model-benchmarking

Rapid Evaluation Framework for climate data

Open source code for AIOpsServing

All-in-one AI workbench program for Vision AI model inference, evaluation, benchmarking, optimization & dataset management

A modular deep learning evaluation framework for benchmarking multiple CNN architectures across varied optimization strategies and training configurations. Built for scalable experimentation and transferability to real-world image classification tasks.

This repo contains a study on performance of LLMs on STS(Semantic Textual Similarity) Data.

Real-time AI model comparison and benchmarking platform with live benchmark data, streaming model responses, and side-by-side evaluation.

Predict stock prices using Linear Regression and LSTM models. Includes data preprocessing, visualization, and benchmarking tools for analyzing historical stock data.

A reproducible, leak-free machine learning benchmarking lab for regression model comparison, cross-validation, diagnostics, experiment tracking, and interactive Streamlit-based inference using the California Housing dataset.

Machine Learning Model using Decision Trees on US Voting Dataset

🌿 Analyze, optimize, and track LLM prompt energy usage, cost, and carbon footprint with research-backed recommendations and real-time dashboards.

A Streamlit web app that uses a Groq-powered LLM (Llama 3) to act as an impartial judge for evaluating and comparing two model outputs. Supports custom criteria, presets like creativity and brand tone, and returns structured scores, explanations, and a winner. Built end-to-end with Python, Groq API, and Streamlit.

This project demonstrates a production-grade Evaluation (Evals) Framework used to benchmark multiple Large Language Models (LLMs) against a "Source of Truth" NBA dataset.

Deep learning benchmarking project in PyTorch comparing MLP, CNN, and ResNet architectures on FashionMNIST with structured training pipelines, data augmentation, and evaluation across accuracy, F1-score, and per-class metrics.

Anonymized consumer classification model benchmarking high-dimensional K-Nearest Neighbors (KNN) against Logistic Regression, achieving 96% predictive accuracy.

Structured failure mode evaluation across GPT-4o-mini, Mistral-small, and Llama 3.1/3.3. 1,728 prompts, 7 failure categories, statistical analysis

Causal analysis framework using Double Machine Learning to quantitatively isolate the effect of model size on deep learning performance while controlling for confounders such as dataset size, training time, and hyperparameters.

Side-by-side benchmark of BiLSTM + Attention vs BERT for multi-class emotion classification with real-time inference comparison

InfernoBench: open-source AI model installers, local inference benchmarks, and reproducible browser labs. Wan2.1 and Sulphur-2 included.

Interactive Python CLI for benchmarking and comparing Amazon Bedrock and multi-provider foundation models side-by-side by latency, token usage, and cost, with timestamped Markdown reports.