Llama-3.1-8B-dpo-10k

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

• Loss: 0.7481
• Rewards/real: 0.4771
• Rewards/generated: 0.1826
• Rewards/accuracies: 0.6538
• Rewards/margins: 0.2945
• Logps/generated: -119.3277
• Logps/real: -137.5565
• Logits/generated: -1.4543
• Logits/real: -1.5698

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: 5e-07
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/real	Rewards/generated	Rewards/accuracies	Rewards/margins	Logps/generated	Logps/real	Logits/generated	Logits/real
0.3318	0.6494	200	0.7481	0.4771	0.1826	0.6538	0.2945	-119.3277	-137.5565	-1.4543	-1.5698

Framework versions

• Transformers 4.43.3
• Pytorch 2.2.2+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1