RewardModel

This model is a fine-tuned version of roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4247
• Accuracy: 0.9675

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: 2e-06
• train_batch_size: 4
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7081	0.4	250	0.6320	0.9555
0.7515	0.8	500	0.5131	0.9615
0.6210	1.2	750	0.4501	0.968
0.5654	1.6	1000	0.4255	0.9675
0.4647	2.0	1250	0.4247	0.9675

Framework versions

• Transformers 5.17.0
• Pytorch 2.11.0+cu128
• Datasets 5.0.1
• Tokenizers 0.23.2