Nadav Timor's repositories
deep-learning-research-lab
A distributed computing infrastructure for simulations, model training & fine-tuning.
keyboardAnt.github.io
Personal blog about research, machine learning, deep learning, data science and software engineering.
distributed-speculative-inference
The fastest off-the-shelf inference algorithm of LLMs
nn-mem-vision
An infrastructure for empirical research: Studying the limits of neural networks for computer vision tasks.
optimal-stopping-game
A game of probability & optimization simulating Optimal Stopping Problem variants. Supports multiple agents with different policies.
simpleaichat
Python package for easily interfacing with chat apps, with robust features and minimal code complexity.
wis-glucose-challenge
This is the 1st Place Winner (1 out of 45) in a data science competition led by Prof. Eran Segal on time-series forecasting via neural networks (Kaggle-like).
bigcode-evaluation-harness
A framework for the evaluation of autoregressive code generation language models.
ai4code_repair_workshop
AI4Code syntax repair tutorial for IAP 2023
containerapps-albumapi-python
Container Apps Quickstart: Python Album API
datasets
🤗 The largest hub of ready-to-use datasets for ML models with fast, easy-to-use and efficient data manipulation tools
DeepSpeed
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
detecting-fake-text
Giant Language Model Test Room
Forked-VulnerabilityDetection
vulnerability detection in python source code with LSTM networks
GPTQ-for-SantaCoder
4 bits quantization of SantaCoder using GPTQ
Implicit-Regularization-Towards-Rank-Minimization-in-ReLU-Networks--presentation
Slides to present our research paper on the foundations of deep learning, “Implicit Regularization Towards Rank Minimization in ReLU Networks”.
openai-cookbook
Examples and guides for using the OpenAI API
optimum-benchmark
A repository for benchmarking HF Optimum's optimizations for inference and training.
picoGPT
An unnecessarily tiny implementation of GPT-2 in NumPy.
SciComPy
An introduction to Scientific Computing with Python, an advanced course for computer science students at IDC Herzliya
smartbugs-curated
SB Curated is a curated dataset of Solidity smart contracts annotated with tagged vulnerabilities. The dataset was created to evaluate the accuracy of automated analysis tools.
speculative-sampling
Simple implementation of Speculative Sampling in NumPy for GPT-2.