BlackHC/toma
Helps you write algorithms in PyTorch that adapt to the available (CUDA) memory
Past: AIMS DPhil in Oxford at @OATML; Fellow at @nwspk; RE at DeepMind; SWE at Google.
Repositories
BlackHC/BatchBALD
Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning.
BlackHC/llm-strategy
Directly Connecting Python to LLMs via Strongly-Typed Functions, Dataclasses, Interfaces & Generic Types
BlackHC/batchbald_redux
Reusable BatchBALD implementation
BlackHC/dart_repl
Proof of concept REPL shell for Dart
BlackHC/neural_net_checklist
BlackHC/continual_learning_iid
BlackHC/tfpyth
Putting TensorFlow back in PyTorch, back in TensorFlow (differentiable TensorFlow PyTorch adapters).
BlackHC/lean-codex-formalization
BlackHC/tinker-cookbook
Post-training with Tinker
BlackHC/mnist_by_zip
Compression algorithms (like the well-known zip file compression) can be used for machine learning purposes, specifically for classifying hand-written digits (MNIST)
BlackHC/notebook_setup
Some batteries for Jupyter notebooks
BlackHC/implicit_lambda
This package adds support for implicit lambdas, so you can write `map(_ + 5, a_list)` instead of `map(lambda x: x + 5, a_list)`.
BlackHC/pytorch_datadiet
BlackHC/dvc
🦉Data Version Control | Git for Data & Models | ML Experiments Management
BlackHC/Dayflow
Generate a timeline of your day, automatically
BlackHC/infer_return_type
BlackHC/modded-nanogpt
NanoGPT (124M) in 3 minutes
BlackHC/mathlib4_docs
BlackHC/gemini-cli
An open-source AI agent that brings the power of Gemini directly into your terminal.
BlackHC/chatplayground
Chat Playground for LLMs
BlackHC/iot-river-crossing-jules
BlackHC/2409.02628
(Implicit) Ensembles of Ensembles: Epistemic Uncertainty Collapse in Large Models
BlackHC/iot-river-crossing-codex
BlackHC/fasthtml
The fastest way to create an HTML app
BlackHC/ddu_dirty_mnist
Dirty-MNIST dataset introduced in "Deterministic Neural Networks with Inductive Biases Capture Epistemic and Aleatoric Uncertainty" (https://arxiv.org/abs/2102.11582)
BlackHC/perplexity-correlations
Simple and scalable tools for data-driven pretraining data selection.
BlackHC/llmtracer
Trace LLM calls (and others) and visualize them in WandB, as interactive SVG or using a streaming local webapp
BlackHC/batch_pong_poc
Instead of running one environment at a time or one per thread, run everything in batch using numpy on a single core.