llama3.2-1B-dpo

This model is a fine-tuned version of meta-llama/Llama-3.2-1B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.7077
• Rewards/chosen: -7.1636
• Rewards/rejected: -9.0302
• Rewards/accuracies: 0.4015
• Rewards/margins: 1.8666
• Logps/rejected: -197.2289
• Logps/chosen: -195.0906
• Logits/rejected: 0.2668
• Logits/chosen: 0.3537

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-05
• train_batch_size: 6
• eval_batch_size: 4
• seed: 42
• 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
0.6114	0.3817	700	0.8334	-7.6439	-9.5517	0.3914	1.9077	-202.4431	-199.8938	0.4338	0.5238
0.6772	0.7634	1400	0.7077	-7.1636	-9.0302	0.4015	1.8666	-197.2289	-195.0906	0.2668	0.3537

Framework versions

• PEFT 0.8.2
• Transformers 4.45.1
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.20.0