added open-ai integration

.gitignore

@@ -25,6 +25,11 @@ share/python-wheels/
 .installed.cfg
 *.egg
 MANIFEST
+langcache-venv/*
+langcache/.DS_Store
+test/.DS_Store
+.DS_Store
+.vscode/*
 
 # PyInstaller
 #  Usually these files are written by a python script from a template
@@ -50,6 +55,7 @@ coverage.xml
 .hypothesis/
 .pytest_cache/
 cover/
+test/evadb_data/*
 
 # Translations
 *.mo

README.md

@@ -1 +1,50 @@
-# langcache
+# langcache
+LangCache is a semantic caching library developed for Large Language Model (LLM) queries. Its primary purpose is to address the cost concerns associated with LLM API calls and to improve the speed of LLM applications. 
+
+### Installation of langcache for developer setup
+To install langcache, we recommend using the pip package manager.
+
+1. Create a new virtual environment called langcache-venv.
+```python -m venv langcache-venv```
+
+2. Now, activate the virtual environment:
+```source langcache-venv/bin/activate```
+
+3. Install the dependecies
+```pip install .```
+
+### OpenAI ChatGPT 3.5 ChatCompletion API usage with langcache enabled
+
+Before running the example, make sure the `OPENAI_API_KEY` environment variable is set by executing `echo $OPENAI_API_KEY`.
+If it is not already set, it can be set by using `export OPENAI_API_KEY=YOUR_API_KEY` on Unix/Linux/MacOS systems or `set OPENAI_API_KEY=YOUR_API_KEY` on Windows systems.
+
+#### OpenAI API original usage
+```
+import openai
+
+question = "What is ChatGPT?"
+completion = client.chat.completions.create(
+   model="gpt-3.5-turbo",
+   messages=[
+     {"role": "system", "content":"You are an helpful assistant to answer all my questions within 15 words limit"},
+     {"role": "user", "content": question}
+   ]
+ )
+```
+#### OpenAI API + LangCache, similar search cache
+```
+from langcache.adapter.openai import OpenAI
+from langcache.core import Cache
+
+cache = Cache(tune_frequency=5, tune_policy="recall")
+client = OpenAI(cache)
+
+question = "What is ChatGPT?"
+completion = client.chat.completions.create(
+   model="gpt-3.5-turbo",
+   messages=[
+     {"role": "system", "content":"You are an helpful assistant to answer all my questions within 15 words limit"},
+     {"role": "user", "content": question}
+   ]
+ )
+```

langcache/adapter/openai.py

@@ -0,0 +1,24 @@
+import openai
+from ..core import Cache
+
+class LangCacheCompletions(openai.resources.chat.completions.Completions):
+    # TODO: Add remaining args
+    def create(self, messages, model):
+        # Get the last message with role "user"
+        last_user_message = [message for message in messages if message["role"] == "user"][-1]["content"]
+        cached_value = self._client.cache.get(last_user_message)
+
+        if cached_value is None:
+            llm_output = openai.resources.chat.completions.Completions.create(self, messages=messages, model=model).choices[0].message.content
+            self._client.cache.put(last_user_message, llm_output)
+            return llm_output
+        else:
+            return cached_value
+
+
+class OpenAI(openai.OpenAI):
+    # TODO: Add remaining args
+    def __init__(self, cache):
+        super(OpenAI, self).__init__()
+        self.cache = cache
+        self.chat.completions = LangCacheCompletions(self)

langcache/core.py

@@ -1,6 +1,6 @@
 import os
 import evadb
-import openai
+from openai import OpenAI
 import string
 import random
 import multiprocessing as mp
@@ -17,6 +17,7 @@
 class Cache:
     def __init__(self, name=None, tune_frequency=5, tune_policy="precision"):
         self.cursor = evadb.connect().cursor()
+        self.client = OpenAI()
 
         # Setup needed functions.
         self.cursor.query(
@@ -72,7 +73,7 @@ def _evaluate_and_tune(self, key: str, ret_key: str, distance: float, **kwargs):
             # LLM tuning.
             self.llm_msg[1]["content"] = f""" "{ret_key}" , "{key}" """
             response = (
-                openai.ChatCompletion.create(
+                self.client.chat.completions.create(
                     model="gpt-3.5-turbo",
                     messages=self.llm_msg,
                 )
@@ -165,4 +166,4 @@ def put(self, key: str, value: str):
                 f"""
                 INSERT INTO {self.cache_name} (key, value) VALUES ("{key}", "{value}")
             """
-            ).df()
+            ).df()

requirements.txt

@@ -0,0 +1,131 @@
+aenum==3.1.15
+aiohttp==3.9.0
+aiosignal==1.3.1
+anyio==3.7.1
+appnope==0.1.3
+asttokens==2.4.1
+attrs==23.1.0
+backcall==0.2.0
+beautifulsoup4==4.12.2
+bs4==0.0.1
+certifi==2023.11.17
+cffi==1.16.0
+charset-normalizer==3.3.2
+click==8.1.7
+comm==0.2.0
+contourpy==1.2.0
+cryptography==41.0.5
+cycler==0.12.1
+dataclasses-json==0.6.2
+debugpy==1.8.0
+decorator==5.1.1
+diskcache==5.6.3
+distro==1.8.0
+evadb==0.3.9
+executing==2.0.1
+faiss-cpu==1.7.4
+fastjsonschema==2.19.0
+filelock==3.13.1
+fonttools==4.45.1
+frozenlist==1.4.0
+fsspec==2023.10.0
+gpt4all==2.0.2
+h11==0.14.0
+httpcore==1.0.2
+httpx==0.25.1
+huggingface-hub==0.19.4
+idna==3.4
+iniconfig==2.0.0
+ipykernel==6.26.0
+ipython==8.12.3
+ipywidgets==8.1.1
+jedi==0.19.1
+Jinja2==3.1.2
+joblib==1.3.2
+jsonpatch==1.33
+jsonpointer==2.4
+jsonschema==4.20.0
+jsonschema-specifications==2023.11.1
+jupyter_client==8.6.0
+jupyter_core==5.5.0
+jupyterlab-widgets==3.0.9
+kiwisolver==1.4.5
+langcache==0.0.1
+langchain==0.0.340
+langsmith==0.0.66
+lark==1.1.8
+MarkupSafe==2.1.3
+marshmallow==3.20.1
+matplotlib==3.8.2
+matplotlib-inline==0.1.6
+mpmath==1.3.0
+multidict==6.0.4
+mypy-extensions==1.0.0
+nbclient==0.6.8
+nbformat==5.9.2
+nbmake==1.4.6
+nest-asyncio==1.5.8
+networkx==3.2.1
+nltk==3.8.1
+numpy==1.26.2
+openai==1.3.5
+packaging==23.2
+pandas==2.1.3
+parso==0.8.3
+pdfminer.six==20221105
+pexpect==4.9.0
+pickleshare==0.7.5
+Pillow==10.1.0
+platformdirs==4.0.0
+pluggy==1.3.0
+prompt-toolkit==3.0.41
+protobuf==4.25.1
+psutil==5.9.6
+ptyprocess==0.7.0
+pure-eval==0.2.2
+py==1.11.0
+pycparser==2.21
+pydantic==1.10.13
+Pygments==2.17.2
+PyMuPDF==1.22.5
+pyparsing==3.1.1
+pytest==7.4.3
+python-dateutil==2.8.2
+pytz==2023.3.post1
+PyYAML==6.0.1
+pyzmq==25.1.1
+rapidfuzz==3.5.2
+referencing==0.31.0
+regex==2023.10.3
+requests==2.31.0
+retry==0.9.2
+rpds-py==0.13.1
+safetensors==0.4.0
+scikit-learn==1.3.2
+scipy==1.11.4
+sentence-transformers==2.2.2
+sentencepiece==0.1.99
+six==1.16.0
+sniffio==1.3.0
+soupsieve==2.5
+SQLAlchemy==2.0.23
+SQLAlchemy-Utils==0.41.1
+stack-data==0.6.3
+sympy==1.12
+tenacity==8.2.3
+thefuzz==0.20.0
+threadpoolctl==3.2.0
+tokenizers==0.15.0
+torch==2.1.1
+torchvision==0.16.1
+tornado==6.3.3
+tqdm==4.66.1
+traitlets==5.13.0
+transformers==4.35.2
+typing-inspect==0.9.0
+typing_extensions==4.8.0
+tzdata==2023.3
+urllib3==2.1.0
+wcwidth==0.2.12
+widgetsnbextension==4.0.9
+yarl==1.9.3

setup.py

@@ -0,0 +1,27 @@
+import setuptools
+
+with open("README.md", "r", encoding="utf-8") as fh:
+    long_description = fh.read()
+
+# Function to read the list of dependencies from requirements.txt
+def load_requirements(filename='requirements.txt'):
+    with open(filename, 'r') as file:
+        return file.read().splitlines()
+
+setuptools.setup(
+    name="langcache",
+    version="0.0.1",
+    author="Georgia Tech Database Group",
+    author_email="jiashenc@gatech.edu",
+    description="A small example package",
+    long_description=long_description,
+    long_description_content_type="text/markdown",
+    url="https://github.com/jiashenC/langcache",
+    packages=setuptools.find_packages(),
+    classifiers=[
+        "Programming Language :: Python :: 3",
+        "Development Status :: 1 - Planning"
+    ],
+    python_requires='>=3.6',
+    install_requires=load_requirements(),
+)