prox-doc-Llama-3.2-3B

This model is a fine-tuned version of meta-llama/Llama-3.2-3B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.3655
• Accuracy: 0.8328

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0	0	1.2404	0.4912
0.6571	0.1002	502	0.7007	0.6618
0.4499	0.2003	1004	0.4616	0.7799
0.4078	0.3005	1506	0.4103	0.8091
0.3714	0.4006	2008	0.3913	0.8154
0.3897	0.5008	2510	0.3780	0.8256
0.3565	0.6010	3012	0.3748	0.8275
0.3413	0.7011	3514	0.3726	0.8282
0.3325	0.8013	4016	0.3667	0.8344
0.3722	0.9014	4518	0.3655	0.8328

Framework versions

• PEFT 0.15.2
• Transformers 4.51.3
• Pytorch 2.4.1+cu124
• Datasets 3.5.0
• Tokenizers 0.21.1