Langchain integration

README.md

@@ -1,10 +1,11 @@
+![Banner](https://raw.githubusercontent.com/pavanjava/bootstrap-rag/refs/heads/main/assets/bootstrap-rag.png)
 # bootstrap-rag
 this project will bootstrap and scaffold the projects for specific semantic search and RAG applications along with regular boiler plate code.
 
 ### Architecture
 ![Arch](assets/architecture.png)
 
-### prerequisite
+### Installing prerequisite
 #### Option-1
 - install ollama following this [guide](https://ollama.com/download)
 - install qdrant follwing this [guide](https://qdrant.tech/documentation/guides/installation/)
@@ -36,8 +37,8 @@ Note: Only `llamaindex` and `Qdrant Search` are functional for now, others frame
 
 #### Resources
 
-![Demo GIF](assets/demo.gif)
+![Demo GIF](https://raw.githubusercontent.com/pavanjava/bootstrap-rag/refs/heads/main/assets/demo.gif)
 
-![Qdrant](assets/qdrant.png)
+![Qdrant](https://raw.githubusercontent.com/pavanjava/bootstrap-rag/refs/heads/main/assets/qdrant.png)
 
-![Arize Phoenix](assets/observability.png)
+![Arize Phoenix](https://raw.githubusercontent.com/pavanjava/bootstrap-rag/refs/heads/main/assets/observability.png)

bootstraprag/templates/langchain/readme.md

@@ -1 +1,6 @@
-### under development
+## LangChain Templates
+This section of the project has stable PoC Ready various RAG codes starting from simple to advanced techniques along with the LangGraph and LangSmith.
+Every LangChain code base has its own readme.md for more specific guidance.
+Go to a specific project and follow the respective readme.md fore more details.
+As a backend layer all the LangChain projects uses Qdrant and for LLM the project templates use Ollama.
+These can be just replaced by their substitutes to make the code working normally.

bootstraprag/templates/langchain/simple_rag/main.py

@@ -0,0 +1,30 @@
+from langchain_community.embeddings.fastembed import FastEmbedEmbeddings
+from langchain_qdrant import QdrantVectorStore
+from langchain_community.document_loaders import WebBaseLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from qdrant_client import QdrantClient
+from qdrant_client.http.models import Distance, VectorParams
+
+embeddings = FastEmbedEmbeddings()
+
+client = QdrantClient(url='http://localhost:6333', api_key='th3s3cr3tk3y')
+
+client.create_collection(
+    collection_name="demo_collection_1",
+    vectors_config=VectorParams(size=384, distance=Distance.COSINE),
+)
+
+vector_store = QdrantVectorStore(
+    client=client,
+    collection_name="demo_collection_1",
+    embedding=embeddings
+)
+
+WEBSITE_URL = "https://medium.com/@jaintarun7/multimodal-using-gemini-and-llamaindex-f622a190cc32"
+data = WebBaseLoader(WEBSITE_URL)
+docs = data.load()
+
+text_split = RecursiveCharacterTextSplitter(chunk_size=512,chunk_overlap=50)
+chunks = text_split.split_documents(docs)
+
+qdrant = vector_store.from_documents(chunks, embedding=embeddings, api_key='th3s3cr3tk3y')

bootstraprag/templates/langchain/simple_rag/requirements.txt

@@ -0,0 +1,7 @@
+langchain==0.3.3
+langchain-core==0.3.10
+langchain-qdrant==0.1.4
+langchain-ollama==0.2.0
+langchain-community==0.3.2
+qdrant-client==1.12.0
+fastembed==0.3.6

bootstraprag/templates/llamaindex/readme.md

@@ -0,0 +1,6 @@
+## LlamaIndex Templates
+This section of the project has stable PoC Ready various RAG codes starting from simple to advanced techniques along with the LlamaDeploy and llamaAgents. 
+Every LlamaIndex code base has its own readme.md for more specific guidance. 
+Go to a specific project and follow the respective readme.md fore more details. 
+As a backend layer all the LlamaIndex projects uses Qdrant and for LLM the project templates use Ollama. 
+These can be just replaced by their substitutes to make the code working normally.

bootstraprag/templates/qdrant/retrieval_quality/measure_retrieval_quality.py

@@ -8,6 +8,7 @@
 
 class MeasureRetrievalQuality:
     def __init__(self, dataset_path: str, collection_name: str, streaming: bool = True):
+        _ = load_dotenv(find_dotenv())
         # path = "Qdrant/arxiv-titles-instructorxl-embeddings"
         dataset = load_dataset(
             path=dataset_path, split="train", streaming=True,