Home - KrArunT/InfobellIT-Gen-AI GitHub Wiki

Welcome to the InfobellIT Gen-AI wiki master!

Projects

Infobell-IT Gen-AI

Scalable LLM deployment on On-Prem/Cloud Servers and Clusters.

  • OpenSift/Nutanix/Vmware/SUSE/Rancher/Vanilla Kubernetes
  • LLama2/LLama3/LLama3.1/LLama3.2 many others and variants.

LLM Model Inference & Inference Server optimizations.

LLM Model Training/Fine-Tuning

Audio Projects

  • Chatbot for diastheria patients.
  • Audio Fine-Tuning (STT/TTS)

Datasets

RAG

  • Augmented Generation (AG) Context Augmentation to prompt either by zero shot , one-shot few shot or web search.
  • [Retrieval Augmented Generations (RAG)]
  • [RAG with Web search]

Agentic AI and Workflows

Copilots & AI Assistants

LLM Performance Benchmarking (Echo-Swift)

Community LLM Benchmark Tools

  • vLLM Benchmark Tools
  • TGI Benchmark
  • Ollama Benchmark metadata

MLPerf LLM Benchmarking

  • LLama2-70B
  • LLama2-7B
  • Mixtral
  • GPT-J

AIOPs/MLOPs/LLMOPs

  • MLFlow
  • KubeFlow

Model Engineering

  • Model Quantization
  • Model Pruning
  • Knowledge Distillation
  • Model Performance Optimizations

LLM Applications and Use-cases

  • Convogene.ai A universal chatbot based on RAG. Having doc-chat with (pdf and web scrapped data)

Publications

  1. OpenShift
  2. Nutanix
  3. Vmware
  4. Suse/Rancher
  5. NVIDIA NIM
  6. Paper
  7. EcoSwift

Challenges & Roadmap

  • Faster models with longer context.
  • AI Based Automation of Enterprise complex workflows.
  • RAG database update with evolution of time.

Small Language Models(SLM)

Vision Models

Multi-modal Modals

Contributing

  • Fork the repository
  • Create a feature branch: git checkout -b feature/amazing-feature
  • Commit changes: git commit -m 'Add amazing feature'
  • Push to branch: git push origin feature/amazing-feature
  • Open a Pull Request