ModernBERT-large_massive_modernbert_large_crf_v1

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

• Loss: 15.5718
• Slot P: 0.5398
• Slot R: 0.6408
• Slot F1: 0.5860
• Slot Exact Match: 0.6001
• Intent Acc: 0.7831

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	43.3614	0.0	0.0	0.0	0.3178	0.0821
160.2669	2.0	90	27.2292	0.3143	0.2269	0.2635	0.3586	0.2548
66.654	3.0	135	19.2474	0.4379	0.4	0.4181	0.4481	0.4575
38.629	4.0	180	15.3625	0.4023	0.5408	0.4614	0.4801	0.5903
23.3498	5.0	225	12.4194	0.4446	0.5706	0.4998	0.5411	0.6695
12.7922	6.0	270	12.3227	0.5013	0.5980	0.5454	0.5691	0.6990
7.8613	7.0	315	12.8060	0.4926	0.6	0.5410	0.5642	0.7324
5.4037	8.0	360	12.9247	0.5086	0.6294	0.5626	0.5809	0.7388
3.6892	9.0	405	13.9871	0.5260	0.6343	0.5751	0.5986	0.7605
2.6797	10.0	450	14.0965	0.5562	0.6204	0.5865	0.6011	0.7742
2.6797	11.0	495	13.8520	0.5105	0.6398	0.5679	0.5775	0.7698
2.0031	12.0	540	15.0858	0.5491	0.6289	0.5863	0.6080	0.7698
1.3894	13.0	585	15.5718	0.5398	0.6408	0.5860	0.6001	0.7831

Framework versions

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