qwen3-8b-base-r-dpo-ultrafeedback-4xh200-batch-128-20260422-131855

• Loss: 0.5512
• R Dpo/chosen Len: 294.4800
• R Dpo/rejected Len: 249.8700
• R Dpo/length Delta: 44.6100
• R Dpo/regularization Term: 0.0
• Logps/chosen: -313.3703
• Logps/rejected: -344.1740
• Logps/ref Chosen: -281.4589
• Logps/ref Rejected: -261.8495
• Logits/chosen: 1.5782
• Logits/rejected: 1.6653

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

• learning_rate: 5e-07
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• total_eval_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1

Training results

Framework versions

• Transformers 4.51.0
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.21.4