Downloads · 30 days
12
2% of all-time downloads
bert_emails
kudoshinichi/bert_emails
bert_emails is a text classification model from kudoshinichi. Use it when you need a label for a piece of text. It is set up for transformers. The card lists the license as mit.
The model fine-tunes the bert-base-cased model. It is able to classify text into one of 150 intent classes. The goal is to use this model to classify intents in user input (e.g. emails, customer service, search queries).
All-time downloads
633
Public
Repo size
1.3 GB
Likes
1
Public
Repo makeup
Click a slice to open those files.
.bin434 MB · 100%
At a glance
- Task
- Text Classification
- Library
- transformers
- License
- mit
- Model type
- bert
- Access
- Public
- Created
- Sep 22, 2023
- Updated
- Sep 22, 2023
- SHA
- 9e6576bc
Try a prompt
Datasets
Model card
From the Hugging Face model README
The model fine-tunes the bert-base-cased model. It is able to classify text into one of 150 intent classes. The goal is to use this model to classify intents in user input (e.g. emails, customer service, search queries).
The model is trained over 3 epochs on a subset of the training set (2000/7500). This yielded a performance of 77.6% on the test set. By using the full dataset, the accuracy is expected to increase drastically. For reference, the model accuracy improved from 32% to 78% by moving to training on 1000/7500 of the dataset to 2000/7500. Hence, there is significant potential to fine-tune a model for intent classification if the full training set is used. The author did not attempt training on the full dataset due to time constraints.
- Task
- Text Classification
- Library
- transformers
- Type
- bert
- License
- mit
- Languages
- en
- Created
- Sep 22, 2023
- Updated
- Sep 22, 2023