LKD_Experience

This model is a fine-tuned version of distilbert-base-uncased on an dataset of Reddit comments and posts related to Living Kidney Donation (LKD). This model identifies documents as either describing a personal experience with LKD, or simply sharing news like headlines and noise/nonsense. The first token/word in each document is the name of the subreddit where the post was written. It achieves the following results on the evaluation set:

• Loss: 0.1651
• Accuracy: 0.9574

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: 8
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	63	0.3311	0.8723
No log	2.0	126	0.1708	0.9574
No log	3.0	189	0.1651	0.9574

Framework versions

• Transformers 4.27.0
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2