llama-3-8b-base-margin-dpo-hh-harmless-batch-size-64

<!-- 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. -->

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: 0.5259
• Margin Dpo/margin Mean: 9.3649
• Margin Dpo/margin Std: 14.8097
• Logps/chosen: -92.0386
• Logps/rejected: -106.0930
• Logps/ref Chosen: -74.8595
• Logps/ref Rejected: -79.5490
• Logits/chosen: 0.3798
• Logits/rejected: 0.3285

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

Framework versions

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