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: 2.4530

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
1.6176	1.25	25	1.4646
0.7045	2.5	50	1.5828
0.3188	3.75	75	1.7073
0.1796	5.0	100	1.8367
0.0882	6.25	125	2.0182
0.038	7.5	150	2.3088
0.0196	8.75	175	2.2476
0.0256	10.0	200	2.1310
0.0117	11.25	225	2.1974
0.0066	12.5	250	2.2862
0.0058	13.75	275	2.3694
0.0071	15.0	300	2.2468
0.0063	16.25	325	2.2986
0.0031	17.5	350	2.3789
0.0024	18.75	375	2.4145
0.003	20.0	400	2.4298
0.0025	21.25	425	2.4402
0.0021	22.5	450	2.4474
0.0022	23.75	475	2.4513
0.0019	25.0	500	2.4530

Framework versions

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