AI Stock Adviser
Developed an AI chatbot with OpenAI as backend to streamline stock investment advising. Implemented tools for real-time and historical data gathering, including stock prices, financial statements, and news.
Features : Precise stock info, Forecasting future prices, Analysis, Risk assesment, Visualization
Tools/Language: OpenAI LLM, Langchain, Agents, LSTMs, Yahoo finance, Web scrapping
View Project →
Retrieval-Augmented Generation (RAG) in Large Language Models
Developed a Retrieval Augmented Generation (RAG) bot featuring integrated vector storage using LangChain and Llama3 via Groq API. Implemented vector storage with Qdrant Cloud to optimize document search efficiency.
Tools/Language: Langchain, OpenAI LLM, Vector DB
View Project →
QLORA fine tuning LLama
Implemented QLoRA fine-tuning on Llama 1.1B using the Guanaco chat dataset.Utilized HuggingFace Transformers libraries (PEFT, TRL, Trainer) for efficient model adaptation.
Tools/Language: LLMs, Transfomers, QLORA
View Project →
Flower Identification System
Developed CNN model fine tuning EfficientNetB7 and ResNet50 architectures. Trained on a dataset of 13,000 images spanning 102 class labels for precise flower classification. Achieved significant accuracy improvements.
Tools/Language: TensorFlow, Transfer learning, EfficientNetB7, ResNet50
View Project →
In Context Learning using Flan T5
Developed a dialogue summarization system using the FLAN-T5 model, employing zero-shot, one-shot, and few-shot inference methods. Utilized the "knkarthick/dialogsum" dataset to enhance summarization accuracy
Tools/Language: Transformers, FlanT5, LLMs
View Project →
