Llama-3-8B-sft-lora-en-tw

This model is a fine-tuned version of meta-llama/Meta-Llama-3-8B-Instruct on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 1.6918

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: 2e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss
2.0752	1.0	161	2.0532
1.9732	2.0	322	1.9112
1.8481	3.0	483	1.8377
1.7678	4.0	644	1.7926
1.7318	5.0	805	1.7617
1.6638	6.0	966	1.7381
1.6334	7.0	1127	1.7235
1.6117	8.0	1288	1.7130
1.6084	9.0	1449	1.7019
1.6187	10.0	1610	1.6975
1.6255	11.0	1771	1.6941
1.5094	12.0	1932	1.6934
1.5754	13.0	2093	1.6908
1.5813	14.0	2254	1.6916
1.5725	15.0	2415	1.6918

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.3.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1