Skip to content
View pjo256's full-sized avatar
21 followers · 7 following

Achievements

Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Mars 2020 ContributorAchievement: Arctic Code Vault ContributorAchievement: Pull Sharkx2

Achievements

Achievement: Pair ExtraordinaireAchievement: QuickdrawAchievement: Mars 2020 ContributorAchievement: Arctic Code Vault ContributorAchievement: Pull Sharkx2

Organizations

@NYUHackDays

Block or report pjo256

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
pjo256/README.md

Hi, I'm Philip 👋

📍 Stuttgart ↔ NYC | 🚀 Co-founded Vimeo's Applied ML team

Python Go TypeScript Codex Claude Web

Building ML systems for video and audio. Previously Principal ML Engineer at Vimeo, shipping AI features to 450k daily uploads.

Dabbling in OSS while exploring what's next.

Contributions

  • kubernetes - Check rollout status across multiple deployments at once
  • PufferLib - Train agent to dock spacecraft in orbit with RL
  • datadog-agent - Cut GPU monitoring overhead by 9–23x, freeing ~250ms of CPU per second on 8-GPU nodes
  • 📜 microrlhf - Dependency-free RLHF in Python, fine-tuning GPT to generate Tolkien-style names. Inspired by @karpathy's microgpt
  • NeMo-RL - Eliminate a multi-GPU bottleneck in distributed RL training
  • tunix - Catch silent sample loss in micro-batches during JAX RL fine-tuning
  • transformers - Faster tokenization for multilingual models
  • vllm - 2.5x faster tensor copies during inference batching
  • 🦙 LlamaFactory - Support mixed-modality inputs (image, video, audio) during LLM fine-tuning
  • verifiers - Accurate scoring metrics across rubrics for RLHF
  • PathOfBuilding - Fixed Coiling Whisper disabling all curses
  • Bugcrowd - Reporting vulnerabilities through bug bounty programs

Pinned Loading

  1. microrlhf.py microrlhf.py
    1 
    """
    2 
    An atomic way to RLHF a language model in pure, dependency-free Python.
    3 
    Uses a discriminative n-gram classifier P(LOTR | name) + REINFORCE to steer microgpt toward Tolkienian names.
    4 
    Inspired by @karpathy's microgpt.py.
    5
  2. PufferLib PufferLib Public

    Forked from PufferAI/PufferLib

    Simplifying reinforcement learning for complex game environments

    C

  3. vllm vllm Public

    Forked from vllm-project/vllm

    A high-throughput and memory-efficient inference and serving engine for LLMs

    Python

  4. kubernetes kubernetes Public

    Forked from kubernetes/kubernetes

    Production-Grade Container Scheduling and Management

    Go 1

  5. transformers transformers Public

    Forked from huggingface/transformers

    🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

    Python

  6. PathOfBuildingCommunity/PathOfBuilding PathOfBuildingCommunity/PathOfBuilding Public

    Offline build planner for Path of Exile.

    Lua 5.1k 2.3k