Merge remote-tracking branch 'workshop-llm/main'

.gitignore

@@ -4,5 +4,7 @@
 .local
 .npm
 .cache
+.env
 shared
-data/model
+data/model
+__pycache__

.vscode/settings.json

@@ -0,0 +1,13 @@
+{
+    "objectscript.conn": {
+        "username": "superuser",
+        "password": "SYS",
+        "active": true,
+        "ns": "LLMRAG",
+        "host": "localhost",
+        "port": 52774,
+        "docker-compose": {
+            "service": "iris"
+        }
+    }
+}

LICENSE

@@ -1,6 +1,6 @@
 MIT License
 
-Copyright (c) 2019 InterSystems Developer Community
+Copyright (c) 2024 InterSystems Developer Community
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,8 +1,11 @@
 # workshop-llm
-Workshop to create a RAG application to use LLM models. This workshop is developed in Python using Jupyter Notebook connected using the official libraries to IRIS.
+Workshop to create a RAG application using LLM models. 
 
-You can find more in-depth information in https://learning.intersystems.com.
+This workshop is developed in Python 🐍 (Jupyter Notebook) and InterSystems IRIS.
+
+The main purpose is to show you the main steps to create a RAG application using an LLM and a vector database.
 
+You can find more in-depth information in https://learning.intersystems.com.
 
 # What do you need to install? 
 * [Git](https://git-scm.com/downloads) 
@@ -13,42 +16,97 @@ You can find more in-depth information in https://learning.intersystems.com.
 # Setup
 Build the image we will use during the workshop:
 
-```console
-$ git clone https://github.com/intersystems-ib/workshop-llm
-$ cd workshop-llm
-$ docker-compose build
+Clone the repository:
+```bash
+git clone https://github.com/intersystems-ib/workshop-llm
+cd workshop-llm
+```
+
+Build the image:
+```bash
+docker compose build
+```
+
+Run the containers:
+```bash
+docker compose up -d
 ```
 
-# Configuration
+After running the containers, you should be able to access to:
+* InterSystems IRIS [Management Portal](http://localhost:52774/csp/sys/UtilHome.csp). You can login using `superuser` / `SYS`
+* [Jupyter Notebook](http://localhost:8888) 
+
+# Explore RAG applications using Jupyter
+
+## Medicine Leaflet examples
+
+You have some medicine leaflets (in spanish) in [./data](./data).
+
+This example is about creating a RAG Q&A application that can answer questions about those medicine leaflets.
+
+Open [Jupyter Notebook](http://localhost:8888), there you can find:
+* [QA-PDF-LLM.ipynb](./jupyter/QA-PDF-LLM.ipynb) - RAG example using [MistralAI](https://mistral.ai) LLM 
+* [QA-PDF-Local.ipynb](./jupyter/QA-PDF-Local.ipynb) - RAG example using a local LLM
+
+![alt text](/images/jupyter.png)
+
+You can test the project step by step or execute it at one time, feel free.
 
-The main purpose of this example is to identify the main steps to create a RAG application using MISTRAL as LLM and IRIS as vector database to save and search the specific context.
+## Hoolefoods data model text to SQL
 
-## **ATTENTION** Docker configuration
+This example is about a company called Holefoods that sells food with some hole on it :)
 
-There is a known issue related to the permission request from Docker Desktop to access to the folders of the project, this permission has to be granted before to launch **docker-compose up -d**. To allow the file sharing in Docker Desktop you have to open settings option, select **Resources** and **File Sharing**, from that screen you have to include the path to the project, you can see here an example:
-![alt text](/images/fileSharing.png)
+Using the sales data model of the company, the goal is to create an assistant that can translate natural language questions into valid SQL that answer the question.
 
-If you don't share this folder previously PostgreSQL database won't be initialized and the project will fail.
+In [Jupyter Notebook](http://localhost:8888), you will find:
+* [QA-SQL-LLM.ipynb](./jupyter/QA-SQL-LLM.ipynb) - text to SQL example using OpenAI LLM.
 
-## Test Production 
-* Run the containers that we will use in the workshop:
+# Create other applications
+
+There are some other examples you can try to build and modify in your local environment.
+
+First of all, create a new environment and install some requirements:
+
+```bash
+# create a local venv environment
+python3 -m venv .venv
+
+# activate venv
+source .venv/bin/activate
+
+# install dependencies
+pip3 install -r requirements.txt
+```
+
+Create an `.env` file for storing API keys for OpenAI / MistralAI. They will be used in the applications.
+
+```
+OPENAI_API_KEY="your-api"
+MISTRAL_API_KEY="your-api"
 ```
-docker-compose build
 
-docker-compose up -d
+## Text to SQL service API 
+You can find a sample Text to SQL based on [QA-SQL-LLM.ipynb](./jupyter/QA-SQL-LLM.ipynb) [here](python/holefoods_text2sql/main.py).
+
+You can run it like this:
+
+```bash
+cd python/holefoods_text2sql
+fastapi dev main.py
 ```
-Automatically an IRIS instance, will be deployed, a Jupyter Notebook is deployed under (http://localhost:8888) too.
 
-## IRIS database
+Then open http://127.0.0.1:8000/docs to explore the API and try it out using the web client.
 
-* Open the [Management Portal](http://localhost:52774/csp/sys/UtilHome.csp).
-* Login using the default `superuser`/ `SYS` account.
-* Open System Explorer --> SQL
-* Select NAMESPACE USER and Schema `Test`
+## Streamlit Assistant
+There is also a great example of a langchain / streamlit chatbot assitant in https://alejandro-ao.com/how-to-use-streaming-in-langchain-and-streamlit/
 
-# Testing with Jupyter Notebook
+You can play with it here as well:
 
-This project is devolped in Python using Jupyter Notebook, you can access to it from [here](http://localhost:8888) and open LLMTests.ipnyb file.
-![alt text](/images/jupyter.png)
+```bash
+cd python/assitant
+streamlit run chatbot.py
+```
+
+Then open http://localhost:8501 and have a look at it.
 
-You can test the project step by step or execute it at one time, feel free.
+Are you able to add the logic to reproduce the Medicine Leaflet example in the assitant ?

docker-compose.yml

@@ -2,7 +2,6 @@ services:
 
   # iris
   iris:
-    init: true
     container_name: iris
     build:
       context: .
@@ -14,7 +13,6 @@ services:
     - ./iris/shared:/shared
     environment:
     - ISC_DATA_DIRECTORY=/shared/durable
-    command: --check-caps false --ISCAgent false
 
   # jupyter notebook
   jupyter:

iris/Dockerfile

@@ -11,15 +11,10 @@ USER irisowner
 # copy files to image
 WORKDIR /opt/irisapp
 COPY --chown=irisowner:irisowner src src
+COPY --chown=irisowner:irisowner iris/iris.script iris.script
+COPY --chown=irisowner:irisowner iris/sql sql
 
-COPY --chown=$ISC_PACKAGE_MGRUSER:$ISC_PACKAGE_IRISGROUP /iris/irissession.sh /
-RUN chmod +x /irissession.sh
-
-SHELL ["/irissession.sh"]
-
-RUN \
-zn "%SYS" \
-set sc = $SYSTEM.OBJ.LoadDir("/opt/irisapp/src/Install", "ck", , 1) \
-if sc do ##class(LLMRAG.Installer).Install() \
-zn "LLMRAG" \
-DO $SYSTEM.OBJ.LoadDir("/opt/irisapp/src/LLMRAG", "ck", , 1) 
+# run iris.script
+RUN iris start IRIS \
+    && iris session IRIS < /opt/irisapp/iris.script \
+    && iris stop IRIS quietly

iris/iris.script

@@ -0,0 +1,21 @@
+
+    // unexpire default passwords
+    set $namespace = "%SYS"
+    do ##class(Security.Users).UnExpireUserPasswords("*")
+
+    // create namespace
+    do $SYSTEM.OBJ.Load("/opt/irisapp/src/LLMRAG/Installer.cls", "ck")
+    do ##class(LLMRAG.Installer).Install()
+
+    // import source code & schemas
+    set $namespace="LLMRAG"
+    do $SYSTEM.SQL.Schema.ImportDDL("/opt/irisapp/sql/dump.sql",,"IRIS")
+
+    // install zpm
+    s r=##class(%Net.HttpRequest).%New(),r.Server="pm.community.intersystems.com",r.SSLConfiguration="ISC.FeatureTracker.SSL.Config" d r.Get("/packages/zpm/latest/installer"),$system.OBJ.LoadStream(r.HttpResponse.Data,"c")
+
+    // install some sample data (Holefoods, etc.)
+    zpm "install samples-bi"
+
+    // load some modifications on Holefoods (e.g. ForeignKeys, etc.)
+    do $SYSTEM.OBJ.LoadDir("/opt/irisapp/src/HoleFoods/", "ck", ,1)

jupyter/Dockerfile

@@ -1,13 +1,9 @@
 FROM jupyter/minimal-notebook
 
-USER root
-
+# install IRIS DBAPI driver
 COPY  /jupyter/intersystems_irispython-3.2.0-py3-none-any.whl /
-RUN chmod 777 /intersystems_irispython-3.2.0-py3-none-any.whl
-
-COPY  /jupyter/requirements.txt /
-RUN chmod 777 /requirements.txt
-
 RUN pip install /*.whl
 
+# install python requirements
+COPY  /jupyter/requirements.txt /
 RUN pip install -r /requirements.txt