best-model-v1

This model is a fine-tuned version of FacebookAI/xlm-roberta-large on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6069
• Accuracy: 0.4410
• Precision: 0.5118
• Recall: 0.6489
• F1 Score: 0.5711

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: 8
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 16
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_steps: 500
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1 Score
1.7357	0.3911	500	0.8218	0.3192	0.2798	0.2506	0.2551
1.5848	0.7822	1000	0.7275	0.3937	0.4789	0.4157	0.4387
1.398	1.1729	1500	0.6628	0.4123	0.4885	0.5846	0.5251
1.2999	1.5639	2000	0.6264	0.4295	0.5046	0.6402	0.5626
1.2964	1.9550	2500	0.6101	0.4260	0.4991	0.6553	0.5655
1.2479	2.3457	3000	0.6069	0.4381	0.5082	0.6506	0.5695
1.267	2.7368	3500	0.6069	0.4410	0.5118	0.6489	0.5711

Framework versions

• Transformers 4.47.1
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0