LLama3-3B-finetuning-italian-version – AI Model by msab97 | AlphaNeural AI

You can deploy this model and start earning money today!

msab97/
LLama3-3B-finetuning-italian-version

like

0

generated_from_trainer

meta-llama/Llama-3.2-3B

Views

No views yet

LLama3-3B-finetuning-italian-version

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: 0.4762
Model Preparation Time: 0.0067
Accuracy: 0.8388
F1 Macro: 0.8459

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: 3.9922302694761015e-05
train_batch_size: 16
eval_batch_size: 16
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: 0.0960656012161834
num_epochs: 7
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Accuracy	F1 Macro
0.4458	1.0	735	0.4802	0.0067	0.8034	0.8130
0.3723	2.0	1470	0.4281	0.0067	0.8184	0.8270
0.2744	3.0	2205	0.4820	0.0067	0.8218	0.8305
0.1358	4.0	2940	0.6884	0.0067	0.8156	0.8254

Framework versions

PEFT 0.14.0
Transformers 4.49.0
Pytorch 2.5.1+cu124
Datasets 3.3.2
Tokenizers 0.21.0