feat: Get answers using preferred number of chunks

README.md

@@ -22,6 +22,7 @@ PERSIST_DIRECTORY: is the folder you want your vectorstore in
 MODEL_PATH: Path to your GPT4All or LlamaCpp supported LLM
 MODEL_N_CTX: Maximum token limit for the LLM model
 EMBEDDINGS_MODEL_NAME: SentenceTransformers embeddings model name (see https://www.sbert.net/docs/pretrained_models.html)
+TARGET_SOURCE_CHUNKS: The amount of chunks (sources) that will be used to answer a question
 ```
 
 Note: because of the way `langchain` loads the `SentenceTransformers` embeddings, the first time you run the script it will require internet connection to download the embeddings model itself.

example.env

@@ -2,4 +2,5 @@ PERSIST_DIRECTORY=db
 MODEL_TYPE=GPT4All
 MODEL_PATH=models/ggml-gpt4all-j-v1.3-groovy.bin
 EMBEDDINGS_MODEL_NAME=all-MiniLM-L6-v2
-MODEL_N_CTX=1000
+MODEL_N_CTX=1000
+TARGET_SOURCE_DOCUMENTS=4

privateGPT.py

@@ -16,6 +16,7 @@
 model_type = os.environ.get('MODEL_TYPE')
 model_path = os.environ.get('MODEL_PATH')
 model_n_ctx = os.environ.get('MODEL_N_CTX')
+target_source_chunks = int(os.environ.get('TARGET_SOURCE_CHUNKS',4))
 
 from constants import CHROMA_SETTINGS
 
@@ -24,7 +25,7 @@ def main():
     args = parse_arguments()
     embeddings = HuggingFaceEmbeddings(model_name=embeddings_model_name)
     db = Chroma(persist_directory=persist_directory, embedding_function=embeddings, client_settings=CHROMA_SETTINGS)
-    retriever = db.as_retriever()
+    retriever = db.as_retriever(search_kwargs={"k": target_source_chunks})
     # activate/deactivate the streaming StdOut callback for LLMs
     callbacks = [] if args.mute_stream else [StreamingStdOutCallbackHandler()]
     # Prepare the LLM