Search results

Unexpected keyword argument 'beta' in DPOTrainer initialization

Status: Open.

#32 In RLHFlow/Online-RLHF;

· YijuGuo opened on Dec 19, 2024

Inquiry about the version of alignment-handbook used in the installation guide

Status: Open.

#31 In RLHFlow/Online-RLHF;

· YijuGuo opened on Dec 17, 2024

Unable to install axolotl due to missing bitsandbytes==0.45.0 dependency

Status: Open.

#30 In RLHFlow/Online-RLHF;

· YijuGuo opened on Dec 16, 2024

Questions About Data chosen Strategies

Status: Open.

#29 In RLHFlow/Online-RLHF;

· nantenT opened on Dec 9, 2024

About the results of vanilla DPO

Status: Open.

#28 In RLHFlow/Online-RLHF;

· lucasliunju opened on Nov 18, 2024

Reward-KL Comparison

Status: Open.

#27 In RLHFlow/Online-RLHF;

· vincezh2000 opened on Nov 10, 2024

SFT training objective

Status: Open.

#24 In RLHFlow/Online-RLHF;

· ljb121002 opened on Oct 18, 2024

Negative reward when serving ArmoRM-Llama3-8B-v0.1

Status: Open.

#23 In RLHFlow/Online-RLHF;

· maoliyuan opened on Aug 29, 2024

Question about CUDA/NVCC setups

Status: Open.

#22 In RLHFlow/Online-RLHF;

· rqzhangberkeley opened on Aug 16, 2024

Questions about Nectar Datasets

Status: Open.

#20 In RLHFlow/Online-RLHF;

· XinZhao0211 opened on Aug 9, 2024

Phi3 has a nearly constant DPO loss of 0.69xx

Status: Open.

#17 In RLHFlow/Online-RLHF;

· Arnav0400 opened on Jul 18, 2024

Fail to load weight from pair-preference-model-LLaMA3-8B

Status: Open.

#4 In RLHFlow/Online-RLHF;

· matouk98 opened on May 31, 2024