xlm-roberta-base_massive_crf_v1

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

• Loss: 4.4117
• Slot P: 0.6934
• Slot R: 0.7706
• Slot F1: 0.7300
• Slot Exact Match: 0.6995
• Intent Acc: 0.8495

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	22.8757	0.0	0.0	0.0	0.3187	0.0300
95.1993	2.0	90	15.1787	0.3194	0.2164	0.2580	0.3015	0.1117
36.1644	3.0	135	10.7793	0.4180	0.4502	0.4335	0.4506	0.1864
24.5568	4.0	180	7.5359	0.5813	0.6333	0.6062	0.5706	0.3586
16.5092	5.0	225	5.7306	0.6266	0.7020	0.6621	0.6203	0.5957
11.609	6.0	270	4.9020	0.6610	0.7363	0.6966	0.6626	0.7280
8.4757	7.0	315	4.4249	0.6701	0.7448	0.7055	0.6744	0.7762
6.8454	8.0	360	4.3691	0.6841	0.7532	0.7170	0.6960	0.7973
5.6898	9.0	405	4.4460	0.6747	0.7647	0.7169	0.6886	0.8141
4.6831	10.0	450	4.2133	0.7067	0.7552	0.7302	0.7073	0.8342
4.6831	11.0	495	4.4300	0.6954	0.7542	0.7236	0.6995	0.8347
3.9992	12.0	540	4.3942	0.6977	0.7637	0.7292	0.7024	0.8416
3.5154	13.0	585	4.4117	0.6934	0.7706	0.7300	0.6995	0.8495

Framework versions

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