waxal-lid-pseudo

This model is a fine-tuned version of sulaimank/waxal-lid on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0158
• Model Preparation Time: 0.0071
• Acc: 0.9967

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0.1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Model Preparation Time	Acc
0.0763	0.2501	500	0.0483	0.0071	0.9903
0.0335	0.5003	1000	0.0199	0.0071	0.9948
0.0327	0.7504	1500	0.0231	0.0071	0.9962
0.0211	1.0005	2000	0.0155	0.0071	0.9957
0.0101	1.2506	2500	0.0155	0.0071	0.9960
0.0122	1.5008	3000	0.0143	0.0071	0.9962
0.0105	1.7509	3500	0.0129	0.0071	0.9965
0.0052	2.0010	4000	0.0137	0.0071	0.9965
0.0054	2.2511	4500	0.0155	0.0071	0.9972
0.0039	2.5013	5000	0.0168	0.0071	0.9969
0.0139	2.7514	5500	0.0138	0.0071	0.9965
0.0008	3.0	5997	0.0158	0.0071	0.9967

Framework versions

• Transformers 5.13.1
• Pytorch 2.13.0+cu130
• Datasets 3.6.0
• Tokenizers 0.22.2