zephyr-spin-nll

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

• Loss: 0.1067
• Rewards/real: 3.1337
• Rewards/generated: -10.7944
• Rewards/accuracies: 1.0
• Rewards/margins: 13.9281
• Logps/generated: -201.2369
• Logps/real: -103.3285
• Logits/generated: -2.7994
• Logits/real: -2.6149

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
• total_train_batch_size: 32
• 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: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/real	Rewards/generated	Rewards/accuracies	Rewards/margins	Logps/generated	Logps/real	Logits/generated	Logits/real
0.1098	0.32	500	0.1129	2.7263	-8.9091	1.0	11.6354	-182.3844	-107.4027	-2.8260	-2.6182
0.0963	0.64	1000	0.1078	3.0493	-10.6129	1.0	13.6622	-199.4227	-104.1727	-2.8201	-2.6740
0.0957	0.96	1500	0.1067	3.1337	-10.7944	1.0	13.9281	-201.2369	-103.3285	-2.7994	-2.6149

Framework versions

• Transformers 4.37.0
• Pytorch 2.1.2+cu121
• Datasets 2.14.6
• Tokenizers 0.15.2