chore: remove unnecessary deps

Author: hassiebp

poetry.lock

@@ -23,37 +23,14 @@ opentelemetry-exporter-otlp = "^1.33.1"
 
 [tool.poetry.group.dev.dependencies]
 pytest = ">=7.4,<9.0"
-chromadb = ">=0.4.2,<0.6.0"
-tiktoken = "0.7.0"
 pytest-timeout = "^2.1.0"
 pytest-xdist = "^3.3.1"
-respx = ">=0.20.2,<0.22.0"
-google-search-results = "^2.4.2"
-huggingface_hub = ">=0.16.4,<0.25.0"
 pre-commit = "^3.2.2"
-anthropic = ">=0.17.0,<1"
-bs4 = ">=0.0.1,<0.0.3"
-lark = "^1.1.7"
 pytest-asyncio = ">=0.21.1,<0.24.0"
 pytest-httpserver = "^1.0.8"
-boto3 = "^1.28.59"
 ruff = ">=0.1.8,<0.6.0"
 mypy = "^1.0.0"
-langchain-mistralai = ">=0.0.1,<0.3"
-google-cloud-aiplatform = "^1.38.1"
-cohere = ">=4.46,<6.0"
-langchain-google-vertexai = ">=1.0.0,<3.0.0"
 langchain-openai = ">=0.0.5,<0.3"
-dashscope = "^1.14.1"
-pymongo = "^4.6.1"
-llama-index-llms-anthropic = ">=0.1.1,<0.6"
-bson = "^0.5.10"
-langchain-anthropic = ">=0.1.4,<0.4"
-langchain-groq = ">=0.1.3,<0.3"
-langchain-aws = ">=0.1.3,<0.3"
-langchain-ollama = "^0.2.0"
-langchain-cohere = "^0.3.3"
-langchain-community = ">=0.2.14,<0.4"
 langgraph = "^0.2.62"
 
 [tool.poetry.group.docs.dependencies]
@@ -62,7 +39,6 @@ pdoc = "^14.4.0"
 [tool.poetry.extras]
 openai = ["openai"]
 langchain = ["langchain"]
-llama-index = ["llama-index"]
 
 [build-system]
 requires = ["poetry-core>=1.0.0"]

pyproject.toml

@@ -7,7 +7,6 @@
 from unittest.mock import patch
 
 import pytest
-from bson import ObjectId
 from langchain.schema.messages import HumanMessage
 from pydantic import BaseModel
 
@@ -129,11 +128,3 @@ def test_observation_level():
     result = json.dumps(ObservationLevel.ERROR, cls=EventSerializer)
 
     assert result == '"ERROR"'
-
-
-def test_mongo_cursor():
-    test_id = ObjectId("5f3e3e3e3e3e3e3e3e3e3e3e")
-
-    result = json.dumps(test_id, cls=EventSerializer)
-
-    assert isinstance(result, str)

tests/test_extract_model.py

@@ -7,20 +7,14 @@
 
 import pytest
 from langchain.chains import (
-    ConversationalRetrievalChain,
     ConversationChain,
     LLMChain,
-    RetrievalQA,
     SimpleSequentialChain,
 )
 from langchain.chains.openai_functions import create_openai_fn_chain
 from langchain.memory import ConversationBufferMemory
 from langchain.prompts import ChatPromptTemplate, PromptTemplate
 from langchain.schema import HumanMessage, SystemMessage
-from langchain.text_splitter import CharacterTextSplitter
-from langchain_community.document_loaders import TextLoader
-from langchain_community.embeddings import OpenAIEmbeddings
-from langchain_community.vectorstores import Chroma
 from langchain_core.callbacks.manager import CallbackManagerForLLMRun
 from langchain_core.language_models.llms import LLM
 from langchain_core.output_parsers import StrOutputParser
@@ -35,7 +29,6 @@
 from langfuse._client.client import Langfuse
 from langfuse.langchain import CallbackHandler
 from langfuse.langchain.CallbackHandler import LANGSMITH_TAG_HIDDEN
-from tests.api_wrapper import LangfuseAPI
 from tests.utils import create_uuid, encode_file_to_base64, get_api
 
 
@@ -226,89 +219,6 @@ def test_basic_chat_openai():
     assert generation.output is not None
 
 
-def test_callback_retriever():
-    langfuse = Langfuse()
-
-    with langfuse.start_as_current_span(name="retriever_test") as span:
-        trace_id = span.trace_id
-        handler = CallbackHandler()
-
-        loader = TextLoader("./static/state_of_the_union.txt", encoding="utf8")
-        llm = OpenAI()
-
-        documents = loader.load()
-        text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
-        texts = text_splitter.split_documents(documents)
-
-        embeddings = OpenAIEmbeddings()
-        docsearch = Chroma.from_documents(texts, embeddings)
-
-        query = "What did the president say about Ketanji Brown Jackson"
-
-        chain = RetrievalQA.from_chain_type(
-            llm,
-            retriever=docsearch.as_retriever(),
-        )
-
-        chain.run(query, callbacks=[handler])
-
-    langfuse.flush()
-
-    trace = get_api().trace.get(trace_id)
-
-    assert len(trace.observations) == 6
-    for observation in trace.observations:
-        if observation.type == "GENERATION":
-            assert observation.usage_details["input"] > 0
-            assert observation.usage_details["output"] > 0
-            assert observation.usage_details["total"] > 0
-            assert observation.input is not None
-            assert observation.input != ""
-            assert observation.output is not None
-            assert observation.output != ""
-
-
-def test_callback_retriever_with_sources():
-    langfuse = Langfuse()
-
-    with langfuse.start_as_current_span(name="retriever_with_sources_test") as span:
-        trace_id = span.trace_id
-        handler = CallbackHandler()
-
-        loader = TextLoader("./static/state_of_the_union.txt", encoding="utf8")
-        llm = OpenAI()
-
-        documents = loader.load()
-        text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
-        texts = text_splitter.split_documents(documents)
-
-        embeddings = OpenAIEmbeddings()
-        docsearch = Chroma.from_documents(texts, embeddings)
-
-        query = "What did the president say about Ketanji Brown Jackson"
-
-        chain = RetrievalQA.from_chain_type(
-            llm, retriever=docsearch.as_retriever(), return_source_documents=True
-        )
-
-        chain(query, callbacks=[handler])
-
-    langfuse.flush()
-
-    trace = get_api().trace.get(trace_id)
-
-    assert len(trace.observations) == 6
-    for observation in trace.observations:
-        if observation.type == "GENERATION":
-            assert observation.usage_details["input"] > 0
-            assert observation.usage_details["output"] > 0
-            assert observation.usage_details["total"] > 0
-            assert observation.input is not None
-            assert observation.input != ""
-            assert observation.output is not None
-            assert observation.output != ""
-
-
 def test_callback_retriever_conversational_with_memory():
     langfuse = Langfuse()
 
@@ -347,54 +257,6 @@ def test_callback_retriever_conversational_with_memory():
         assert generation.usage_details["output"] is not None
 
 
-def test_callback_retriever_conversational():
-    langfuse = Langfuse()
-
-    with langfuse.start_as_current_span(name="retriever_conversational_test") as span:
-        trace_id = span.trace_id
-        api_wrapper = LangfuseAPI()
-        handler = CallbackHandler()
-
-        loader = TextLoader("./static/state_of_the_union.txt", encoding="utf8")
-
-        documents = loader.load()
-        text_splitter = CharacterTextSplitter(chunk_size=1000, chunk_overlap=0)
-        texts = text_splitter.split_documents(documents)
-
-        embeddings = OpenAIEmbeddings(openai_api_key=os.environ.get("OPENAI_API_KEY"))
-        docsearch = Chroma.from_documents(texts, embeddings)
-
-        query = "What did the president say about Ketanji Brown Jackson"
-
-        chain = ConversationalRetrievalChain.from_llm(
-            ChatOpenAI(
-                openai_api_key=os.environ.get("OPENAI_API_KEY"),
-                temperature=0.5,
-                model="gpt-3.5-turbo-16k",
-            ),
-            docsearch.as_retriever(search_kwargs={"k": 6}),
-            return_source_documents=True,
-        )
-
-        chain({"question": query, "chat_history": []}, callbacks=[handler])
-
-    handler.client.flush()
-
-    trace = api_wrapper.get_trace(trace_id)
-
-    # Add 1 to account for the wrapping span
-    assert len(trace["observations"]) == 6
-    for observation in trace["observations"]:
-        if observation["type"] == "GENERATION":
-            assert observation["promptTokens"] > 0
-            assert observation["completionTokens"] > 0
-            assert observation["totalTokens"] > 0
-            assert observation["input"] is not None
-            assert observation["input"] != ""
-            assert observation["output"] is not None
-            assert observation["output"] != ""
-
-
 def test_callback_simple_openai():
     langfuse = Langfuse()