tulu2-13b-cost-tulumix-5e-7-nojudge

This model is a fine-tuned version of allenai/tulu-2-13b on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.6915
• Rewards/chosen: 0.0201
• Rewards/rejected: 0.0161
• Rewards/accuracies: 0.5510
• Rewards/margins: 0.0040
• Rewards/margins Max: 0.0645
• Rewards/margins Min: -0.0595
• Rewards/margins Std: 0.0411
• Logps/rejected: -314.0515
• Logps/chosen: -322.9527
• Logits/rejected: -0.8949
• Logits/chosen: -1.0172

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: 2
• total_train_batch_size: 8
• 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
0.6659	1.0	4424	0.6915	0.0201	0.0161	0.5510	0.0040	0.0645	-0.0595	0.0411	-314.0515	-322.9527	-0.8949	-1.0172

Framework versions

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