zephyr-7b-dpo-lora-ultrafeedback

This model is a fine-tuned version of EllieS/zephyr-7b-sft-qlora on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:

• Loss: 0.5939
• Rewards/chosen: -0.2154
• Rewards/rejected: -0.5247
• Rewards/accuracies: 0.6920
• Rewards/margins: 0.3093
• Logps/rejected: -320.4754
• Logps/chosen: -313.5164
• Logits/rejected: -2.1488
• Logits/chosen: -2.2262

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-06
• train_batch_size: 4
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• 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

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.5796	0.92	7000	0.5940	-0.2151	-0.5243	0.6940	0.3091	-320.4359	-313.4948	-2.1487	-2.2261

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.6
• Tokenizers 0.15.0