README.md

JW_chatbot

Python, Django, tensorflow

Virtual Environment Settings

* Python == 3.7
* Django == 3.1.1
* Tensorflow == 2.4.1
* scikit-learn == 0.24.1
* colorama == 0.4.4

Model

* Model: "sequential_2"
  _________________________________________________________________
  Layer (type)                 Output Shape              Param #   
  =================================================================
  embedding_2 (Embedding)      (None, 20, 16)            16000     
  _________________________________________________________________
  global_average_pooling1d_2 ( (None, 16)                0         
  _________________________________________________________________
  dense_6 (Dense)              (None, 16)                272       
  _________________________________________________________________
  dense_7 (Dense)              (None, 16)                272       
  _________________________________________________________________
  dense_8 (Dense)              (None, 8)                 136       
  =================================================================
  Total params: 16,680
  Trainable params: 16,680
  Non-trainable params: 0
  _________________________________________________________________
  
  
 * Epoch -> 490
 * 최종 loss: 0.5963 - accuracy: 0.8775

개발 노트

  * 채팅 환경 구축 완료 , 모델 로드 및 JWbot answering fuction 완료

  * Django 환경 구축 -> Flask사용 배포 (진행중)
  
  * 모델 성능 개선 연구중 (model on 21.02.13 = loss: 0.5963 - accuracy: 0.8775)