ajay's starred repositories
screenshot-to-code
Drop in a screenshot and convert it to clean code (HTML/Tailwind/React/Vue)
PaddleOCR
Awesome multilingual OCR toolkits based on PaddlePaddle (practical ultra lightweight OCR system, support 80+ languages recognition, provide data annotation and synthesis tools, support training and deployment among server, mobile, embedded and IoT devices)
llm-course
Course to get into Large Language Models (LLMs) with roadmaps and Colab notebooks.
LLMs-from-scratch
Implementing a ChatGPT-like LLM in PyTorch from scratch, step by step
unstructured
Open source libraries and APIs to build custom preprocessing pipelines for labeling, training, or production machine learning pipelines.
search_with_lepton
Building a quick conversation-based search demo with Lepton AI.
FlagEmbedding
Retrieval and Retrieval-augmented LLMs
AlphaCodium
Official implementation for the paper: "Code Generation with AlphaCodium: From Prompt Engineering to Flow Engineering""
AutoPrompt
A framework for prompt tuning using Intent-based Prompt Calibration
hallucination-leaderboard
Leaderboard Comparing LLM Performance at Producing Hallucinations when Summarizing Short Documents
local-llms-analyse-finance
In this project, I explored how local LLMs can be used to label data and support analyses. Specifically, I used Llama2 model to automatically categorise my bank transaction data.
minimal-trainer-zoo
Minimal example scripts of the Hugging Face Trainer, focused on staying under 150 lines
Indic-gemma-7b-Navarasa
Repository for fine-tuning gemma models using unsloth for indic languages
RAG-with-Cross-Encoder-Reranker
Testing speed and accuracy of RAG with, and without Cross Encoder Reranker.
blitz-embed
C++ inference wrappers for running blazing fast embedding services on your favourite serverless like AWS Lambda. By Prithivi Da, PRs welcome.