1. X
  2. Dev Patel
Log inSign up
Dev Patel
267 posts
Image
user avatar
Dev Patel
@devpatelio
making magic machines @ucberkeley | @novaskyai • @amd • @healthenginecal
SF, CA
devpatelio.com
Joined June 2022
1,141
Following
531
Followers
RepliesRepliesMediaMedia
  • user avatar
    Dev Patel
    @devpatelio
    Feb 17
    SkyRL 🤝 Harbor is the best way to train terminal agents.
    user avatar
    Charlie Ruan
    @charlie_ruan
    Feb 17
    Releasing the official SkyRL + Harbor integration: a standardized way to train terminal-use agents with RL. From the creators of Terminal-Bench, Harbor is a widely adopted framework for evaluating terminal-use agents on any task expressible as a Dockerfile + instruction + test
    Image
  • user avatar
    Dev Patel
    @devpatelio
    Feb 13
    We’re super excited to have SkyRL be fully implementing Tinker!
    user avatar
    Tyler Griggs
    @tyler_griggs_
    Feb 13
    SkyRL now implements the Tinker API. Now, training scripts written for Tinker can run on your own GPUs with zero code changes using SkyRL's FSDP2, Megatron, and vLLM backends. Blog: novasky-ai.notion.site/skyrl-tinker 🧵
    Image
  • user avatar
    Dev Patel
    @devpatelio
    Feb 3
    One of the most unique aspects of venture in today’s climate is that as the LP base changes and the timelines to revenue compress dramatically for your AI-enabled businesses, there’s gonna be a huge explosion in creative vehicles and infrastructure to support the next generation
    user avatar
    Will Manidis
    @WillManidis
    Feb 2
    Article cover image
    Article
    Patient Capital Will Eat the World
    Venture capital has become a very, very big asset class. Just a couple of weeks ago, Andreessen Horowitz announced a $15 billion in newly committed capital. This is a single fund representing a half...
  • user avatar
    Dev Patel
    @devpatelio
    Jan 26
    a very awesome project made with @NovaSkyAI SkyRL!
    user avatar
    Kanishk Gandhi
    @gandhikanishk
    Jan 26
    (1/9) We built Endless Terminals: a fully autonomous pipeline that procedurally generates terminal tasks for RL training with no human annotation needed. Simple PPO + scaled environments give consistent improvements on downstream tasks like Terminal Bench 2.0!
    Image
  • user avatar
    Dev Patel
    @devpatelio
    Dec 8, 2025
    v0.3.0 is a major step change, lots more exciting stuff on the roadmap!
    user avatar
    NovaSky
    @NovaSkyAI
    Dec 8, 2025
    We recently released SkyRL-Train v0.3.0! Highlights include: - Experimental support for Pipeline-RL style Async-RL - Updated E2E Recipes page with Math, Search, SQL runs - Migration from mbridge -> Megatron-Bridge - 14 new OSS contributors! (1/n) 🧵
    Image

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
Advertisement
Advertisement