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-4xh200	alignment-handbook margin-dpo generated_from_trainer	Anthropic/hh-rlhf	name results llama-3-8b-base-margin-dpo-hh-harmless

llama-3-8b-base-margin-dpo-hh-harmless

This model is a fine-tuned version of W-61/llama-3-8b-base-sft-hh-harmless-4xh200 on the Anthropic/hh-rlhf dataset. It achieves the following results on the evaluation set:

• Loss: 0.5348
• Margin Dpo/margin Mean: 60.1785
• Margin Dpo/margin Std: 94.6210
• Logps/chosen: -211.5150
• Logps/rejected: -274.1422
• Logps/ref Chosen: -75.3065
• Logps/ref Rejected: -77.7551
• Logits/chosen: 0.8960
• Logits/rejected: 0.8635

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: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• total_eval_batch_size: 32
• 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	Margin Dpo/margin Mean	Margin Dpo/margin Std	Logps/chosen	Logps/rejected	Logps/ref Chosen	Logps/ref Rejected	Logits/chosen	Logits/rejected
1.1015	0.4535	300	0.5574	48.6165	83.3105	-200.8800	-251.9452	-75.3065	-77.7551	0.9108	0.8757
1.1391	0.9070	600	0.5348	60.1785	94.6210	-211.5150	-274.1422	-75.3065	-77.7551	0.8960	0.8635

Framework versions

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