Dylan's Projects
Resources for skilling up in AI alignment research engineering. Covers basics of deep learning, mechanistic interpretability, and RL.
An experimental open-source attempt to make GPT-4 fully autonomous.
An improved version of the original AutoDD - with installer.
Awesome Papers related to Mamba.
Let ChatGPT teach your own chatbot in hours with a single GPU!
Don't let these adorable strays go unnoticed - add your sightings to the map today!
the AI-native open-source embedding database
DigitsUI is a highly configurable and ML-leveraged quantitative analysis platform for options traders and aspiring market makers š
digits-api-altdata is the Digits service responsible for retrieving alternative data, which is sort of a catchall for "not scientifically evaluated" in the investing world.
digits-api-main is the authentication, preprocessing, and tertiary API bridge in the Digits service mesh.
digits-api-ml is a large suite of API endpoints that directly respond with the inference of a given machine learning function.
digits-api-tda is the service responsible for handling TDAmeritrade API data to the main API and the UI.
The official helm chart repository for open-source market analysis platform, Digits.
Open-source Embeddings
:mag: LLM orchestration framework to build customizable, production-ready LLM applications. Connect components (models, vector DBs, file converters) to pipelines or agents that can interact with your data. With advanced retrieval methods, it's best suited for building RAG, question answering, semantic search or conversational agent chatbots.
hygiene (šŖ„) is a data preprocessing toolkit that makes it easy to create common LLM-related data structures; from training data to chain payloads!
This is based on instagrambot/instabot and is basically a nice wrapper to do essential functions. Also bundled into an .exe with pyinstaller.
I wrote a script that miserably hacks the instagram page of a user and parses data after cleaning it up and using lists. I will come up with a better way to do it but Instagram really does not want bots scraping their site, I've found out.
NATS on Kubernetes with Helm Charts
LlamaIndex (GPT Index) is a project that provides a central interface to connect your LLM's with external data.
OpenCV implementation using a webcam as input for a haar-cascade-based vision recognition system.
small, efficient embedding (SL, 'small language') model toolkit ~ finetune SOTA models on knowledge bases, rapidly ~
Convert PDF to markdown quickly with high accuracy
This repository is for Milvus technical documentation update and maintenance. Visit Milvus.io for fully rendered technical documents.
A trojan horse that drops a portable runtime of MinerGate's CLI CPU cryptocurrency miner. Includes basic persistance.
A minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training
MicroPython NATS.io Client
Self-referencing