final_wo_mrp_0.5_seed52_no_ckpy_ep4

final_wo_mrp_0.5_seed52_no_ckpy_ep4

xlm-roberta-base Finetuning

Model Description

Base Model: xlm-roberta-base

Intermediate Task:

Pre-Finetuned Model: xlm-roberta-base

Wandb Run URL: https://wandb.ai/s_haturusinghe-university-of-kelaniya/baseline-without-augmented-2-seed52-no-pre/runs/gg0yltmw

Training Information

Epochs: 4

Steps: 0

Validation Time: 0.01s

Performance Metrics

Main Metrics

Validation Loss: 0.332881

Validation Accuracy: 0.860000

Validation F1: 0.853412

Classification Report

{'0': {'precision': 0.8701298701298701, 'recall': 0.8993288590604027, 'f1-score': 0.8844884488448845, 'support': 447.0}, '1': {'precision': 0.84375, 'recall': 0.801980198019802, 'f1-score': 0.8223350253807107, 'support': 303.0}, 'accuracy': 0.86, 'macro avg': {'precision': 0.856939935064935, 'recall': 0.8506545285401024, 'f1-score': 0.8534117371127976, 'support': 750.0}, 'weighted avg': {'precision': 0.8594724025974025, 'recall': 0.86, 'f1-score': 0.8593784657653583, 'support': 750.0}}

Trainer Arguments:

                "learning_rate": 2e-05
                "epochs": 5
                "batch_size": 16
                "model": xlm-roberta-base
                "intermediate_task": False
                "n_tk_label": 2
                "mask_ratio": 0.5
                "seed": 52
                "dataset": sold
                "finetuning_stage": final
                "val_int": 10000
                "patience": 3
                "skip_empty_rat": False