tulu2-13b-cost-tulumix-5e-6

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: 0.9352
• Rewards/chosen: -3.1557
• Rewards/rejected: -4.3037
• Rewards/accuracies: 0.6135
• Rewards/margins: 1.1480
• Rewards/margins Max: 6.7338
• Rewards/margins Min: -3.1653
• Rewards/margins Std: 3.2319
• Logps/rejected: -746.0304
• Logps/chosen: -640.5291
• Logits/rejected: -1.0143
• Logits/chosen: -1.1338

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-06
• 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
0.1926	1.0	3174	0.9352	-3.1557	-4.3037	0.6135	1.1480	6.7338	-3.1653	3.2319	-746.0304	-640.5291	-1.0143	-1.1338

Framework versions

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