Deep dive into understanding GenAI model working

 Pricing Gemini api Cloud platforms

 Model deployment of LLAMA2 7b Model deployment steps: Understand the Model: Before deploying an LLM, it’s important to understand its capabilities, limitations, and the specific use case it will serve. Choose the Right Platform: Decide whether you want to deploy the model locally or on a cloud platform. Each has its own advantages and disadvantages. Prepare the Model: This could involve fine-tuning the model on your specific task or data1. Set Up the Infrastructure: This involves setting up the servers or cloud resources where the model will run. You’ll need to consider the computational resources required by the model. Deploy the Model: Use a model serving tool or platform to deploy the model. This could be a cloud service or an open-source tool. Ensure Security and Privacy: Implement measures to ensure the security of the model and the privacy of the data it handles. Monitor the Model: Once the model is deployed, monitor its performance and usage to ensure it’s working as expect...

 Giftguru  As per the discussion on 15th May 2024 cohere API is only free for developers and it is not free for commercial use which doesn't align with our objective.  Tried and tested various Large language models for our needs in text generation but there were some challenges faced during the testing of different LLM models that were suitable for our use case Facebook Bart model Challenges: Gave repetitive output similar to the input provided Flan T5: Challenges: Fulfilled all the requirements as needed for the model still the model didn't result in any output. Gemma model: Challenges: Requires GPU for running but there was a timeout for GPU in Colab and Kaggle is not verifying the license key used by me. Pros: It can work on our use case. Will resolve this issue by tomorrow LLama3 by meta: Challenges: Takes too much time to run a single prompt. Note: Gemini 1.0 API is a free tier but has some restrictions 1 minute only 15 request can be made and 15000 requests per day ...

 

giftguru.co.in Objective:   To build a website that will recommend gifts based on users' queries. Giftguru.co.in is a website-based platform to generate gift result recommendations for users based on their prompt and results will also include product links for the products. The initial development plan: Building a basic website with an input field Domain name selection Testing for the gen ai API for generating the result. Integrating Amazon PA API for result generation First Draft of the website: Testing the API's Challenges in Testing the APIs  We tried using Google Cloud API, but we were required to pay to access them. We tried using Openai's API money was the issue here. To resolve this we searched for publically available API and found co-here API. Result: Next steps Moving ahead we are planning for the generation of a new UI.  Trying to parse the output generated for the link generation. Mousam- Understanding the working of React js and working on link generat...

 Health report pdf->text OnlineOCR.net: It is a free tool available on the internet Input: Heath report pdf output:  VIJAYA   VIJAYA DIAGNOSTIC CENTRE®  H No. 2-137/10, Plot No.42, NH65, Opposite R.S. Brothers, Gangaram, Chanda Nagar, Hyderabad - 500050, Telangana  TEST REPORT  Name : Mr. VIJAY KUMAR GUNTIREDDY Registered on : 03-Apr-2024 08:33 Age/Gender : 41 Years / Male BirthDate : 09-Dec-1982 Collected on : 03-Apr-2024 08:42  Registration ID : 240840005878 Released on : 03-Apr-2024 13:49 Ref. By : Self Printed on : 03-Apr-2024 14:36 Sample Type : Serum Regn Centre : Nallagandla - 84  THYROID PROFILE  TEST NAME RESULT UNIT BIOLOGICAL REFERENCE INTERVAL T3 Total : 1.04 ng/mL 0.60 - 1.81 Method: Chemiluminiscence Immunoassay T4 Total : 8.30 p.g/dL 3.2 - 12.6 Method: Chemiluminiscence Immunoassay TSH - Ultrasensitive : 1.215 i-tIU/mL 0.55 - 4.78  Method: Chemiluminiscence Immunoassay  Interpretation / Comments :  • Pati...

Objective: Take an MRI of the brain as input and generate the probability of the disease and also disease associated with the MRI Input:  MRI image of the brain    Expected Outcome: Types of disease and the probability associated with it. Overview of the steps:- Researching the pre-existing LLM model and testing it for our requirements Collection of Brain MRI image dataset  Preparing a Multimodal LLM  Training the multimodal LLM on the image dataset Testing it for new data Evaluating the result Date- wise Plan 8/05/2024-10/05/2024 Research work on the LLM models 13/05/2024  - 15-05-2024 Coding for the LLM model  16-05-2024- 22-05-2024 Training the model 22-05-2024-23-05-2024 Testing and Evaluation 

 Input: Expected Output: Patient Information: Name: QWERTY Age: XX Date: DD/MM/YYYY Disease: Parkinson Disease Description: Parkinson's disease is a progressive disorder that affects the nervous system and the parts of the body controlled by the nerves. Symptoms start slowly. The first symptom may be a barely noticeable tremor in just one hand. Tremors are common, but the disorder also may cause stiffness or slowing of movement. In Parkinson's disease, certain nerve cells called neurons in the brain gradually break down or die. Many of the symptoms of Parkinson's are due to a loss of neurons that produce a chemical messenger in your brain called dopamine. When dopamine levels decrease, it causes irregular brain activity, leading to problems with movement and other symptoms of Parkinson's disease. Risk factors: Genetics Age Gender Environmental exposure: Exposure to pesticides, herbicides, solvents, oils, metals, toxins, and airborne pollutants may increase the risk of P...

 Work Review till 6.5.24 The deep learning model that we have: takes input as an image of an MRI of the brain output: classification as Alzheimer's and Parkinson's disease. We have completed the research part on the disease for the text generation on the disease and documented how it will be generated. As we have only two types of disease in this classification we have two approaches for this either template format or taking a pre-trained LLM model and training them on data related to Alzheimer's and Parkinson's disease. We are moving forward with the template format first. The here objective is to have our second model ready which will generate the output based on the output of our deep learning model.  Our deep learning model gives output as the classification if the MRI is of Alzheimer's or Parkinson's disease it doesn't classify the MRI in a mild case or severe case so we can't have a detailed report we can only have the report as what is the diseas...