tom-pollak
Follow
Lists (12)
Stars
rl from zero pretrain, can it be done? yes.
Postmodern immutable and persistent data structures for C++ — value semantics at scale
Helpful kernel tutorials and examples for tile-based GPU programming
https://marketplace.visualstudio.com/items?itemName=TomPollak.lazygit-vscode
A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.
MoE training for Me and You and maybe other people
Finally a Fabioulous & Fast Fuzzy File Finder for neovim
llms can learn their own context compression via RL
A debugging and profiling tool that can trace and visualize python code execution
Official Problem Sets / Reference Kernels for the GPU MODE Leaderboard!
We want to compare how good Qwen3-1.7B-Base using B200 to continue pretraining on Malaysian multi-lingual corpus on different mixed precision training with proper truncated multi-packing.
Super basic implementation (gist-like) of RLMs with REPL environments.
Triton-based Symmetric Memory operators and examples
From baby GPT to diffusion GPT: An annotated implementation of a character-level discrete diffusion model (adapted from Karpathy’s baby GPT).
This is a beginner-friendly tutorial on MLIR from the perspective of a user of MLIR, not a compiler engineer. This tutorial will introduce why MLIR exists and how it is used to compile code at diff…
Use Claude Code with any LLM provider - GLM-4.5, Kimi-K2, Qwen3-Coder, DeepSeek, etc.
A non-saturating, open-ended environment for evaluating LLMs in Factorio
Renderer for the harmony response format to be used with gpt-oss
Hierarchical Reasoning Model Official Release
Official Repository of Absolute Zero Reasoner
Atropos is a Language Model Reinforcement Learning Environments framework for collecting and evaluating LLM trajectories through diverse environments