JINO ROHIT

advanced_ml. Jupyter Notebook

fastcv. fastcv is a CUDA rewrite of the opencv filters with python bindings

inferGPT. a simple c++ inference engine for gpt based architecture

rag_with_kg. RAG with knowledge graphs implemented from scratch

fastapi_with_celery_redis. serving a torch model using Celery, Redis and RabbitMQ to serve users asynchronously

nano-paged-attention. a minimal paged attention implementation

kernels. writing really fast kernels

Tune-RAG-Parameters-With-LlamaIndex. Python

tachyon. a LLM inference engine to run on consumer hardware

ml-math-in-depth.

Full-Stack-RAG. JavaScript

tachyon-deprecated-. a C++ inference engine for LLM

llm-inference-techniques. a guide to learn and implement inference techniques from scratch

nano-llama.cpp. a repo to understand llama.cpp

c-gym. c++ algorithms and problems

RAG-with-Memory. An attempt to adding memory to RAG

gpt-triton. Python

ORCA. implementation for ORCA - A Distributed Serving System for Transformer-Based Generative Models

ml_reading_list. fundamental ml paper notes

k8_modelserve. Model deploy with FASTAPI and K8s

real_time_rag_kafka. rag with kafka pipelines

ml_batch_server_fastapi_postgres. a batch prediction setup using fastapi and postgres as a database

inferLlama. a simple c++ inference engine for llama based architecture

Flash-Attention-Series. a series of softmax experiments to derive flash attention

june. LLM Agent Assistants

llamaindex_chat_assistant_aws. Python

RAG-semantic-caching. Python

Computer-Vision-Zero-to-Hero. Jupyter Notebook

airflow_orchestration_ml_pipeline. Python

rag_with_slackbot. rag with a slackbot

Embedding-Quantization. Python

mnist-cuda. CUDA rewrite of a simple NN in cuda c++

Image-Captioning-System. Python

Subway-Surfers-AI-using-Behavioral-Cloning. Jupyter Notebook

fastapi_react_stablediffusion. fastapi + react = stablediffusion

Airflow-ML-Docker. Python

ultramicrograd. a playground autograd framework

mlflow_training_pipeline. torch pipeline with mlflow tracking

mini_gpt_engineer. Python

JINO-ROHIT.github.io. HTML

cutlass. CUDA Templates and Python DSLs for High-Performance Linear Algebra

datasets. 🤗 The largest hub of ready-to-use datasets for ML models with fast, easy-to-use and efficient data manipulation tools