Alpay Ariyak's Projects
Banner readme for the runpod-workers organization.
A web app that uses a BERT model of your choice to answer questions based on a reference text.
Using a Deep Q Network(DQN) to play Atari Breakout
Go ahead and axolotl questions
Fine-Tuning and Deploying BERT to predict the next element in a SMILES (Simplified Molecular Input Line Entry System) string.
Software developed for Brigham Women's Hospital in Massachusetts in a team of 10.
EvalPlus for rigourous evaluation of LLM-synthesized code
My implementation of WizardLM's Evol-Instruct
Classifying Fake News with 99.6% Accuracy using ML and Exploring NLP Data Processing Techniques
Fast & more realistic evaluation of chat language models. Includes leaderboard.
A scalable, efficient and Open Source Mixture of Experts (MoE) architecture to enable OSS AI to achieve SOTA (GPT-4 level) performance.
Open sourced backend for Martian's LLM Inference Provider Leaderboard
Enhanced ChatGPT Clone: Features OpenAI, GPT-4 Vision, Bing, Anthropic, OpenRouter, Google, AI model switching, message search, langchain, DALL-E-3, ChatGPT Plugins, OpenAI Functions, Secure Multi-User System, Presets, completely open-source for self-hosting. More features in development
Improving Mathematical Capabilities of Large Language Models
LLMPerf is a library for validating and benchmarking LLMs
An open-source framework for training large multimodal models.
OpenChat: Advancing Open-source Language Models with Imperfect Data
Expanding the GrIT Dataset for Kosmos 2
quick mvp for fun
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent