metadata

library_name: transformers
license: llama3.1
base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
tags:
  - alignment-handbook
  - trl
  - dpo
  - generated_from_trainer
  - trl
  - dpo
  - generated_from_trainer
datasets:
  - tanliboy/orca_dpo_pairs
model-index:
  - name: lambda-llama-3-8b-dpo-test-orca
    results: []

lambda-llama-3-8b-dpo-test-orca

This model is a fine-tuned version of meta-llama/Meta-Llama-3.1-8B-Instruct on the tanliboy/orca_dpo_pairs dataset. It achieves the following results on the evaluation set:

Loss: 0.1235
Rewards/chosen: -2.8028
Rewards/rejected: -6.6852
Rewards/accuracies: 0.9643
Rewards/margins: 3.8824
Logps/rejected: -970.0546
Logps/chosen: -562.0943
Logits/rejected: -1.9611
Logits/chosen: -2.4346

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-07
train_batch_size: 4
eval_batch_size: 4
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 128
total_eval_batch_size: 32
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Framework versions

Transformers 4.44.2
Pytorch 2.4.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1