metadata

library_name: peft
license: llama3.2
base_model: meta-llama/Llama-3.2-3B
tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: LLama3-3B-finetuning
    results: []

LLama3-3B-finetuning

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: 0.4744
Accuracy: 0.8077
F1 Macro: 0.8046

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: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro
2.4745	1.0	149	1.1147	0.5471	0.5000
1.0995	2.0	298	0.5006	0.8047	0.8023
0.7117	3.0	447	0.4292	0.8215	0.8218
0.5494	4.0	596	0.4039	0.8468	0.8469
0.4276	5.0	745	0.3924	0.8418	0.8415
0.2451	6.0	894	0.4101	0.8300	0.8312

Framework versions

PEFT 0.14.0
Transformers 4.47.1
Pytorch 2.5.1+cu121
Datasets 3.2.0
Tokenizers 0.21.0