IE_M2_50steps_1e7rate_03beta_SFT

This model is a fine-tuned version of tsavage68/IE_M2_1000steps_1e7rate_SFT on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3782
• Rewards/chosen: -0.1103
• Rewards/rejected: -2.3818
• Rewards/accuracies: 0.4600
• Rewards/margins: 2.2716
• Logps/rejected: -48.9613
• Logps/chosen: -42.5731
• Logits/rejected: -2.9040
• Logits/chosen: -2.8424

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: 1e-07
• train_batch_size: 2
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• training_steps: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.4682	0.4	50	0.3782	-0.1103	-2.3818	0.4600	2.2716	-48.9613	-42.5731	-2.9040	-2.8424

Framework versions

• Transformers 4.44.2
• Pytorch 2.0.0+cu117
• Datasets 3.0.0
• Tokenizers 0.19.1