gpt2-large-simpo

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.2764
• Rewards/chosen: -4.4922
• Rewards/rejected: -4.8516
• Rewards/accuracies: 0.5655
• Rewards/margins: 0.3589
• Logps/rejected: -2.4258
• Logps/chosen: -2.2461
• Logits/rejected: -7.4219
• Logits/chosen: -7.1797

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: 6e-07
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
1.342	0.8373	400	1.2764	-4.4922	-4.8516	0.5655	0.3589	-2.4258	-2.2461	-7.4219	-7.1797

Framework versions

• Transformers 4.42.0
• Pytorch 2.2.2+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1