caretech-owl / gerd

Generating and evaluating relevant documentation (GERD)
https://towardsdatascience.com/running-llama-2-on-cpu-inference-for-document-q-a-3d636037a3d8
MIT License
4 stars 0 forks source link

[UC]: QA: A user can select a document and ask any question about it #17

Closed Depo14 closed 2 months ago

Depo14 commented 12 months ago

Summary

A user can select a document in the website's file explorer and then ask a question using the questiontextfield. After the user received the answer, he can ask another question. If he want to question another document he can select another one.

Rationale

To be able to ask a question about a document, you must first select the document. Especially if more than one document are uploaded. Thre must also be an option to ask the question and send it to the model. A textfield is used for this.

Level

user goal

Actors

User

Preconditions

Postconditions

An answer is displayed under the questiontextfield

Basic Flow

  1. The user selects the document he about which he has questions
  2. The user ask one or multiple questions
  3. If he want to ask the model about another document he can switch the document

Alternative Paths

  1. The user selects the document he want to ask questions
  2. An error is displayed that the document cannot be selected

Visualisation

flowchart LR;
  1[The user selects the document he about which he has questions]
  1-->2.1[The user ask one or multiple questions]
  1-->2.2[An error is displayed that the document cannot be selected]
  2.1-->3[If he want to ask the model about another document he can switch the document]
  3-->2.1
  3-->2.2

Other related issues, use cases, features

https://github.com/caretech-owl/Llama-2-Open-Source-LLM-CPU-Inference/issues/16