Add local models

app.py

@@ -18,6 +18,7 @@ from constants import (
     PROJECT_URL,
     TEMPERATURE,
     USAGE_HELP,
+    MODEL_N_CTX,
     K,
 )
 from utils import (
@@ -45,12 +46,12 @@ SESSION_DEFAULTS = {
     "usage": {},
     "chat_history": [],
     "generated": [],
-    "data_source": DEFAULT_DATA_SOURCE,
-    "uploaded_file": None,
     "auth_ok": False,
     "openai_api_key": None,
     "activeloop_token": None,
     "activeloop_org_name": None,
+    "uploaded_file": None,
+    "data_source": DEFAULT_DATA_SOURCE,
     "model": MODEL,
     "embeddings": EMBEDDINGS,
     "k": K,
@@ -59,6 +60,7 @@ SESSION_DEFAULTS = {
     "chunk_overlap": CHUNK_OVERLAP,
     "temperature": TEMPERATURE,
     "max_tokens": MAX_TOKENS,
+    "model_n_ctx": MODEL_N_CTX,
 }
 # Initialise session state variables
 for k, v in SESSION_DEFAULTS.items():

constants.py

@@ -15,6 +15,11 @@ TEMPERATURE = 0.7
 MAX_TOKENS = 3357
 ENABLE_ADVANCED_OPTIONS = True
 
+MODEL_N_CTX = 1000
+LLAMACPP_MODEL_PATH = ""
+GPT4ALL_MODEL_PATH = ""
+ENABLE_LOCAL_MODELS = False
+
 DATA_PATH = Path.cwd() / "data"
 DEFAULT_DATA_SOURCE = "https://github.com/gustavz/DataChad.git"
 

utils.py

@@ -37,8 +37,10 @@ from langchain.embeddings.openai import Embeddings, OpenAIEmbeddings
 from langchain.schema import Document
 from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain.vectorstores import DeepLake, VectorStore
+from langchain.llms import GPT4All, LlamaCpp
+from langchain.embeddings import HuggingFaceEmbeddings
 
-from constants import APP_NAME, DATA_PATH, PAGE_ICON, PROJECT_URL
+from constants import APP_NAME, DATA_PATH, PAGE_ICON, PROJECT_URL, LLAMACPP_MODEL_PATH, GPT4ALL_MODEL_PATH
 
 # loads environment variables
 load_dotenv()
@@ -190,7 +192,7 @@ WEB_LOADER_MAPPING = {
 }
 
 
-def get_loader(file_path: str, mapping: dict, default_loader:BaseLoader) -> BaseLoader:
+def get_loader(file_path: str, mapping: dict, default_loader: BaseLoader) -> BaseLoader:
     # Choose loader from mapping, load default if no match found
     ext = "." + file_path.rsplit(".", 1)[-1]
     if ext in mapping:
@@ -238,7 +240,7 @@ def load_data_source() -> List[Document]:
         st.stop()
 
 
-def get_data_source_string() -> str:
+def get_dataset_name() -> str:
     # replace all non-word characters with dashes
     # to get a string that can be used to create a new dataset
     dashed_string = re.sub(r"\W+", "-", st.session_state["data_source"])
@@ -254,6 +256,21 @@ def get_model() -> BaseLanguageModel:
                 temperature=st.session_state["temperature"],
                 openai_api_key=st.session_state["openai_api_key"],
             )
+        case "LlamaCpp":
+            model = LlamaCpp(
+                model_path=LLAMACPP_MODEL_PATH,
+                n_ctx=st.session_state["model_n_ctx"],
+                temperature=st.session_state["temperature"],
+                verbose=True,
+            )
+        case "GPT4All":
+            model = GPT4All(
+                model=GPT4ALL_MODEL_PATH,
+                n_ctx=st.session_state["model_n_ctx"],
+                backend="gptj",
+                temp=st.session_state["temperature"],
+                verbose=True,
+            )
         # Add more models as needed
         case _default:
             msg = f"Model {st.session_state['model']} not supported!"
@@ -269,6 +286,8 @@ def get_embeddings() -> Embeddings:
             embeddings = OpenAIEmbeddings(
                 disallowed_special=(), openai_api_key=st.session_state["openai_api_key"]
             )
+        case "huggingface-Fall-MiniLM-L6-v2":
+            embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
         # Add more embeddings as needed
         case _default:
             msg = f"Embeddings {st.session_state['embeddings']} not supported!"
@@ -281,8 +300,8 @@ def get_embeddings() -> Embeddings:
 def get_vector_store() -> VectorStore:
     # either load existing vector store or upload a new one to the hub
     embeddings = get_embeddings()
-    data_source_name = get_data_source_string()
-    dataset_path = f"hub://{st.session_state['activeloop_org_name']}/{data_source_name}-{st.session_state['chunk_size']}"
+    dataset_name = get_dataset_name()
+    dataset_path = f"hub://{st.session_state['activeloop_org_name']}/{dataset_name}-{st.session_state['chunk_size']}"
     if deeplake.exists(dataset_path, token=st.session_state["activeloop_token"]):
         with st.spinner("Loading vector store..."):
             logger.info(f"Dataset '{dataset_path}' exists -> loading")