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: 2.5133

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
1.7585	1.25	25	1.5229
0.755	2.5	50	1.6669
0.3627	3.75	75	1.7460
0.1677	5.0	100	1.8874
0.0916	6.25	125	2.1126
0.0345	7.5	150	2.2322
0.0218	8.75	175	2.3397
0.0222	10.0	200	2.2314
0.016	11.25	225	2.1066
0.0061	12.5	250	2.3260
0.0049	13.75	275	2.2982
0.0037	15.0	300	2.3236
0.0028	16.25	325	2.4239
0.0025	17.5	350	2.4615
0.0025	18.75	375	2.4799
0.0025	20.0	400	2.4930
0.002	21.25	425	2.5027
0.0018	22.5	450	2.5087
0.0016	23.75	475	2.5120
0.0021	25.0	500	2.5133

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1