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: 2.3230

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
1.616	1.25	25	1.4605
0.7705	2.5	50	1.4575
0.3691	3.75	75	1.5446
0.1573	5.0	100	1.6476
0.1067	6.25	125	1.7580
0.0391	7.5	150	2.0052
0.0248	8.75	175	2.0620
0.0178	10.0	200	2.0495
0.0142	11.25	225	2.0332
0.0044	12.5	250	2.1315
0.0043	13.75	275	2.2304
0.0034	15.0	300	2.2290
0.0047	16.25	325	2.2549
0.0027	17.5	350	2.2630
0.0023	18.75	375	2.2856
0.0028	20.0	400	2.3005
0.0025	21.25	425	2.3111
0.002	22.5	450	2.3187
0.0018	23.75	475	2.3216
0.0021	25.0	500	2.3230

Framework versions

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