ferrazzipietro/meshTask-gemma-3-1b-it-ood

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

meshTask-gemma-3-1b-it-ood

This model is a fine-tuned version of google/gemma-3-1b-it on an unknown dataset. It achieves the following results on the evaluation set:

• —Loss: 2.4488
• —F1 Micro: 0.2475
• —F1 Macro: 0.0098
• —F1 Weighted: 0.3698
• —Class/f1 Results Per Class: {}
• —Items/f1 Scores Per Item: {'Female': 0.08525899541766016, 'Child, Preschool': 0.041326728826728824, 'Retrospective Studies': 0.04188565058130275, 'Cell Proliferation': 0.06603927099153829, 'Aged, 80 and over': 0.029086255583612386, 'Infant': 0.06091162485653386, 'Rats': 0.03277777777777778, 'Adult': 0.06061887372669614, 'Prospective Studies': 0.029296535625649548, 'Young Adult': 0.0723076763936156, 'Cross-Sectional Studies': 0.03263983114729383, 'Middle Aged': 0.051254024461159516, 'Child': 0.06218699601052542, 'Reproducibility of Results': 0.026745329400196655, 'Follow-Up Studies': 0.019956092339876913}

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: 0.0003
• —trainbatchsize: 128
• —evalbatchsize: 32
• —seed: 42
• —distributed_type: multi-GPU
• —optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-07 and optimizerargs=No additional optimizer arguments
• —lrschedulertype: cosine
• —lrschedulerwarmup_ratio: 0.1
• —num_epochs: 1

Training results

Framework versions

• —PEFT 0.18.1
• —Transformers 4.51.0
• —Pytorch 2.8.0+cu128
• —Datasets 3.6.0
• —Tokenizers 0.21.0