Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Jun 7, 2024

• integretaed gold reward model testing in dpo

  Nicolinho

  committed
  0948b91

  Copy full SHA for 0948b91

Commits on May 26, 2024

• setup init training

  Nicolinho

  committed
  f46ceb5

  Copy full SHA for f46ceb5

Commits on May 23, 2024

• Apply deprecated evaluation_strategy (huggingface#1559)

  Show description for a02513c
  

  muellerzr

  and

  kashif

  authored
  a02513c

  Copy full SHA for a02513c

Commits on May 22, 2024

• PPO / Reinforce Trainers (huggingface#1540)

  Show description for 13454d2
  vwxyzjn

  authored
  13454d2

  Copy full SHA for 13454d2

Commits on May 10, 2024

• Update sft_llama2.py to work with the latest API (huggingface#1637)

  Show description for 5aeb752
  xianbaoqian

  authored
  5aeb752

  Copy full SHA for 5aeb752

• visualize rm prediction (huggingface#1636)

  Show description for 8799952
  vwxyzjn

  authored
  8799952

  Copy full SHA for 8799952

Commits on May 2, 2024

• Update HH dataset on helpful only subset (huggingface#1613)

  Show description for 7075cec
  vwxyzjn

  authored
  7075cec

  Copy full SHA for 7075cec

• Use auto device map (huggingface#1596)

  lewtun

  authored
  5f09131

  Copy full SHA for 5f09131

Commits on Apr 30, 2024

• [KTOTrainer] add BCO (reward shift and underlying distribution matching) (huggingface#1599)

  Show description for d1aa0b6
  seanexp

  authored
  d1aa0b6

  Copy full SHA for d1aa0b6

Commits on Apr 23, 2024

• [SFT] add SFT Trainer Config dataclass (huggingface#1530)

  Show description for f30daa4
  

  kashif

  and

  lewtun

  authored
  f30daa4

  Copy full SHA for f30daa4

• [DPO] DPOConfig class (huggingface#1554)

  Show description for 24fd8dd
  

  kashif

  and

  lewtun

  authored
  24fd8dd

  Copy full SHA for 24fd8dd

Commits on Apr 18, 2024

• enable multiple eos tokens (huggingface#1553)

  lvwerra

  authored
  9afd901

  Copy full SHA for 9afd901

• FIX: make the train / test fields modulable (huggingface#1551)

  Show description for e04432d
  younesbelkada

  authored
  e04432d

  Copy full SHA for e04432d

Commits on Apr 12, 2024

• VSFT hotfix - adds gen prompt to template and processor to hub (huggingface#1532)

  Show description for 1c0d8bc
  edbeeching

  authored
  1c0d8bc

  Copy full SHA for 1c0d8bc

Commits on Apr 11, 2024

• Adds VLM Training support to SFTTrainer + VSFT script (huggingface#1518)

  Show description for 346c99d
  

  edbeeching

  and

  younesbelkada

  authored
  346c99d

  Copy full SHA for 346c99d

Commits on Mar 27, 2024

• Update KTO example to use better model and ChatML support (huggingface#1485)

  Show description for 0ee349d
  lewtun

  authored
  0ee349d

  Copy full SHA for 0ee349d

Commits on Mar 25, 2024

• Update KTO example with good dataset & chat format (huggingface#1481)

  Show description for c4f0f41
  lewtun

  authored
  c4f0f41

  Copy full SHA for c4f0f41

Commits on Mar 24, 2024

• Fix hyperparameters in KTO example (huggingface#1474)

  Show description for 9ce7ac6
  lewtun

  authored
  9ce7ac6

  Copy full SHA for 9ce7ac6

Commits on Mar 22, 2024

• ORPO trainer (huggingface#1435)

  Show description for 2ce8e45
  

  3 peopleauthored
  2ce8e45

  Copy full SHA for 2ce8e45

• Add CPOTrainer (huggingface#1382)

  Show description for d1df79f
  

  3 peopleauthored
  d1df79f

  Copy full SHA for d1df79f

Commits on Mar 20, 2024

• Use the standard dataset for DPO CLI (huggingface#1456)

  Show description for 423991c
  vwxyzjn

  authored
  423991c

  Copy full SHA for 423991c

• add eos token to generate (huggingface#1459)

  lvwerra

  authored
  5095e7f

  Copy full SHA for 5095e7f

• Fix chat CLI for model revisions (huggingface#1458)

  Show description for 9fcf61d
  lewtun

  authored
  9fcf61d

  Copy full SHA for 9fcf61d

Commits on Mar 19, 2024

• model --> model_name_or_path (huggingface#1452)

  Show description for 3f7ff60
  lvwerra

  authored
  3f7ff60

  Copy full SHA for 3f7ff60

• chat cli (huggingface#1431)

  Show description for 4e622a9
  lvwerra

  authored
  4e622a9

  Copy full SHA for 4e622a9

Commits on Mar 18, 2024

• FEAT: Add CLIs in TRL ! (huggingface#1419)

  Show description for a2aa0f0
  

  younesbelkada

  and

  lvwerra

  authored
  a2aa0f0

  Copy full SHA for a2aa0f0

Commits on Mar 14, 2024

• Create standard dataset for TRL (huggingface#1424)

  Show description for 304e208
  vwxyzjn

  authored
  304e208

  Copy full SHA for 304e208

Commits on Mar 7, 2024

• fix RM script (huggingface#1393)

  younesbelkada

  authored
  22b4f54

  Copy full SHA for 22b4f54

Commits on Mar 5, 2024

• Fix the pad_token_id error (huggingface#1394)

  Show description for 4219cbf
  yuanwu2017

  authored
  4219cbf

  Copy full SHA for 4219cbf

Commits on Mar 4, 2024

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

  skavulya

  authored
  3bd0238

  Copy full SHA for 3bd0238

Commits on Feb 29, 2024

• fix bugs in KTO implementation (huggingface#1380)

  Show description for 14e0d78
  

  3 peopleauthored
  14e0d78

  Copy full SHA for 14e0d78

Commits on Feb 27, 2024

• Add some arguments for support XPU (huggingface#1366)

  Show description for ddeebce
  yuanwu2017

  authored
  ddeebce

  Copy full SHA for ddeebce

Commits on Feb 23, 2024

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

  Show description for 2a2676e
  sywangyi

  authored
  2a2676e

  Copy full SHA for 2a2676e

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

  Show description for ca90cba
  

  fancyerii

  and

  ubuntu

  authored
  ca90cba

  Copy full SHA for ca90cba

Commits on Feb 19, 2024

• Kto trainer (huggingface#1181)

  Show description for a46cd84
  

  4 peopleauthored
  a46cd84

  Copy full SHA for a46cd84

Pagination

Previous

Next