Fast Api Open ai chat bot with intent classification and Entity extraction
To run this project follow the command below.
./run.sh
Further work:
- Improve slot filling by training a custom slot fill.
- evaluate the intent mapping and validate which label is suitable for our use case
- Experiment and log prompts experiments for the open ai chatbot
- Run evaluation
- Improve the redis storage conversation from last to three or more conversations backward
- Set triggers for the intent and slot filling
Resources:
- https://github.com/NVIDIA/NeMo/blob/stable/tutorials/nlp/Joint_Intent_and_Slot_Classification.ipynb
- https://github.com/anuragmishra1/alter-nlu
- https://nvidia.github.io/NeMo/
- https://github.com/fractalego/zero-shot-relation-extractor
- https://github.com/sebastianruder/NLP-progress/blob/master/english/intent_detection_slot_filling.md
- https://paperswithcode.com/paper/massive-a-1m-example-multilingual-natural#code
- https://github.com/Rishabbh-Sahu/intent_and_slot_classification
- https://medium.com/holler-developers/slot-filling-using-sequence-models-78ecd928b300