parent
08b771c863
commit
851f9b81c2
@ -0,0 +1,28 @@
|
||||
# Legal Chatbot
|
||||
|
||||
**Tier:** 3-Advanced
|
||||
|
||||
An AI powered legal assistant chatbot that is powered by a Large Language Model and is fine-tuned on local laws
|
||||
|
||||
- The main purpose of this application is to make complicate seeming laws accessile to the general public.
|
||||
- For this, one would require good knowledge of fine-tuning llms and data preprocessing along with access to datasets of local laws.
|
||||
|
||||
## User Stories
|
||||
|
||||
- [x] User can get a summary on any legal document
|
||||
- [x] User can ask any question from any of the local laws
|
||||
- [x] User can ask the AI powered chat-bot to create a legal document like rent agreement based on his own conditions
|
||||
- [x] User can chat with AI about any of the judicial processes with human-like responses like that of ChatGPT.
|
||||
|
||||
## Bonus features
|
||||
|
||||
- [ ] User has the option to contact an expert
|
||||
- [ ] The model can refer to past cases and offer legal advice based on past proceedings (would require an additional dataset of past cases)
|
||||
|
||||
## Useful links and resources
|
||||
|
||||
[Fine-tuning LLMs using PEFT and QLORA](https://medium.com/@ud.chandra/instruction-fine-tuning-llama-2-with-pefts-qlora-method-d6a801ebb19)
|
||||
|
||||
## Example projects
|
||||
|
||||
[Lilith AI](https://lilithai.tech/)
|
Loading…
Reference in new issue