| { |
| "model_type": "modernbert-token-classifier", |
| "base_model": "answerdotai/ModernBERT-large", |
| "num_labels": 3, |
| "hidden_size": 1024, |
| "max_seq_length": 8192, |
| "entity_types": [ |
| "Dataset" |
| ], |
| "entity_description": "Names of datasets, corpus, collections, databases, benchmarks used in scientific research", |
| "bio_labels": [ |
| "O", |
| "B-Dataset", |
| "I-Dataset" |
| ], |
| "label2id": { |
| "O": 0, |
| "B-Dataset": 1, |
| "I-Dataset": 2 |
| }, |
| "hyperparameters": { |
| "encoder_lr": 5e-05, |
| "head_lr": 0.001, |
| "dropout": 0.1, |
| "batch_size": 1, |
| "grad_accum_steps": 4, |
| "effective_batch_size": 4, |
| "num_epochs": 10, |
| "warmup_ratio": 0.1, |
| "weight_decay": 0.01, |
| "freeze_layers": 0, |
| "precision": "bf16", |
| "gpu": "NVIDIA GeForce RTX 5070 Ti" |
| }, |
| "training_data": { |
| "source": "data-annotations.json", |
| "ood_source": "151-eval.json", |
| "total_documents": 542, |
| "ood_documents": 151, |
| "train_examples": 2324, |
| "val_examples": 309, |
| "test_examples": 466, |
| "ood_examples": 1142, |
| "chunk_size_chars": 6000, |
| "chunk_overlap_chars": 500, |
| "negative_sampling": true, |
| "negative_sample_ratio": 0.15, |
| "augmentation": true, |
| "augmentation_window_size": 4000, |
| "augmentation_max_per_doc": 3 |
| } |
| } |
