legalcase_outcomepred_model_v1

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.3580
• Accuracy: 0.3340

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.4956	0.9981	132	2.0711	0.3174
1.5006	1.9962	264	2.0215	0.2848
1.4925	2.9943	396	2.0069	0.2796
1.429	4.0	529	1.9503	0.2947
1.2188	4.9981	661	2.1001	0.3240
1.0163	5.9962	793	2.1491	0.3297
0.8554	6.9943	925	2.2008	0.3236
0.7692	8.0	1058	2.2889	0.3316
0.7553	8.9981	1190	2.3550	0.3349
0.6845	9.9811	1320	2.3580	0.3340

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0
• Datasets 3.0.0
• Tokenizers 0.19.1