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: 2.6704

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 30
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
1.7321	1.25	25	1.6900
0.8071	2.5	50	1.8501
0.3599	3.75	75	1.8955
0.1408	5.0	100	2.0946
0.1067	6.25	125	2.1324
0.0366	7.5	150	2.4187
0.025	8.75	175	2.3309
0.0215	10.0	200	2.3936
0.0094	11.25	225	2.5407
0.0063	12.5	250	2.4951
0.0035	13.75	275	2.5288
0.0051	15.0	300	2.5601
0.0042	16.25	325	2.5644
0.002	17.5	350	2.6005
0.002	18.75	375	2.6281
0.002	20.0	400	2.6451
0.0019	21.25	425	2.6574
0.0018	22.5	450	2.6649
0.0017	23.75	475	2.6694
0.0018	25.0	500	2.6704

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.4.0+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1