Update model folder

.gitignore

@@ -1,3 +1,7 @@
+# GGML Models
+models/*.bin
+
+
 # Byte-compiled / optimized / DLL files
 __pycache__/
 *.py[cod]

models/model_download.txt

@@ -0,0 +1 @@
+Download the desired quantized MPT-7B-Instruct model from https://huggingface.co/TheBloke/MPT-7B-Instruct-GGML/tree/main

notebooks/main_CTransformers_LangChain.ipynb

@@ -131,7 +131,7 @@
   {
    "cell_type": "code",
    "execution_count": 8,
-   "id": "da64c2ce",
+   "id": "03928a6c",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -360,7 +360,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.4"
+   "version": "3.9.12"
   }
  },
  "nbformat": 4,

pyproject.toml

@@ -22,6 +22,7 @@ ipykernel = "^6.23.1"
 ctransformers = "^0.2.5"
 fastapi = "^0.96.0"
 uvicorn = "^0.22.0"
+ctransformers
 
 [build-system]
 requires = ["poetry-core"]

src/prompts.py

@@ -6,7 +6,7 @@
 # Note that the spacing and indentation of the prompt template is important for MPT-7B-Instruct, as it is highly sensitive to these
 # whitespace changes. For example, it could have problems generating a summary from the pieces of context
 
-mpt_7b_qa_template = """You are an expert HR assistant. Use the following pieces of information to answer the user's question.
+qa_template = """You are an expert HR assistant. Use the following pieces of information to answer the user's question.
 If you don't know the answer, just say that you don't know, don't try to make up an answer.
 
 Context: {context}

src/utils.py

@@ -10,8 +10,7 @@
 from langchain.chains import RetrievalQA
 from langchain.embeddings import HuggingFaceEmbeddings
 from langchain.vectorstores import FAISS
-from src.prompts import mpt_7b_qa_template
-from src.utils import set_qa_prompt, build_retrieval_qa
+from src.prompts import qa_template
 from src.llm import build_llm
 
 # Import config vars