MLP-FinLLM-dpo-7b

This model is a fine-tuned version of seanmemery/MLP-FinLLM-7b-it on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6707
• Rewards/chosen: 0.0674
• Rewards/rejected: -0.0381
• Rewards/accuracies: 0.2975
• Rewards/margins: 0.1055
• Logps/rejected: -22.125
• Logps/chosen: -29.875
• Logits/rejected: -5.8438
• Logits/chosen: -5.875

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: 1e-06
• train_batch_size: 32
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.6367	0.69	50	0.6707	0.0674	-0.0381	0.2975	0.1055	-22.125	-29.875	-5.8438	-5.875

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2