library_name, base_model, tags, datasets, model-index

library_name	base_model	tags	datasets	model-index
transformers	W-61/llama-3-8b-base-sft-hh-harmless-8xh200	alignment-handbook beta-dpo generated_from_trainer	Anthropic/hh-rlhf	name results llama-3-8b-base-beta-dpo-hh-harmless-8xh200-20260410-223557

llama-3-8b-base-beta-dpo-hh-harmless-8xh200-20260410-223557

This model is a fine-tuned version of W-61/llama-3-8b-base-sft-hh-harmless-8xh200 on the Anthropic/hh-rlhf dataset. It achieves the following results on the evaluation set:

• Loss: 0.5633
• Beta Dpo/gap Mean: 8.8052
• Beta Dpo/gap Std: 15.1783
• Beta Dpo/beta Used Raw: 0.1070
• Beta Dpo/beta Used: 0.1070
• Beta Dpo/mask Keep Frac: 1.0
• Logits/chosen: -0.4218
• Logits/rejected: -0.4089

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-07
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 128
• total_eval_batch_size: 128
• 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

Training Loss	Epoch	Step	Validation Loss	Beta Dpo/gap Mean	Beta Dpo/gap Std	Beta Dpo/beta Used Raw	Beta Dpo/beta Used	Beta Dpo/mask Keep Frac	Logits/chosen	Logits/rejected
0.6231	0.3030	100	0.6186	1.9525	4.8480	0.1117	0.1117	1.0	-0.5574	-0.5400
0.498	0.6061	200	0.5506	6.7801	11.7207	0.1056	0.1056	1.0	-0.4723	-0.4582
0.5615	0.9091	300	0.5633	8.8052	15.1783	0.1070	0.1070	1.0	-0.4218	-0.4089

Framework versions

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