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: 2.5135

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
1.7173	1.25	25	1.6227
0.8099	2.5	50	1.5926
0.3572	3.75	75	1.7290
0.1907	5.0	100	1.8269
0.101	6.25	125	2.0608
0.0383	7.5	150	2.2592
0.0268	8.75	175	2.1757
0.0202	10.0	200	2.1768
0.0159	11.25	225	2.2593
0.0063	12.5	250	2.2801
0.004	13.75	275	2.3952
0.0032	15.0	300	2.4333
0.0031	16.25	325	2.4696
0.0021	17.5	350	2.4861
0.0025	18.75	375	2.4947
0.0022	20.0	400	2.5028
0.002	21.25	425	2.5074
0.0019	22.5	450	2.5114
0.0017	23.75	475	2.5130
0.0022	25.0	500	2.5135

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1