feat: Add support for gpt-4o model (#2589)

This pull request adds support for the gpt-4o model to the existing
codebase. It includes changes to the BrainConfig, openAiFreeModels,
defineMaxTokens, model_compatible_with_function_calling, create_graph,
main, and process_assistant functions.

backend/modules/assistant/ito/difference.py

@@ -101,7 +101,7 @@ async def process_assistant(self):
         document_1_to_langchain = document_1_llama_parsed[0].to_langchain_format()
         document_2_to_langchain = document_2_llama_parsed[0].to_langchain_format()
 
-        llm = ChatLiteLLM(model="gpt-4-turbo-2024-04-09")
+        llm = ChatLiteLLM(model="gpt-4o")
 
         human_prompt = """Given the following two documents, find the difference between them:
 

backend/modules/brain/integrations/GPT4/Brain.py

@@ -166,9 +166,7 @@ def create_graph(self):
         return app
 
     def get_chain(self):
-        self.function_model = ChatOpenAI(
-            model="gpt-4-turbo", temperature=0, streaming=True
-        )
+        self.function_model = ChatOpenAI(model="gpt-4o", temperature=0, streaming=True)
 
         self.function_model = self.function_model.bind_tools(self.tools)
 

backend/modules/brain/qa_interface.py

@@ -42,6 +42,7 @@ def generate_stream(
 
     def model_compatible_with_function_calling(self, model: str):
         if model in [
+            "gpt-4o",
             "gpt-4-turbo",
             "gpt-4-turbo-2024-04-09",
             "gpt-4-turbo-preview",

backend/modules/brain/rags/quivr_rag.py

@@ -4,14 +4,14 @@
 from uuid import UUID
 
 from langchain.chains import ConversationalRetrievalChain
-from langchain_community.embeddings import OllamaEmbeddings
 from langchain.llms.base import BaseLLM
 from langchain.prompts import HumanMessagePromptTemplate, SystemMessagePromptTemplate
 from langchain.retrievers import ContextualCompressionRetriever
 from langchain.retrievers.document_compressors import FlashrankRerank
 from langchain.schema import format_document
 from langchain_cohere import CohereRerank
 from langchain_community.chat_models import ChatLiteLLM
+from langchain_community.embeddings import OllamaEmbeddings
 from langchain_core.output_parsers import StrOutputParser
 from langchain_core.prompts import ChatPromptTemplate, PromptTemplate
 from langchain_core.pydantic_v1 import BaseModel as BaseModelV1
@@ -136,6 +136,7 @@ def prompt_to_use(self):
 
     def model_compatible_with_function_calling(self):
         if self.model in [
+            "gpt-4o",
             "gpt-4-turbo",
             "gpt-4-turbo-2024-04-09",
             "gpt-4-turbo-preview",

backend/modules/chat/controller/chat/brainful_chat.py

@@ -34,6 +34,7 @@
     "gpt-4-0125-preview",
     "gpt-3.5-turbo",
     "gpt-4-turbo",
+    "gpt-4o",
 ]
 
 

backend/requirements.txt

@@ -17,8 +17,8 @@ backoff==2.2.1; python_version >= '3.7' and python_version < '4.0'
 beautifulsoup4==4.12.3; python_full_version >= '3.6.0'
 billiard==4.2.0; python_version >= '3.7'
 black==24.4.2; python_version >= '3.8'
-boto3==1.34.101; python_version >= '3.8'
-botocore==1.34.101; python_version >= '3.8'
+boto3==1.34.104; python_version >= '3.8'
+botocore==1.34.104; python_version >= '3.8'
 cachetools==5.3.3; python_version >= '3.7'
 celery[redis,sqs]==5.4.0; python_version >= '3.8'
 certifi==2024.2.2; python_version >= '3.6'
@@ -29,15 +29,15 @@ click==8.1.7; python_version >= '3.7'
 click-didyoumean==0.3.1; python_full_version >= '3.6.2'
 click-plugins==1.1.1
 click-repl==0.3.0; python_version >= '3.6'
-cohere==5.3.5; python_version >= '3.8' and python_version < '4.0'
+cohere==5.4.0; python_version >= '3.8' and python_version < '4.0'
 coloredlogs==15.0.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
 colorlog==6.8.2; python_version >= '3.6'
 contourpy==1.2.1; python_version >= '3.9'
 cryptography==42.0.7; python_version >= '3.7'
 cssselect==1.2.0; python_version >= '3.7'
 curl-cffi==0.7.0b4; python_version >= '3.8'
 cycler==0.12.1; python_version >= '3.8'
-dataclasses-json==0.6.5; python_version >= '3.7' and python_version < '4.0'
+dataclasses-json==0.6.6; python_version >= '3.7' and python_version < '4.0'
 datasets==2.19.1; python_full_version >= '3.8.0'
 debugpy==1.8.1; python_version >= '3.8'
 decorator==5.1.1; python_version >= '3.5'
@@ -109,35 +109,35 @@ jsonpath-python==1.0.6; python_version >= '3.6'
 jsonpointer==2.4; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6'
 kiwisolver==1.4.5; python_version >= '3.7'
 kombu[sqs]==5.3.7; python_version >= '3.8'
-langchain==0.1.19; python_version < '4.0' and python_full_version >= '3.8.1'
+langchain==0.1.20; python_version < '4.0' and python_full_version >= '3.8.1'
 langchain-cohere==0.1.4; python_version < '4.0' and python_full_version >= '3.8.1'
 langchain-community==0.0.38; python_version < '4.0' and python_full_version >= '3.8.1'
 langchain-core==0.1.52; python_version < '4.0' and python_full_version >= '3.8.1'
 langchain-openai==0.1.6; python_version < '4.0' and python_full_version >= '3.8.1'
 langchain-text-splitters==0.0.1; python_version < '4.0' and python_full_version >= '3.8.1'
 langdetect==1.0.9
-langfuse==2.29.3; python_version < '4.0' and python_full_version >= '3.8.1'
+langfuse==2.30.0; python_version < '4.0' and python_full_version >= '3.8.1'
 langgraph==0.0.48; python_version < '4.0' and python_full_version >= '3.9.0'
-langsmith==0.1.56; python_version < '4.0' and python_full_version >= '3.8.1'
+langsmith==0.1.57; python_version < '4.0' and python_full_version >= '3.8.1'
 layoutparser[layoutmodels,tesseract]==0.3.4; python_version >= '3.6'
-litellm==1.36.4; python_version not in '2.7, 3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7' and python_version >= '3.8'
+litellm==1.37.5; python_version not in '2.7, 3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7' and python_version >= '3.8'
 llama-cpp-python==0.2.67; python_version >= '3.8'
-llama-index==0.10.35; python_version < '4.0' and python_full_version >= '3.8.1'
+llama-index==0.10.36; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-agent-openai==0.2.4; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-cli==0.1.12; python_version < '4.0' and python_full_version >= '3.8.1'
-llama-index-core==0.10.35.post1; python_version < '4.0' and python_full_version >= '3.8.1'
+llama-index-core==0.10.36; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-embeddings-openai==0.1.9; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-indices-managed-llama-cloud==0.1.6; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-legacy==0.9.48; python_version < '4.0' and python_full_version >= '3.8.1'
-llama-index-llms-openai==0.1.18; python_version < '4.0' and python_full_version >= '3.8.1'
-llama-index-multi-modal-llms-openai==0.1.5; python_version < '4.0' and python_full_version >= '3.8.1'
+llama-index-llms-openai==0.1.19; python_version < '4.0' and python_full_version >= '3.8.1'
+llama-index-multi-modal-llms-openai==0.1.6; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-program-openai==0.1.6; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-question-gen-openai==0.1.3; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-readers-file==0.1.22; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-index-readers-llama-parse==0.1.4; python_version < '4.0' and python_full_version >= '3.8.1'
 llama-parse==0.4.2; python_version < '4.0' and python_full_version >= '3.8.1'
 llamaindex-py-client==0.1.19; python_version >= '3.8' and python_version < '4'
-lxml[html_clean]==5.2.1; python_version >= '3.6'
+lxml[html_clean]==5.2.2; python_version >= '3.6'
 lxml-html-clean==0.1.1
 markdown==3.6
 markdown-it-py==3.0.0; python_version >= '3.8'
@@ -162,7 +162,7 @@ olefile==0.47; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2,
 omegaconf==2.3.0; python_version >= '3.6'
 onnx==1.16.0
 onnxruntime==1.17.3
-openai==1.27.0; python_full_version >= '3.7.1'
+openai==1.29.0; python_full_version >= '3.7.1'
 opencv-python==4.9.0.80; python_version >= '3.6'
 openpyxl==3.1.2
 ordered-set==4.1.0; python_version >= '3.7'
@@ -209,7 +209,7 @@ pyinstrument==4.6.2; python_version >= '3.7'
 pypandoc==1.13; python_version >= '3.6'
 pyparsing==3.1.2; python_full_version >= '3.6.8'
 pypdf==4.2.0; python_version >= '3.6'
-pypdfium2==4.29.0; python_version >= '3.6'
+pypdfium2==4.30.0; python_version >= '3.6'
 pyright==1.1.362; python_version >= '3.7'
 pysbd==0.3.4; python_version >= '3'
 pytesseract==0.3.10; python_version >= '3.7'
@@ -232,7 +232,7 @@ ragas==0.1.7
 rapidfuzz==3.9.0; python_version >= '3.8'
 realtime==1.0.4; python_version >= '3.8' and python_version < '4.0'
 redis==5.0.4; python_version >= '3.7'
-regex==2024.4.28; python_version >= '3.8'
+regex==2024.5.10; python_version >= '3.8'
 requests==2.31.0; python_version >= '3.7'
 requests-file==2.0.0
 resend==1.0.1; python_version >= '3.7'
@@ -261,7 +261,7 @@ sympy==1.12; python_version >= '3.8'
 tabulate==0.9.0; python_version >= '3.7'
 tavily-python==0.3.3; python_version >= '3.6'
 tenacity==8.3.0; python_version >= '3.8'
-tiktoken==0.6.0; python_version >= '3.8'
+tiktoken==0.7.0; python_version >= '3.8'
 timm==0.9.16; python_version >= '3.8'
 tinysegmenter==0.3
 tldextract==5.1.2; python_version >= '3.8'

backend/tests/ragas_evaluation/run_evaluation.py

@@ -70,7 +70,7 @@ def main(
     score = evaluate(
         response_dataset,
         metrics=ragas_metrics,
-        llm=ChatOpenAI(model="gpt-3.5-turbo-0125", temperature=0.1),
+        llm=ChatOpenAI(model="gpt-4o", temperature=0.1),
         embeddings=LangchainEmbeddingsWrapper(
             OpenAIEmbeddings(model="text-embedding-3-large", dimensions=1536)
         ),

frontend/lib/helpers/defineMaxTokens.ts

@@ -21,7 +21,9 @@ export const defineMaxTokens = (
       return 2000;
     case "mistral/mistral-large-latest":
       return 2000;
+    case "gpt-4o":
+      return 2000;
     default:
-      return 1000;
+      return 2000;
   }
 };

frontend/lib/types/BrainConfig.ts

@@ -39,6 +39,7 @@ export type BrainConfig = {
 
 export const openAiFreeModels = [
   "gpt-3.5-turbo",
+  "gpt-4o",
   "gpt-3.5-turbo-1106",
   "gpt-3.5-turbo-16k",
   "gpt-4-0125-preview",