---
tags:
- generated_from_trainer
model-index:
- name: bt-rm
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# bt-rm

This model was trained from LLaMA 3.1 8B Instruct with dataset `hendrydong/preference_700K` (Preprocessed dataset `RyanYr/preference_700K_llama31_tokenized`). Training script is https://github.com/yurun-yuan/RLHF-Reward-Modeling/blob/4b827117dc9a85062c396eb62200b48e6dbfd596/bradley-terry-rm/llama3_rm.py 

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 2e-06
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- gradient_accumulation_steps: 256
- total_train_batch_size: 1024
- total_eval_batch_size: 4
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.03
- num_epochs: 1

### Training results



### Framework versions

- Transformers 4.43.3
- Pytorch 2.1.2+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1