CafeBERT_massive_v2

This model is a fine-tuned version of uitnlp/CafeBERT on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9485
• Slot P: 0.0090
• Slot R: 0.0199
• Slot F1: 0.0124
• Slot Exact Match: 0.0418
• Intent Acc: 0.8647

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: 5e-05
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 256
• 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
• lr_scheduler_warmup_ratio: 0.06
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Slot P	Slot R	Slot F1	Slot Exact Match	Intent Acc
No log	1.0	45	4.1897	0.0	0.0	0.0	0.4088	0.3478
12.6262	2.0	90	1.3454	0.0090	0.0129	0.0106	0.1943	0.8406
4.1265	3.0	135	1.0228	0.0102	0.0193	0.0134	0.1181	0.8564
1.987	4.0	180	0.9293	0.0100	0.0188	0.0130	0.1313	0.8692
1.382	5.0	225	0.9144	0.0096	0.0205	0.0131	0.0644	0.8706
1.0689	6.0	270	0.9485	0.0090	0.0199	0.0124	0.0418	0.8647

Framework versions

• Transformers 4.55.0
• Pytorch 2.7.0+cu126
• Datasets 3.6.0
• Tokenizers 0.21.4