IKEA-7b-UFB-0

This model is a fine-tuned version of weijie210/zephyr-UFB-7b-sft on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0776
• Rewards/chosen: -6.9486
• Rewards/rejected: -22.9186
• Rewards/accuracies: 0.9583
• Rewards/margins: 15.9701
• Logps/rejected: -316.0304
• Logps/chosen: -200.6633
• Logits/rejected: -2.8329
• Logits/chosen: -2.8989

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: 4
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 16
• total_eval_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.1091	0.53	500	0.2325	-3.8178	-23.3989	0.9583	19.5811	-320.8331	-169.3557	-2.8725	-2.9272
0.0336	1.06	1000	0.0688	-5.4221	-18.0263	0.9722	12.6042	-267.1074	-185.3986	-2.8786	-2.9367
0.0247	1.59	1500	0.0597	-5.4161	-19.3860	0.9444	13.9699	-280.7044	-185.3390	-2.8687	-2.9293

Framework versions

• Transformers 4.36.1
• Pytorch 2.0.1+cu117
• Datasets 2.16.1
• Tokenizers 0.15.0