finetuned_ClinicalLongformer_negsampling_llama_2575

This model is a fine-tuned version of yikuan8/Clinical-Longformer on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1546
  • F1: 0.9281

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: 1e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 15

Training results

Training Loss Epoch Step Validation Loss F1
0.2101 1.0 76 0.2707 0.825
0.2161 2.0 152 0.1818 0.9178
0.372 3.0 228 0.1546 0.9281
0.1678 4.0 304 0.1724 0.9054
0.1615 5.0 380 0.2070 0.9079
0.1595 6.0 456 0.2347 0.9103
0.0517 7.0 532 0.2861 0.8936
0.08 8.0 608 0.3332 0.8732
0.1124 9.0 684 0.3660 0.8767
0.0279 10.0 760 0.2867 0.9028
0.0043 11.0 836 0.3631 0.8889
0.0299 12.0 912 0.3500 0.9041
0.0005 13.0 988 0.3502 0.9041
0.0027 14.0 1064 0.3698 0.8966
0.0004 15.0 1140 0.3696 0.9041

Framework versions

  • Transformers 4.48.0
  • Pytorch 2.5.1+cu124
  • Datasets 3.2.0
  • Tokenizers 0.21.0
Downloads last month
4
Safetensors
Model size
149M params
Tensor type
F32
·
Inference Providers NEW
This model is not currently available via any of the supported third-party Inference Providers, and the model is not deployed on the HF Inference API.

Model tree for npark95/finetuned_ClinicalLongformer_negsampling_llama_2575

Finetuned
(15)
this model