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: 2.3502

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
1.6893	1.3158	25	1.4438
0.7874	2.6316	50	1.5368
0.4613	3.9474	75	1.5765
0.152	5.2632	100	1.8493
0.08	6.5789	125	2.1247
0.0453	7.8947	150	2.0704
0.0233	9.2105	175	2.1779
0.0148	10.5263	200	2.1609
0.0112	11.8421	225	2.2039
0.0083	13.1579	250	2.2150
0.0081	14.4737	275	2.2033
0.0034	15.7895	300	2.2145
0.0025	17.1053	325	2.2772
0.0015	18.4211	350	2.3073
0.0015	19.7368	375	2.3235
0.0017	21.0526	400	2.3341
0.0012	22.3684	425	2.3415
0.0022	23.6842	450	2.3460
0.0015	25.0	475	2.3494
0.0011	26.3158	500	2.3502

Framework versions

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