17Lab/qwen14b-dpo-rank256-s42

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

qwen14bdpofromsftrank_s42

This model is a fine-tuned version of Qwen/Qwen2.5-14B on the assimilationdpov1 dataset. It achieves the following results on the evaluation set:

• —Loss: 0.0272
• —Rewards/chosen: -0.1747
• —Rewards/rejected: -6.2640
• —Rewards/accuracies: 1.0
• —Rewards/margins: 6.0893
• —Logps/chosen: -5.2166
• —Logps/rejected: -80.8057
• —Logits/chosen: -1.5318
• —Logits/rejected: -1.5291

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
• —trainbatchsize: 2
• —evalbatchsize: 1
• —seed: 42
• —distributed_type: multi-GPU
• —num_devices: 2
• —gradientaccumulationsteps: 16
• —totaltrainbatch_size: 64
• —totalevalbatch_size: 2
• —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_steps: 0.1
• —num_epochs: 1

Training results

Framework versions

• —PEFT 0.18.1
• —Transformers 5.6.0
• —Pytorch 2.7.0+cu128
• —Datasets 4.0.0
• —Tokenizers 0.22.2