xlm-roberta-large_massive_crf_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: 5.3967
• Slot P: 0.7375
• Slot R: 0.7801
• Slot F1: 0.7582
• Slot Exact Match: 0.7260
• Intent Acc: 0.8687

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	19.9807	0.0	0.0	0.0	0.3187	0.0649
84.6375	2.0	90	9.4999	0.5191	0.5	0.5094	0.4634	0.3453
26.6202	3.0	135	4.9217	0.6025	0.7060	0.6502	0.6119	0.7708
11.9124	4.0	180	4.0265	0.6567	0.7338	0.6931	0.6749	0.8411
7.1611	5.0	225	3.6002	0.6822	0.7647	0.7211	0.6985	0.8569
5.4223	6.0	270	3.6653	0.7199	0.7687	0.7435	0.7191	0.8578
4.1914	7.0	315	3.6212	0.7156	0.7711	0.7423	0.7201	0.8628
3.4499	8.0	360	3.9382	0.7019	0.7836	0.7405	0.7078	0.8662
2.8646	9.0	405	4.0638	0.7059	0.7856	0.7436	0.7093	0.8652
2.3159	10.0	450	4.1920	0.7117	0.7751	0.7421	0.7127	0.8682
2.3159	11.0	495	4.3891	0.7110	0.7736	0.7410	0.7142	0.8731
1.8513	12.0	540	4.4429	0.7295	0.7821	0.7549	0.7231	0.8741
1.6003	13.0	585	4.7107	0.7317	0.7841	0.7570	0.7211	0.8775
1.3617	14.0	630	4.8732	0.7311	0.7751	0.7525	0.7231	0.8721
1.0467	15.0	675	5.0702	0.7230	0.7816	0.7511	0.7211	0.8687
0.9095	16.0	720	5.1884	0.7377	0.7935	0.7646	0.7304	0.8775
0.7422	17.0	765	5.1458	0.7267	0.7831	0.7538	0.7182	0.8711
0.6169	18.0	810	5.5018	0.7326	0.7836	0.7572	0.7231	0.8711
0.5221	19.0	855	5.3967	0.7375	0.7801	0.7582	0.7260	0.8687

Framework versions

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