feat: add Google embedding support & update setup (#550) bump:patch

flowsettings.py

@@ -26,6 +26,7 @@ if not KH_APP_VERSION:
 
 KH_ENABLE_FIRST_SETUP = True
 KH_DEMO_MODE = config("KH_DEMO_MODE", default=False, cast=bool)
+KH_OLLAMA_URL = config("KH_OLLAMA_URL", default="http://localhost:11434/v1/")
 
 # App can be ran from anywhere and it's not trivial to decide where to store app data.
 # So let's use the same directory as the flowsetting.py file.
@@ -162,7 +163,7 @@ if config("LOCAL_MODEL", default=""):
     KH_LLMS["ollama"] = {
         "spec": {
             "__type__": "kotaemon.llms.ChatOpenAI",
-            "base_url": "http://localhost:11434/v1/",
+            "base_url": KH_OLLAMA_URL,
             "model": config("LOCAL_MODEL", default="llama3.1:8b"),
             "api_key": "ollama",
         },
@@ -171,7 +172,7 @@ if config("LOCAL_MODEL", default=""):
     KH_EMBEDDINGS["ollama"] = {
         "spec": {
             "__type__": "kotaemon.embeddings.OpenAIEmbeddings",
-            "base_url": "http://localhost:11434/v1/",
+            "base_url": KH_OLLAMA_URL,
             "model": config("LOCAL_MODEL_EMBEDDINGS", default="nomic-embed-text"),
             "api_key": "ollama",
         },
@@ -195,11 +196,11 @@ KH_LLMS["claude"] = {
     },
     "default": False,
 }
-KH_LLMS["gemini"] = {
+KH_LLMS["google"] = {
     "spec": {
         "__type__": "kotaemon.llms.chats.LCGeminiChat",
-        "model_name": "gemini-1.5-pro",
+        "model_name": "gemini-1.5-flash",
-        "api_key": "your-key",
+        "api_key": config("GOOGLE_API_KEY", default="your-key"),
     },
     "default": False,
 }
@@ -231,6 +232,13 @@ KH_EMBEDDINGS["cohere"] = {
     },
     "default": False,
 }
+KH_EMBEDDINGS["google"] = {
+    "spec": {
+        "__type__": "kotaemon.embeddings.LCGoogleEmbeddings",
+        "model": "models/text-embedding-004",
+        "google_api_key": config("GOOGLE_API_KEY", default="your-key"),
+    }
+}
 # KH_EMBEDDINGS["huggingface"] = {
 #     "spec": {
 #         "__type__": "kotaemon.embeddings.LCHuggingFaceEmbeddings",

libs/kotaemon/kotaemon/embeddings/__init__.py

@@ -4,6 +4,7 @@ from .fastembed import FastEmbedEmbeddings
 from .langchain_based import (
     LCAzureOpenAIEmbeddings,
     LCCohereEmbeddings,
+    LCGoogleEmbeddings,
     LCHuggingFaceEmbeddings,
     LCOpenAIEmbeddings,
 )
@@ -18,6 +19,7 @@ __all__ = [
     "LCAzureOpenAIEmbeddings",
     "LCCohereEmbeddings",
     "LCHuggingFaceEmbeddings",
+    "LCGoogleEmbeddings",
     "OpenAIEmbeddings",
     "AzureOpenAIEmbeddings",
     "FastEmbedEmbeddings",

libs/kotaemon/kotaemon/embeddings/langchain_based.py

@@ -219,3 +219,38 @@ class LCHuggingFaceEmbeddings(LCEmbeddingMixin, BaseEmbeddings):
             from langchain.embeddings import HuggingFaceBgeEmbeddings
 
         return HuggingFaceBgeEmbeddings
+
+
+class LCGoogleEmbeddings(LCEmbeddingMixin, BaseEmbeddings):
+    """Wrapper around Langchain's Google GenAI embedding, focusing on key parameters"""
+
+    google_api_key: str = Param(
+        help="API key (https://aistudio.google.com/app/apikey)",
+        default=None,
+        required=True,
+    )
+    model: str = Param(
+        help="Model name to use (https://ai.google.dev/gemini-api/docs/models/gemini#text-embedding-and-embedding)",  # noqa
+        default="models/text-embedding-004",
+        required=True,
+    )
+
+    def __init__(
+        self,
+        model: str = "models/text-embedding-004",
+        google_api_key: Optional[str] = None,
+        **params,
+    ):
+        super().__init__(
+            model=model,
+            google_api_key=google_api_key,
+            **params,
+        )
+
+    def _get_lc_class(self):
+        try:
+            from langchain_google_genai import GoogleGenerativeAIEmbeddings
+        except ImportError:
+            raise ImportError("Please install langchain-google-genai")
+
+        return GoogleGenerativeAIEmbeddings

libs/ktem/ktem/embeddings/manager.py

@@ -57,6 +57,7 @@ class EmbeddingManager:
             AzureOpenAIEmbeddings,
             FastEmbedEmbeddings,
             LCCohereEmbeddings,
+            LCGoogleEmbeddings,
             LCHuggingFaceEmbeddings,
             OpenAIEmbeddings,
             TeiEndpointEmbeddings,
@@ -68,6 +69,7 @@ class EmbeddingManager:
             FastEmbedEmbeddings,
             LCCohereEmbeddings,
             LCHuggingFaceEmbeddings,
+            LCGoogleEmbeddings,
             TeiEndpointEmbeddings,
         ]
 

libs/ktem/ktem/pages/setup.py

@@ -9,7 +9,10 @@ from ktem.rerankings.manager import reranking_models_manager as rerankers
 from theflow.settings import settings as flowsettings
 
 KH_DEMO_MODE = getattr(flowsettings, "KH_DEMO_MODE", False)
-DEFAULT_OLLAMA_URL = "http://localhost:11434/api"
+KH_OLLAMA_URL = getattr(flowsettings, "KH_OLLAMA_URL", "http://localhost:11434/v1/")
+DEFAULT_OLLAMA_URL = KH_OLLAMA_URL.replace("v1", "api")
+if DEFAULT_OLLAMA_URL.endswith("/"):
+    DEFAULT_OLLAMA_URL = DEFAULT_OLLAMA_URL[:-1]
 
 
 DEMO_MESSAGE = (
@@ -55,8 +58,9 @@ class SetupPage(BasePage):
         gr.Markdown(f"# Welcome to {self._app.app_name} first setup!")
         self.radio_model = gr.Radio(
             [
-                ("Cohere API (*free registration* available) - recommended", "cohere"),
+                ("Cohere API (*free registration*) - recommended", "cohere"),
-                ("OpenAI API (for more advance models)", "openai"),
+                ("Google API (*free registration*)", "google"),
+                ("OpenAI API (for GPT-based models)", "openai"),
                 ("Local LLM (for completely *private RAG*)", "ollama"),
             ],
             label="Select your model provider",
@@ -92,6 +96,18 @@ class SetupPage(BasePage):
                 show_label=False, placeholder="Cohere API Key"
             )
 
+        with gr.Column(visible=False) as self.google_option:
+            gr.Markdown(
+                (
+                    "#### Google API Key\n\n"
+                    "(register your free API key "
+                    "at https://aistudio.google.com/app/apikey)"
+                )
+            )
+            self.google_api_key = gr.Textbox(
+                show_label=False, placeholder="Google API Key"
+            )
+
         with gr.Column(visible=False) as self.ollama_option:
             gr.Markdown(
                 (
@@ -119,7 +135,12 @@ class SetupPage(BasePage):
                 self.openai_api_key.submit,
             ],
             fn=self.update_model,
-            inputs=[self.cohere_api_key, self.openai_api_key, self.radio_model],
+            inputs=[
+                self.cohere_api_key,
+                self.openai_api_key,
+                self.google_api_key,
+                self.radio_model,
+            ],
             outputs=[self.setup_log],
             show_progress="hidden",
         )
@@ -147,13 +168,19 @@ class SetupPage(BasePage):
             fn=self.switch_options_view,
             inputs=[self.radio_model],
             show_progress="hidden",
-            outputs=[self.cohere_option, self.openai_option, self.ollama_option],
+            outputs=[
+                self.cohere_option,
+                self.openai_option,
+                self.ollama_option,
+                self.google_option,
+            ],
         )
 
     def update_model(
         self,
         cohere_api_key,
         openai_api_key,
+        google_api_key,
         radio_model_value,
     ):
         # skip if KH_DEMO_MODE
@@ -221,12 +248,32 @@ class SetupPage(BasePage):
                     },
                     default=True,
                 )
+        elif radio_model_value == "google":
+            if google_api_key:
+                llms.update(
+                    name="google",
+                    spec={
+                        "__type__": "kotaemon.llms.chats.LCGeminiChat",
+                        "model_name": "gemini-1.5-flash",
+                        "api_key": google_api_key,
+                    },
+                    default=True,
+                )
+                embeddings.update(
+                    name="google",
+                    spec={
+                        "__type__": "kotaemon.embeddings.LCGoogleEmbeddings",
+                        "model": "models/text-embedding-004",
+                        "google_api_key": google_api_key,
+                    },
+                    default=True,
+                )
         elif radio_model_value == "ollama":
             llms.update(
                 name="ollama",
                 spec={
                     "__type__": "kotaemon.llms.ChatOpenAI",
-                    "base_url": "http://localhost:11434/v1/",
+                    "base_url": KH_OLLAMA_URL,
                     "model": "llama3.1:8b",
                     "api_key": "ollama",
                 },
@@ -236,7 +283,7 @@ class SetupPage(BasePage):
                 name="ollama",
                 spec={
                     "__type__": "kotaemon.embeddings.OpenAIEmbeddings",
-                    "base_url": "http://localhost:11434/v1/",
+                    "base_url": KH_OLLAMA_URL,
                     "model": "nomic-embed-text",
                     "api_key": "ollama",
                 },
@@ -345,9 +392,9 @@ class SetupPage(BasePage):
         return default_settings
 
     def switch_options_view(self, radio_model_value):
-        components_visible = [gr.update(visible=False) for _ in range(3)]
+        components_visible = [gr.update(visible=False) for _ in range(4)]
 
-        values = ["cohere", "openai", "ollama", None]
+        values = ["cohere", "openai", "ollama", "google", None]
         assert radio_model_value in values, f"Invalid value {radio_model_value}"
 
         if radio_model_value is not None: