• Model creator: https://huggingface.co/jbjeong91/
• Original model: https://huggingface.co/jbjeong91/llama3.1-cpo_j-full-0911/

Original model description:

• alignment-handbook
• trl
• cpo
• generated_from_trainer
• trl
• cpo
• generated_from_trainer datasets:
• princeton-nlp/llama3-ultrafeedback model-index:
• name: llama3.1-cpo_j-full-0911 results: []

llama3.1-cpo_j-full-0911

• Loss: 1.4373
• Rewards/chosen: -14.1493
• Rewards/rejected: -15.5710
• Rewards/accuracies: 0.6543
• Rewards/margins: 1.4217
• Logps/rejected: -155.7095
• Logps/chosen: -141.4926
• Logits/rejected: -0.1136
• Logits/chosen: -0.1476
• Nll Loss: 0.1725

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

• learning_rate: 1e-06
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• total_eval_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 5

Training results

Framework versions

• Transformers 4.44.2
• Pytorch 2.3.1
• Datasets 2.21.0
• Tokenizers 0.19.1