is hard (huggingface#1043)

  grahamannett

  authored
  c203e47

• [DPO] cDPO loss (huggingface#1035)

  Show description for c84e591
  kashif

  authored
  c84e591

Commits on Nov 29, 2023

• Update utils.py (huggingface#1012)

  Show description for 4b67af3
  

  ZihanWang314

  and

  lvwerra

  authored
  4b67af3

Commits on Nov 24, 2023

• [DPO] IPO Training loss (huggingface#1022)

  Show description for 55d7c95
  

  kashif

  and

  lvwerra

  authored
  55d7c95

Commits on Nov 23, 2023

• Add missing elements to sft_trainer document (huggingface#1029)

  mutichung

  authored
  3719f7a

• Remove duplicate data loading in rl_training.py (huggingface#1020)

  Show description for e7961e4
  viethoangtranduong

  authored
  e7961e4

Commits on Nov 21, 2023

• [Multi-Adapter PPO] Fix and Refactor reward model adapter (huggingface#982)

  Show description for b307faf
  mnoukhov

  authored
  b307faf

Commits on Nov 20, 2023

• Adds requires_grad to input for non-quantized peft models (huggingface#1006)

  Show description for aea1da8
  younesbelkada

  authored
  aea1da8

• Update how_to_train.md (huggingface#1003)

  Show description for e5eb4db
  

  halfrot

  and

  vwxyzjn

  authored
  e5eb4db

Commits on Nov 15, 2023

• Fixed wrong trigger for warning (huggingface#971)

  Show description for 28bdb6a
  zabealbe

  authored
  28bdb6a

• make distributed true for multiple process (huggingface#997)

  Show description for e140d22
  

  allanj

  and

  younesbelkada

  authored
  e140d22

Commits on Nov 14, 2023

• add docs (huggingface#992)

  younesbelkada

  authored
  e23a541

Pagination

PreviousNext