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: 2.6421

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
1.6462	1.25	25	1.6200
0.6375	2.5	50	1.7340
0.3765	3.75	75	1.7938
0.1611	5.0	100	2.0300
0.0918	6.25	125	2.1064
0.0293	7.5	150	2.3839
0.022	8.75	175	2.4332
0.0233	10.0	200	2.3058
0.0147	11.25	225	2.5775
0.0098	12.5	250	2.3925
0.0046	13.75	275	2.4783
0.0062	15.0	300	2.5364
0.0036	16.25	325	2.5107
0.0023	17.5	350	2.5617
0.0021	18.75	375	2.5967
0.0023	20.0	400	2.6159
0.0024	21.25	425	2.6281
0.0017	22.5	450	2.6365
0.0019	23.75	475	2.6405
0.0019	25.0	500	2.6421

Framework versions

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