llama381binstruct_summarize_short

This model is a fine-tuned version of NousResearch/Meta-Llama-3.1-8B-Instruct on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 1.4158

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: 0.0002
• train_batch_size: 1
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 30
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
1.6861	1.1905	25	0.9223
0.7859	2.3810	50	0.8779
0.3887	3.5714	75	0.9867
0.1412	4.7619	100	1.0822
0.0911	5.9524	125	1.2118
0.0391	7.1429	150	1.3553
0.0309	8.3333	175	1.2825
0.0188	9.5238	200	1.2512
0.0145	10.7143	225	1.2936
0.0091	11.9048	250	1.3109
0.0058	13.0952	275	1.2768
0.0042	14.2857	300	1.2963
0.0032	15.4762	325	1.3539
0.0021	16.6667	350	1.3810
0.0024	17.8571	375	1.3974
0.0021	19.0476	400	1.4047
0.002	20.2381	425	1.4103
0.0018	21.4286	450	1.4133
0.0017	22.6190	475	1.4152
0.0015	23.8095	500	1.4158

Framework versions

• PEFT 0.13.2
• Transformers 4.46.1
• Pytorch 2.5.0+cu121
• Datasets 3.0.2
• Tokenizers 0.20.1