AI Assisted Emailing with Radiantloom Email Assist 7b

January 10, 2024


In today’s corporate world, email has become an essential tool for communication. It's not only used by companies, it has become a part of everyone's life. However, it’s not easy to manage a large email inbox every day, either for individuals or for businesses. 

In this blog post, we will explore AI assisted emailing with Large Language Models (LLMs), particularly Radiantloom Email Assist 7b. For there has been mounting evidence suggesting that smaller open access / open-source models are outperforming GPT-4 on specific tasks.

Radiantloom Email Assist 7b is the optimal email assistant. It’s a large language model which can help you summarize, organize, and manage your email. Just by its name, you can understand that this model contains 7 billion parameters. It is fine-tuned on Zephyr-7B Beta model, and it has been trained on a custom-curated dataset of 1,000 email-assistant summarization tasks.

Radiantloom is an AI-powered platform that helps businesses automate their workflows and improve their customer service. It offers a variety of services, like email assistance, customer support products and so on, powered by its proprietary large language models, which have been trained on a massive dataset of text and code.

Let's give a brief introduction to Zephyr-7B Beta as our email assistant LLM is fine-tuned over it.

Zephyr-7B Beta

Zephyr is a series of language models which have been trained to act as helpful assistants. Zephyr-7B-β is the second model in the series, and is a fine-tuned version of mistralai/Mistral-7B-v0.1 that was trained on a mix of publicly available, synthetic datasets using Direct Preference Optimization (DPO).

All of these are decoder type models of transformers. Unlike encoder-decoder models, which process both input and output sequences, decoder-only models process the output sequence only. This makes them well-suited for tasks such as text generation, translation, and question-answering. This requires only two different kinds of input:

An Initial Prompt

An initial prompt is a brief text that offers context for the task at hand. In a generation task, this prompt might describe the task itself, demonstrating how we can employ zero-shot learning, even when some input data is available. This approach allows us to utilize these models in a few-shot manner.

Let's have a look at the prompt template for Radiantloom email assist:

<>Given an input, provide a clear and concise summary and action items in less than 50 words.<>
Answer the question below from context below :
Convert the summary into a voice memo, keeping it under one minute.

The notation ‘<s>’ usually indicates the start of a section or a tag used to signify a specific type of content or formatting. ‘[INST]’ stands for ‘instruction’, denoting a specific set of instructions or guidelines to follow. ‘<<SYS>>’ could indicate a system message or instruction within a structured conversation or script. These tags help organize and interpret different parts of a document or conversation for easier understanding or processing.

Previous Word/Getting Context

Generating the first word is the most difficult. This type of model first understands the prompt about the task and it already has access to a vast vocabulary, learned from the training data. It utilizes this vocabulary to generate a list of potential first words that align with the context and the overall tone of the text. Then it uses statistical methods to calculate the probability of each potential first word and pick the one with the highest probability.

Once it generates the first word, it only takes the prompt and the previous word.

Radiantloom Email Assist 7b Features 

This LLM offers a vast number of features that can help you become efficient and increase your productivity. Just by using a proper prompt you can use these features, which include:

Email Summarization: This LLM is trained on generating automated crisp summaries of your emails. With this, you can quickly grasp the nature of your emails and identify key action items, saving time and enhancing your productivity.

Voice Memo Conversion: This model can convert your emails into voice memos, which allows you to catch up on relevant information in the middle of multi-tasking.

Chat Message Conversion: This model is also capable of converting your emails into chat messages, which you can use to communicate with colleagues or customers.

Use Cases

Radiantloom Email Assist 7b can be used by a large number of individuals and organizations. Here I have given a few examples:

Busy Professionals: These LLMs can help busy professionals save time and improve their productivity by summarizing emails and organizing inboxes, helping to find the gist of specific emails quickly.

Salespeople: Radiantloom Email Assist 7b can help salespeople communicate more effectively with customers by providing solid summaries of emails and converting emails into voice memos or chat messages.

Customer Service Representatives: This model can also help customer service representatives provide better service by helping them find specific emails quickly and providing summaries of customer inquiries.

Now, let’s do a deep dive into using the model and see how it performs with an example email.

AI Assisted Emailing

Launch a GPU node on E2E Networks.

Installing Dependencies

!pip install -q transformers peft accelerate bitsandbytes safetensors sentencepiece langchain sentence-transformers

Import Required Libraries

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from transformers import BitsAndBytesConfig, pipeline

Loading the Model

model_name = "aigeek0x0/radiantloom-email-assist-7b"

def load_quantized_model(model_name: str):
    bnb_config = BitsAndBytesConfig(
        load_in_4bit=True, 				# loading the model with 4bit quantization
        bnb_4bit_use_double_quant=True, # utilize double quantization for processing
        bnb_4bit_quant_type="nf4",		# quantization type: "nf4"

    model = AutoModelForCausalLM.from_pretrained(
    return model

We are loading the model in 4 bit quantization for causal language modeling. bnb_config defines a BitsandBytesConfig with specific settings for quantization. Have a look at the comments in the above script of the function load_quantized_model for understanding more about bnb_config. This function essentially configures and loads a pre-trained model with specific quantization settings (4-bit quantization using bfloat16 data type) for efficient and optimized causal language modeling tasks.

Initializing the Tokenizer

def initialize_tokenizer(model_name: str):
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    tokenizer.bos_token_id = 1  # Set beginning of sentence token id
    return tokenizer

Loading the Model and Tokenizer

model = load_quantized_model(model_name)
tokenizer = initialize_tokenizer(model_name)

# define stop token ids
stop_token_ids = [0]

Now, as our model is ready for generating some cool output, let's take an example email conversation and instruct the model to provide a clear and concise summary with action items in 50 words.

In the below prompt template you can observe that with <<SYS>> token, we are giving instruction to the model.

Specifying the Prompt

prompt = """

<>Given an input, provide a clear and concise summary and action items in less than 50 words.<>

Answer the question below from context below :

Convert the summary into a voice memo, keeping it under one minute:

Subject: Project Update and Next Steps for Marketing Campaign

Hi Emily,

I trust you're doing well. As we progress with our marketing campaign, I wanted to share a brief update on our recent developments.

We've successfully completed the initial phase of market research, identifying key demographics and potential areas for engagement. Our next step involves finalizing the content strategy for social media outreach and refining our targeted messaging.

Your input on the content approach, especially for our upcoming product launch, is crucial. Could you review the proposed social media content plan and suggest enhancements or additions that align with our campaign objectives?

Additionally, let's schedule a brief meeting this week to discuss these updates and finalize the strategy moving forward.

Looking forward to your insights and our upcoming discussion.

Best regards,




Inferencing the Model

The below code snippet performs text generation using a pre-trained language model based on the input prompt. Let's break down each step:

input_ids = tokenizer.encode(prompt, return_tensors="pt")
tokenizer.encode: Converts the prompt text into tokenized IDs using the provided tokenizer.
return_tensors="pt": Returns PyTorch tensors; the input is converted into tensors suitable for model processing.
Text Generation
output = model.generate(input_ids, max_length=512, temperature=0.1, repetition_penalty=1.1, top_p=0.7, top_k=50)
model.generate: Utilizes the model to generate text based on the provided input_ids.
max_length=512: Specifies the maximum length for the generated text.
temperature=0.1: Controls the randomness of the generated text; lower values lead to more deterministic output.
repetition_penalty=1.1: Modifies the likelihood of repeating tokens; values > 1.0 reduce repetitions.
top_p=0.7: Controls the cumulative probability for token selection in the nucleus sampling method.
top_k=50: Limits the selection of next tokens to the top-k most probable tokens.
Decoding Output
output_text = tokenizer.decode(output[0], skip_special_tokens=True)
tokenizer.decode: Converts the generated output tensor IDs back into human-readable text.
output[0]: Accesses the first generated output sequence.
skip_special_tokens=True: Excludes special tokens like padding or separator tokens from the decoded text.
Displaying Generated Text
print(output_text): Prints the decoded generated text.

input_ids = tokenizer.encode(prompt, return_tensors="pt")
output = model.generate(input_ids, max_length=512,
    	 temperature=0.1, repetition_penalty=1.1, top_p=0.7, top_k=50)
output_text = tokenizer.decode(output[0], skip_special_tokens=True)


Radianloom Email Assist Output

Summary: Alex provides an update on the marketing campaign, highlighting the completion of market research and the need for content strategy refinement. He requests Emily's feedback on the social media content plan and suggests scheduling a meeting to finalize the strategy.

Action Items

  • Review the proposed social media content plan.
  • Suggest improvements or additions aligned with campaign objectives.
  • Schedule a meeting with Alex to discuss the updates and finalize the strategy.

Wow, beautiful, right? Now let us compare this with the ChatGPT-3.5’s response and see whether a smaller fine-tuned model can outperform an LLM trained on a substantially large amount of data.

We can see the clear difference between the summarization and action item generation capability of Radiantloom and ChatGPT.

Now, let's integrate it with LangChain and build a ready-to-serve pipeline for the email assistant.

LangChain Integration

Importing LangChain Libraries

from langchain import HuggingFacePipeline
from langchain import PromptTemplate, LLMChain

Configuring Hugging Face Pipeline

pipeline = pipeline(

Running the Chain

llm = HuggingFacePipeline(pipeline=pipeline)

def email_assist(prompt):

  template = """[INST]
  Given an input, provide a clear and concise summary and action items in less than 50 	  	words.

  Answer the question below from context below:
  {question} [/INST] 

  question = """Convert the summary into a voice memo, keeping it under one minute:"""

  context = prompt

  prompt = PromptTemplate(template=template, input_variables=["question","context"])
  llm_chain = LLMChain(prompt=prompt, llm=llm)
  response ={"question":question,"context":context})
  return response

output = email_assist(prompt)


"Alex here with a project update. We've completed initial market research and identified key demographics. Next, we're finalizing social media content and messaging for our product launch. Emily, your input on the content strategy is crucial. Please review the plan and suggest improvements. Let's schedule a meeting this week to finalize the strategy. Thanks, Alex."

Action Items:

  • Review the social media content plan.
  • Suggest improvements and additions.
  • Schedule a meeting with Alex this week.


Evaluating large language model (LLM) based chat assistants is challenging due to their broad capabilities and the inadequacy of existing benchmarks in measuring human preferences. Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena - this paper established a method to address these issues and the Radiantloom Email Assist model received good results, according to initial assessments, with the utilization of GPT-4.


Radiantloom Email Assist 7b is a powerful tool that can help you to be efficient, improve your productivity, and communicate more effectively. If you are looking to manage your email inbox more effectively, I suggest you try Radiantloom Email Assist 7b. 

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