dpo5k2

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

• Loss: 6.1354
• Rewards/chosen: -37.3199
• Rewards/rejected: -114.4451
• Rewards/accuracies: 0.7599
• Rewards/margins: 77.1251
• Logps/rejected: -1316.1266
• Logps/chosen: -541.0429
• Logits/rejected: -13.2895
• Logits/chosen: -11.3220

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: 0.001
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 64
• 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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.147	2.5744	100	6.1354	-37.3199	-114.4451	0.7599	77.1251	-1316.1266	-541.0429	-13.2895	-11.3220

Framework versions

• Transformers 4.43.0.dev0
• Pytorch 2.2.2+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1