Beta
Explore
Marketplace
Neural Labs
Chat
Wallet
Docs
NHS-pubmedbert-binary – AI Model by intermezzo672 | AlphaNeural AI
You can deploy this model and start earning money today!
intermezzo672
/
NHS-pubmedbert-binary
like
0
transformers
tensorboard
safetensors
bert
text-classification
generated_from_trainer
microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract
finetune
mit
autotrain_compatible
endpoints_compatible
us
Views
No views yet
Model card
Files and Versions
Community
API
Deploy
NHS-pubmedbert-binary
This model is a fine-tuned version of
microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract
on the None dataset. It achieves the following results on the evaluation set:
Loss: 0.5861
Accuracy: 0.8246
Precision: 0.8180
Recall: 0.8240
F1: 0.8202
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: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 6
Training results
Training Loss
Epoch
Step
Validation Loss
Accuracy
Precision
Recall
F1
0.0592
1.0
397
0.3980
0.8246
0.8179
0.8220
0.8196
0.0542
2.0
794
0.4714
0.7943
0.7960
0.8064
0.7929
0.6307
3.0
1191
0.5861
0.8246
0.8180
0.8240
0.8202
Framework versions
Transformers 4.38.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.15.2
