Get protein embeddings from protein sequences

🏭 PDF text extraction pipeline: self-hosted, local-first, Docker-based

Hallucinations (Confabulations) Document-Based Benchmark for RAG. Includes human-verified questions and answers.

[ICLR 2024] Trajectory-as-Exemplar Prompting with Memory for Computer Control

たまに追加される論文メモ

Code for building self-expanding knowledge graphs with Outlines, vLLM, neo4j, and Modal.

This repo includes ChatGPT prompts made for any person, any profession for any work curated to use ChatGPT better.

Neural ngram language model in PyTorch.

Material for my VLDB'22 and BTW'23 tutorials on application for language models in data management

CodeT5 LLM fine-tuned with C++ code from KDE

Cyber Anomaly Detection using RNN Language model

Bulgarian spaCy natural language processing pipeline

Fill-in-the-BERT uses pre-trained BERT Masked Language Model for Infering the task of fill in the blanks.

Web application to help develop scraping techniques to create context in LLM prompts.

A GPT-2 powered web app that generates advice in the style of world-class experts:

Русско-мансийский переводчик

Keras/Tensorflow implementation of the decoder from the transformer as described in the paper: "Attention Is All You Need"

Character level language model trained on the British National Formulary (BNF).

Code repository for https://arxiv.org/abs/2004.01549