cls_train_llama3_v1

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: 0.6452

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: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
0.7976	0.2146	50	0.8180
0.7427	0.4292	100	0.7618
0.7449	0.6438	150	0.7284
0.6912	0.8584	200	0.6968
0.5697	1.0730	250	0.6920
0.5641	1.2876	300	0.6837
0.5407	1.5021	350	0.6624
0.5387	1.7167	400	0.6548
0.5464	1.9313	450	0.6452

Framework versions

• PEFT 0.11.1
• Transformers 4.41.0
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1