eurus-7b-cost-UI-UC-5e-7

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

• Loss: 0.7012
• Rewards/chosen: -0.8455
• Rewards/rejected: -0.9219
• Rewards/accuracies: 0.5507
• Rewards/margins: 0.0764
• Rewards/margins Max: 0.9500
• Rewards/margins Min: -0.7111
• Rewards/margins Std: 0.5453
• Logps/rejected: -358.1681
• Logps/chosen: -363.9157
• Logits/rejected: -2.1126
• Logits/chosen: -2.2173

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: 2
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• total_eval_batch_size: 16
• 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

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Rewards/margins Max	Rewards/margins Min	Rewards/margins Std	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.296	1.0	2616	0.7012	-0.8455	-0.9219	0.5507	0.0764	0.9500	-0.7111	0.5453	-358.1681	-363.9157	-2.1126	-2.2173

Framework versions

• PEFT 0.7.1
• Transformers 4.39.0.dev0
• Pytorch 2.1.2+cu121
• Datasets 2.14.6
• Tokenizers 0.15.2