danish-legal-longformer-base-mlm

This model is a fine-tuned version of data/plms/danish-legal-longformer-base on the danish_legal_pile dataset. It achieves the following results on the evaluation set:

• Loss: 0.7374
• Accuracy: 0.8286

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• training_steps: 64000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7425	14.69	32000	0.7502	0.8259
0.7257	29.37	64000	0.7368	0.8287

Framework versions

• Transformers 4.18.0
• Pytorch 1.12.0+cu113
• Datasets 2.0.0
• Tokenizers 0.12.1