zephyr-7b-dpo-full-magpi-3

This model is a fine-tuned version of alignment-handbook/zephyr-7b-sft-full on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0003
• Rewards/chosen: -1.7320
• Rewards/rejected: -49.6169
• Rewards/accuracies: 1.0
• Rewards/margins: 47.8849
• Logps/rejected: -5602.4775
• Logps/chosen: -540.1784
• Logits/rejected: 1.2551
• Logits/chosen: -1.4806

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: 8
• eval_batch_size: 8
• seed: 55
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• total_eval_batch_size: 64
• 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
0.0007	0.2841	100	0.0010	-1.3251	-44.7362	0.9980	43.4111	-5114.4058	-499.4881	-1.4200	-2.7411
0.0007	0.5682	200	0.0004	-1.6225	-48.3903	1.0	46.7678	-5479.8149	-529.2302	0.7429	-1.8781
0.0008	0.8523	300	0.0003	-1.7320	-49.6169	1.0	47.8849	-5602.4775	-540.1784	1.2551	-1.4806

Framework versions

• Transformers 4.44.0.dev0
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1