phi-2-gpo-test-longest-iter-random-1

This model is a fine-tuned version of DUAL-GPO/phi-2-gpo-test-longest-iter-random-0 on the HuggingFaceH4/ultrafeedback_binarized dataset. It achieves the following results on the evaluation set:

• Loss: 0.0022
• Rewards/chosen: 0.0010
• Rewards/rejected: 0.0016
• Rewards/accuracies: 0.4790
• Rewards/margins: -0.0006
• Logps/rejected: -278.5288
• Logps/chosen: -306.2895
• Logits/rejected: 0.0882
• Logits/chosen: -0.0097

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: 4
• seed: 42
• distributed_type: multi-GPU
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.0016	1.6	100	0.0021	0.0005	0.0001	0.4910	0.0004	-278.6740	-306.3399	0.0933	-0.0050
0.0017	3.2	200	0.0022	-0.0007	-0.0006	0.4950	-0.0001	-278.7457	-306.4609	0.0886	-0.0095

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.2.1+cu121
• Datasets 2.14.6
• Tokenizers 0.15.2