How to Speed Up Large Language Model Pipelines Using GPTCache: A Guide and Performance Comparison

November 30, 2023


The incorporation of large language models (LLMs) such as ChatGPT into various applications has created a plethora of new opportunities. These models are utilised in a variety of applications, such as chatbots and in content creation, and are capable of producing human-like language. However, when your application becomes more well-known and experiences more traffic, the price of LLM API calls may increase dramatically and response times may become slower, particularly when a large volume of requests are made. 

In this blog post, we'll look at GPTCache, a novel project that attempts to construct a semantic cache for storing LLM replies. This creative method not only dramatically lowers the costs associated with LLM API calls, but it also speeds up your application.

Quick Start with GPTCache

Before we delve into the details of GPTCache, let's get started:

Installation: You can install GPTCache with a simple pip command:

pip install gptcache

Usage: To use GPTCache, you only need to include a few lines of code. There's no need to modify your existing code. Here's an example

from gptcache import cache
from gptcache.adapter import openai


For more detailed instructions and examples, please refer to the official documentation.

Understanding GPTCache

GPTCache is a system designed to optimize the retrieval process of relevant information by incorporating caching mechanisms. It achieves this by storing precomputed embeddings and their corresponding similar vectors. Let's take a closer look at the components that make up this system:

  1. LLM Adapter: Establishing the Connection between the LLM Model and the Backend
    The LLM Adapter serves as a crucial intermediary between the language model and the backend systems. It establishes a connection that allows the LLM to access and retrieve data from the backend as needed. This integration streamlines communication and ensures that the LLM can seamlessly interact with other components of the pipeline.
  2. Embedding Generator: Generating Query Embeddings
    The Embedding Generator is responsible for converting user queries into embeddings. These embeddings represent the semantic information contained within the query in a numerical form. By doing so, the system can efficiently compare and evaluate the similarity of these embeddings with stored vectors, enabling faster and more accurate results.
  3. Similarity Evaluator: Assessing Vector Similarity
    The Similarity Evaluator is a crucial component of the GPTCache system. It evaluates the resemblance between the embedding of the query and the vectors stored in the cache. By using various similarity metrics, such as cosine similarity, it determines the degree of resemblance between vectors. This aids in identifying the most relevant matches, ensuring that the system responds with the most appropriate information.
  4. Cache Storage: Storage of Vectors and Comparable Vectors
    The Cache Storage component serves as the repository for vectors and their corresponding similar vectors. It stores key-value pairs and organizes them in descending order based on their distance or similarity. This arrangement enables the system to quickly retrieve the most relevant and similar vectors during query processing, thus significantly reducing response times.
  5. Cache Hit: Checking for Vector Existence in the Cache
    During the query processing phase, the Cache Hit functionality is employed to determine whether a given vector already exists in the cache storage. By checking for vector existence, the system can efficiently identify and retrieve previously stored vectors, potentially avoiding redundant computations and further accelerating the response.
  6. LLM: Responding with Relevant Paragraphs
    The LLM at the core of the GPTCache system receives the relevant paragraph, typically extracted from a larger document corpus, and generates a response based on the query and the provided context. Leveraging its language understanding capabilities, the LLM provides accurate and contextually appropriate responses, enhancing the overall user experience.

How GPTCache Works

Traditional cache systems rely on an exact match between a new query and a cached query to determine if the requested content is available in the cache. However, for LLM caches, this approach is less effective due to the complexity and variability of LLM queries, resulting in a low cache hit rate. To address this, GPTCache adopts alternative strategies like semantic caching.

Semantic caching identifies and stores similar or related queries, increasing cache hit probability and enhancing overall caching efficiency. GPTCache uses embedding algorithms to convert queries into embeddings and employs a vector store for similarity searches on these embeddings. This procedure enables GPTCache to recognize and fetch similar or related queries from the cache storage. 

GPTCache is designed with modularity, allowing users to tailor their own semantic cache. The system provides diverse implementations for each module, and users can even create their own implementations to meet their specific requirements.

In a semantic cache, false positives may occur during cache hits, and false negatives may occur during cache misses. GPTCache presents three key metrics to assess its performance:

  1. Hit Ratio: This metric measures the cache's success in fulfilling content requests compared to the total number of requests it receives. The higher the hit ratio, the more effective the cache is.
  2. Latency: This metric gauges the time it takes for a query to be processed and the corresponding data to be retrieved from the cache. A caching system that has reduced latency is more responsive and efficient.
  3. Recall: This metric shows the percentage of queries that the cache has answered out of all the queries that it was supposed to answer. Higher recall percentages indicate that the cache is effectively delivering the appropriate content.

Performance Comparison

The introduction of GPTCache into LLM pipelines has brought about significant improvements in terms of speed and efficiency. By storing precomputed embeddings and their corresponding similar vectors, the system can retrieve relevant information quickly, resulting in faster response times. This has a profound impact on various applications, including chatbots, search engines, and information retrieval systems.

In comparison to traditional LLM pipelines without GPTCache, the performance gain is substantial. The GPTCache system significantly reduces the computational load on the LLM, as it can rely on precomputed embeddings and cache storage for frequently occurring queries. This reduction in computation time not only speeds up response times but also decreases the hardware and energy requirements, making the system more sustainable.

Benefits of GPTCache

GPTCache offers a range of compelling benefits:

Decreased Expenses

Most LLM services charge fees based on the number of requests and token count. GPTCache effectively reduces your expenses by caching query results, minimizing the number of requests and tokens sent to the LLM service. As a result, you can enjoy a more cost-efficient experience when using the service.

Enhanced Performance

LLMs use generative AI algorithms to generate responses in real-time, which can be time-consuming. However, when a similar query is cached, the response time significantly improves. GPTCache fetches results directly from the cache, eliminating the need to interact with the LLM service. In most situations, GPTCache can provide superior query throughput compared to standard LLM services.

Adaptable Development and Testing

GPTCache provides an interface that mirrors LLM APIs and accommodates storage of both LLM-generated and mocking data. This feature enables developers to effortlessly develop and test their applications, eliminating the need to connect to the LLM service. Comprehensive testing of your application is crucial before moving it to a production environment.

Improved Scalability and Availability

LLM services often impose rate limits, which can lead to service outages if exceeded. With GPTCache, you can easily scale to accommodate an increasing volume of queries, ensuring consistent performance as your application's user base expands.


If you require GPU resources for the tutorials ahead, you can explore the offerings on E2E CLOUD. We provide a diverse selection of GPUs, making us a suitable choice for more advanced LLM-based applications as well.

To get one, head over to MyAccount, and sign up. Then launch a GPU node as is shown in the screenshot below:

Make sure you add your ssh keys during launch, or through the security tab after launching. 

Once you have launched a node, you can use VSCode Remote Explorer to ssh into the node and use it as a local development environment. 

Chatting with GPT-3.5 Turbo and Caching Responses with GPTCache

In this tutorial, we will show you how to chat with GPT-3.5 Turbo and cache responses for both exact and similar matches using the GPTCache library. The purpose of caching responses is to save time and API usage when you have similar or identical questions. By caching responses, you can retrieve answers from the cache without sending requests to ChatGPT again.

This tutorial will be divided into the following:

  1. OpenAI API + GPTCache for Exact Match Cache
  2. OpenAI API + GPTCache for Similar Search Cache


Install the OpenAI and GPTCache libraries.

! pip install -q gptcache
! pip install openai

Set up your OpenAI API key. You can set the API key in your code using os.environ. To get a fresh OpenAI API key, you can visit here:

import os
os.environ["OPENAI_API_KEY"] = "YOUR_API_KEY"

Part 1: OpenAI API + GPTCache for Exact Match Cache

In this section, we will show you how to set up and use GPTCache for exact match caching. With exact match caching, you can store responses for the same question and retrieve them from the cache when the same question is asked again.

mport time
from gptcache import cache
from gptcache.adapter import openai as gpt_openai
def get_response_text(openai_resp):
    return openai_resp['choices'][0]['message']['content']
user_question = "what is github?"
for _ in range(2):
    start_time = time.time()
    response = gpt_openai.ChatCompletion.create(
        messages=[{'role': 'user','content': user_question}],
    print(f'User Question: {user_question}')
    print("Time consumed: {:.2f}s".format(time.time() - start_time))
    print(f'AI Answer: {get_response_text(response)}\n')

In this code, we set up the GPTCache with exact match caching. The response for the same question is cached, and when the question is asked again, the answer is retrieved from the cache without making a new request to GPT-3.5 Turbo.

Part 2: OpenAI API + GPTCache for Similar Search Cache

In this section, we will demonstrate how to set up and use GPTCache for similar search caching. Similar search caching allows you to retrieve responses from the cache when questions are similar but not identical.

import time
from gptcache import cache
from gptcache.adapter import openai
from gptcache.embedding import Onnx
from gptcache.manager import CacheBase
from gptcache.manager import VectorBase
from gptcache.manager import get_data_manager
from gptcache.similarity_evaluation.distance import SearchDistanceEvaluation
def get_response_text(openai_resp):
    return openai_resp['choices'][0]['message']['content']
onnx = Onnx()
data_manager = get_data_manager(CacheBase("sqlite"), VectorBase("faiss", dimension=onnx.dimension))
queries= ["What is github?","can you tell me more about github?",    "what is the purpose of github?"]
for user_question in queries:
    start_time = time.time()
    response = openai.ChatCompletion.create(
        messages=[{'role': 'user','content': user_question}],
    print(f'User Question: {user_question}')
    print("Time Consumed: {:.2f}s".format(time.time() - start_time))
    print(f'AI Answer: {get_response_text(response)}\n')

In this code, we set up the GPTCache for similar search caching. Questions that are similar to those in the cache will retrieve answers from the cache without making new requests to GPT-3.5 Turbo.

Now you have learned how to use GPT-3.5 Turbo and GPTCache to cache responses for exact and similar matches, saving time and API usage when interacting with the model. You can adapt this approach to various applications that involve interacting with chatbots and question-answering systems.

Use Cases for GPTCache

GPTCache is not suitable for all LLM applications, as its effectiveness depends on the cache hit rate. To maximize the return on investment, GPTCache is most advantageous in the following practical situations:

  1. Specialized Domains: LLM applications designed for specific domains of expertise, such as law, biology, medicine, finance, and other specialized fields.
  2. Specific Use Cases: LLM applications applied to specific use cases, like internal company chat bots or personal assistants.
  3. User Profiling: LLM applications with large user groups can benefit from using the same cache for user groups with the same profile if user profiling and classification can be done.


GPTCache offers a new way to significantly reduce costs and boost the speed of your LLM-based applications. By implementing semantic caching, it opens up opportunities for developers to create more cost-efficient and responsive solutions. As the project continues to develop, it's important to stay updated with the latest documentation and release notes for the most current information.


Research Paper: GPTCache - An Open-Source Semantic Cache for LLM Applications

Enabling Faster Answers and Cost Savings

GPTCache : A Library for Creating Semantic Cache for LLM Queries

GPTCache Quick Start

Latest Blogs
This is a decorative image for: A Complete Guide To Customer Acquisition For Startups
October 18, 2022

A Complete Guide To Customer Acquisition For Startups

Any business is enlivened by its customers. Therefore, a strategy to constantly bring in new clients is an ongoing requirement. In this regard, having a proper customer acquisition strategy can be of great importance.

So, if you are just starting your business, or planning to expand it, read on to learn more about this concept.

The problem with customer acquisition

As an organization, when working in a diverse and competitive market like India, you need to have a well-defined customer acquisition strategy to attain success. However, this is where most startups struggle. Now, you may have a great product or service, but if you are not in the right place targeting the right demographic, you are not likely to get the results you want.

To resolve this, typically, companies invest, but if that is not channelized properly, it will be futile.

So, the best way out of this dilemma is to have a clear customer acquisition strategy in place.

How can you create the ideal customer acquisition strategy for your business?

  • Define what your goals are

You need to define your goals so that you can meet the revenue expectations you have for the current fiscal year. You need to find a value for the metrics –

  • MRR – Monthly recurring revenue, which tells you all the income that can be generated from all your income channels.
  • CLV – Customer lifetime value tells you how much a customer is willing to spend on your business during your mutual relationship duration.  
  • CAC – Customer acquisition costs, which tells how much your organization needs to spend to acquire customers constantly.
  • Churn rate – It tells you the rate at which customers stop doing business.

All these metrics tell you how well you will be able to grow your business and revenue.

  • Identify your ideal customers

You need to understand who your current customers are and who your target customers are. Once you are aware of your customer base, you can focus your energies in that direction and get the maximum sale of your products or services. You can also understand what your customers require through various analytics and markers and address them to leverage your products/services towards them.

  • Choose your channels for customer acquisition

How will you acquire customers who will eventually tell at what scale and at what rate you need to expand your business? You could market and sell your products on social media channels like Instagram, Facebook and YouTube, or invest in paid marketing like Google Ads. You need to develop a unique strategy for each of these channels. 

  • Communicate with your customers

If you know exactly what your customers have in mind, then you will be able to develop your customer strategy with a clear perspective in mind. You can do it through surveys or customer opinion forms, email contact forms, blog posts and social media posts. After that, you just need to measure the analytics, clearly understand the insights, and improve your strategy accordingly.

Combining these strategies with your long-term business plan will bring results. However, there will be challenges on the way, where you need to adapt as per the requirements to make the most of it. At the same time, introducing new technologies like AI and ML can also solve such issues easily. To learn more about the use of AI and ML and how they are transforming businesses, keep referring to the blog section of E2E Networks.

Reference Links

This is a decorative image for: Constructing 3D objects through Deep Learning
October 18, 2022

Image-based 3D Object Reconstruction State-of-the-Art and trends in the Deep Learning Era

3D reconstruction is one of the most complex issues of deep learning systems. There have been multiple types of research in this field, and almost everything has been tried on it — computer vision, computer graphics and machine learning, but to no avail. However, that has resulted in CNN or convolutional neural networks foraying into this field, which has yielded some success.

The Main Objective of the 3D Object Reconstruction

Developing this deep learning technology aims to infer the shape of 3D objects from 2D images. So, to conduct the experiment, you need the following:

  • Highly calibrated cameras that take a photograph of the image from various angles.
  • Large training datasets can predict the geometry of the object whose 3D image reconstruction needs to be done. These datasets can be collected from a database of images, or they can be collected and sampled from a video.

By using the apparatus and datasets, you will be able to proceed with the 3D reconstruction from 2D datasets.

State-of-the-art Technology Used by the Datasets for the Reconstruction of 3D Objects

The technology used for this purpose needs to stick to the following parameters:

  • Input

Training with the help of one or multiple RGB images, where the segmentation of the 3D ground truth needs to be done. It could be one image, multiple images or even a video stream.

The testing will also be done on the same parameters, which will also help to create a uniform, cluttered background, or both.

  • Output

The volumetric output will be done in both high and low resolution, and the surface output will be generated through parameterisation, template deformation and point cloud. Moreover, the direct and intermediate outputs will be calculated this way.

  • Network architecture used

The architecture used in training is 3D-VAE-GAN, which has an encoder and a decoder, with TL-Net and conditional GAN. At the same time, the testing architecture is 3D-VAE, which has an encoder and a decoder.

  • Training used

The degree of supervision used in 2D vs 3D supervision, weak supervision along with loss functions have to be included in this system. The training procedure is adversarial training with joint 2D and 3D embeddings. Also, the network architecture is extremely important for the speed and processing quality of the output images.

  • Practical applications and use cases

Volumetric representations and surface representations can do the reconstruction. Powerful computer systems need to be used for reconstruction.

Given below are some of the places where 3D Object Reconstruction Deep Learning Systems are used:

  • 3D reconstruction technology can be used in the Police Department for drawing the faces of criminals whose images have been procured from a crime site where their faces are not completely revealed.
  • It can be used for re-modelling ruins at ancient architectural sites. The rubble or the debris stubs of structures can be used to recreate the entire building structure and get an idea of how it looked in the past.
  • They can be used in plastic surgery where the organs, face, limbs or any other portion of the body has been damaged and needs to be rebuilt.
  • It can be used in airport security, where concealed shapes can be used for guessing whether a person is armed or is carrying explosives or not.
  • It can also help in completing DNA sequences.

So, if you are planning to implement this technology, then you can rent the required infrastructure from E2E Networks and avoid investing in it. And if you plan to learn more about such topics, then keep a tab on the blog section of the website

Reference Links

This is a decorative image for: Comprehensive Guide to Deep Q-Learning for Data Science Enthusiasts
October 18, 2022

A Comprehensive Guide To Deep Q-Learning For Data Science Enthusiasts

For all data science enthusiasts who would love to dig deep, we have composed a write-up about Q-Learning specifically for you all. Deep Q-Learning and Reinforcement learning (RL) are extremely popular these days. These two data science methodologies use Python libraries like TensorFlow 2 and openAI’s Gym environment.

So, read on to know more.

What is Deep Q-Learning?

Deep Q-Learning utilizes the principles of Q-learning, but instead of using the Q-table, it uses the neural network. The algorithm of deep Q-Learning uses the states as input and the optimal Q-value of every action possible as the output. The agent gathers and stores all the previous experiences in the memory of the trained tuple in the following order:

State> Next state> Action> Reward

The neural network training stability increases using a random batch of previous data by using the experience replay. Experience replay also means the previous experiences stocking, and the target network uses it for training and calculation of the Q-network and the predicted Q-Value. This neural network uses openAI Gym, which is provided by taxi-v3 environments.

Now, any understanding of Deep Q-Learning   is incomplete without talking about Reinforcement Learning.

What is Reinforcement Learning?

Reinforcement is a subsection of ML. This part of ML is related to the action in which an environmental agent participates in a reward-based system and uses Reinforcement Learning to maximize the rewards. Reinforcement Learning is a different technique from unsupervised learning or supervised learning because it does not require a supervised input/output pair. The number of corrections is also less, so it is a highly efficient technique.

Now, the understanding of reinforcement learning is incomplete without knowing about Markov Decision Process (MDP). MDP is involved with each state that has been presented in the results of the environment, derived from the state previously there. The information which composes both states is gathered and transferred to the decision process. The task of the chosen agent is to maximize the awards. The MDP optimizes the actions and helps construct the optimal policy.

For developing the MDP, you need to follow the Q-Learning Algorithm, which is an extremely important part of data science and machine learning.

What is Q-Learning Algorithm?

The process of Q-Learning is important for understanding the data from scratch. It involves defining the parameters, choosing the actions from the current state and also choosing the actions from the previous state and then developing a Q-table for maximizing the results or output rewards.

The 4 steps that are involved in Q-Learning:

  1. Initializing parameters – The RL (reinforcement learning) model learns the set of actions that the agent requires in the state, environment and time.
  2. Identifying current state – The model stores the prior records for optimal action definition for maximizing the results. For acting in the present state, the state needs to be identified and perform an action combination for it.
  3. Choosing the optimal action set and gaining the relevant experience – A Q-table is generated from the data with a set of specific states and actions, and the weight of this data is calculated for updating the Q-Table to the following step.
  4. Updating Q-table rewards and next state determination – After the relevant experience is gained and agents start getting environmental records. The reward amplitude helps to present the subsequent step.  

In case the Q-table size is huge, then the generation of the model is a time-consuming process. This situation requires Deep Q-learning.

Hopefully, this write-up has provided an outline of Deep Q-Learning and its related concepts. If you wish to learn more about such topics, then keep a tab on the blog section of the E2E Networks website.

Reference Links

This is a decorative image for: GAUDI: A Neural Architect for Immersive 3D Scene Generation
October 13, 2022

GAUDI: A Neural Architect for Immersive 3D Scene Generation

The evolution of artificial intelligence in the past decade has been staggering, and now the focus is shifting towards AI and ML systems to understand and generate 3D spaces. As a result, there has been extensive research on manipulating 3D generative models. In this regard, Apple’s AI and ML scientists have developed GAUDI, a method specifically for this job.

An introduction to GAUDI

The GAUDI 3D immersive technique founders named it after the famous architect Antoni Gaudi. This AI model takes the help of a camera pose decoder, which enables it to guess the possible camera angles of a scene. Hence, the decoder then makes it possible to predict the 3D canvas from almost every angle.

What does GAUDI do?

GAUDI can perform multiple functions –

  • The extensions of these generative models have a tremendous effect on ML and computer vision. Pragmatically, such models are highly useful. They are applied in model-based reinforcement learning and planning world models, SLAM is s, or 3D content creation.
  • Generative modelling for 3D objects has been used for generating scenes using graf, pigan, and gsn, which incorporate a GAN (Generative Adversarial Network). The generator codes radiance fields exclusively. Using the 3D space in the scene along with the camera pose generates the 3D image from that point. This point has a density scalar and RGB value for that specific point in 3D space. This can be done from a 2D camera view. It does this by imposing 3D datasets on those 2D shots. It isolates various objects and scenes and combines them to render a new scene altogether.
  • GAUDI also removes GANs pathologies like mode collapse and improved GAN.
  • GAUDI also uses this to train data on a canonical coordinate system. You can compare it by looking at the trajectory of the scenes.

How is GAUDI applied to the content?

The steps of application for GAUDI have been given below:

  • Each trajectory is created, which consists of a sequence of posed images (These images are from a 3D scene) encoded into a latent representation. This representation which has a radiance field or what we refer to as the 3D scene and the camera path is created in a disentangled way. The results are interpreted as free parameters. The problem is optimized by and formulation of a reconstruction objective.
  • This simple training process is then scaled to trajectories, thousands of them creating a large number of views. The model samples the radiance fields totally from the previous distribution that the model has learned.
  • The scenes are thus synthesized by interpolation within the hidden space.
  • The scaling of 3D scenes generates many scenes that contain thousands of images. During training, there is no issue related to canonical orientation or mode collapse.
  • A novel de-noising optimization technique is used to find hidden representations that collaborate in modelling the camera poses and the radiance field to create multiple datasets with state-of-the-art performance in generating 3D scenes by building a setup that uses images and text.

To conclude, GAUDI has more capabilities and can also be used for sampling various images and video datasets. Furthermore, this will make a foray into AR (augmented reality) and VR (virtual reality). With GAUDI in hand, the sky is only the limit in the field of media creation. So, if you enjoy reading about the latest development in the field of AI and ML, then keep a tab on the blog section of the E2E Networks website.

Reference Links

Build on the most powerful infrastructure cloud

A vector illustration of a tech city using latest cloud technologies & infrastructure