RLHFlow/RLHF-Reward-Modeling

Issues

• Missing code for ODIN

  #46 opened a month ago by maoliyuan
  1

• Question regarding ARMO stage2-train code

  #37 opened 2 months ago by RayWang-iat
  0

• stage1-train:RuntimeError: torch.cat(): expected a non-empty list of Tensors

  #36 opened 3 months ago by RayWang-iat
  0

• Armo-rm env set-up and data processing

  #35 opened 3 months ago by MaxwellJryao
  1

• ArmoRM-Llama3-8B-v0.1's tokenizer is different from Meta-Llama-3-8B-Instruct's

  #32 opened 3 months ago by efsotr
  7

• Can I inquire about some training details about armo-rm？

  #27 opened 3 months ago by xiaotian917
  7

• How to finetune ARMO model with custom dataset?

  #23 opened 3 months ago by Helen-Cheung
  4

• Code to reproduce ArmoRM

  #28 opened 3 months ago by halfrot
  5

• reproduce ArmoRM

  #30 opened 3 months ago by richhh520
  3

• Clarification on Reward Usage in DPO Training

  #33 opened 3 months ago by vincezh2000
  1

• preference dataset 404 not found

  #29 opened 4 months ago by wty500
  2

• tutorial to reproduce ArmoRM

  #17 opened 6 months ago by pluiez
  1

• Regarding the Gemma2 Reward Model Structure

  #26 opened 5 months ago by Loong435
  2

• "Token pattern not found in the list" error

  #24 opened 5 months ago by nshen7
  3

• How to batch inference?

  #25 opened 5 months ago by AIR-hl
  0

• Code for Armo on Reward Bench

  #15 opened 5 months ago by philschmid
  4

• Training and evaluating for pair_pm model.

  #21 opened 6 months ago by t-sifanwu
  5

• Bradley-Terry model removes lm head while saving

  #22 opened 6 months ago by Arnav0400
  1

• question of chat templates

  #16 opened 6 months ago by trueRosun
  6

• environment set up issue

  #18 opened 6 months ago by WayXG
  1

• How do you implement SLic on pair_pm model?

  #20 opened 6 months ago by t-sifanwu
  1

• preference_700K dataset's details?

  #19 opened 6 months ago by yechenzhi
  4

• How to calculate the avg score of reward bench?

  #14 opened 7 months ago by eyuansu62
  2

• Cannot run the training script

  #2 opened 7 months ago by peter-peng-w
  1

• how to serve this model?

  #1 opened 7 months ago by jxgu1016
  1

• Cannot understant the code at README.md of pair-pm

  #8 opened 7 months ago by heyzude
  4

• Does pair-pm supports multi-turn conversation?

  #9 opened 7 months ago by heyzude
  2

• Low Safety Score for RM-Gemma-2B Model

  #13 opened 7 months ago by loss4Wang
  2

• can we say PM is better than BT?

  #12 opened 7 months ago by yechenzhi
  2

• quesion about the output

  #11 opened 7 months ago by yechenzhi
  1

• How to construct new pairs for adding to the dataset

  #10 opened 7 months ago by wlhgtc
  1

• KeyError: 'input_ids_j' in training

  #6 opened 8 months ago by iseesaw
  2