llama3_false_positives_1010_KTO_hp_screening_seeds

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2449
• Eval/rewards/chosen: -0.1725
• Eval/logps/chosen: -212.3262
• Eval/rewards/rejected: -0.2031
• Eval/logps/rejected: -235.8965
• Eval/rewards/margins: 0.0307
• Eval/kl: 0.0010

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-05
• train_batch_size: 1
• eval_batch_size: 2
• seed: 9012
• gradient_accumulation_steps: 8
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 6.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.7665	0.9412	12	1.3091	0.0694
0.7349	1.9608	25	1.2881	0.0051
0.7221	2.9804	38	1.2705	0.0002
0.6999	4.0	51	1.2533	0.0050
0.7169	4.9412	63	1.2452	0.0
0.5263	5.6471	72	1.2449	0.0010

Framework versions

• PEFT 0.11.1
• Transformers 4.44.0
• Pytorch 2.2.0
• Datasets 2.20.0
• Tokenizers 0.19.1