dpo4

This model is a fine-tuned version of deepseek-ai/deepseek-coder-1.3b-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 18.2999
• Rewards/chosen: -55.1127
• Rewards/rejected: -55.1897
• Rewards/accuracies: 0.4073
• Rewards/margins: 0.0770
• Logps/rejected: -625.9051
• Logps/chosen: -588.4946
• Logits/rejected: -8.9525
• Logits/chosen: -8.9519

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: 0.1
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• num_epochs: 4
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
22.9478	2.3088	100	18.2999	-55.1127	-55.1897	0.4073	0.0770	-625.9051	-588.4946	-8.9525	-8.9519

Framework versions

• PEFT 0.10.0
• Transformers 4.45.0
• Pytorch 2.5.1+cu124
• Datasets 2.19.2
• Tokenizers 0.20.3