Charts
DataOn-chain
VIP
Market Cap
API
Rankings
CoinOSNew
CoinClaw🦞
Language
  • 简体中文
  • 繁体中文
  • English
Leader in global market data applications, committed to providing valuable information more efficiently.

Features

  • Real-time Data
  • Special Features
  • AI Grid

Services

  • News
  • Open Data(API)
  • Institutional Services

Downloads

  • Desktop
  • Android
  • iOS

Contact Us

  • Chat Room
  • Business Email
  • Official Email
  • Official Verification

Join Community

  • Telegram
  • Twitter
  • Discord

© Copyright 2013-2026. All rights reserved.

简体繁體English
|Legacy

Gradient releases the Echo-2 RL framework, enhancing AI research efficiency by over 10 times.

CN
PANews
Follow
1 month ago
AI summarizes in 5 seconds.

The distributed AI laboratory Gradient today released the Echo-2 distributed reinforcement learning framework (arxiv.org/pdf/2602.02192), aimed at breaking the training efficiency barrier in AI research. By achieving a complete decoupling of Learner and Actor at the architectural level, Echo-2 reduced the post-training cost of a 30B model from $4,500 to $425. It delivers over 10 times the scientific throughput under the same budget.

The framework utilizes computing and storage separation technology for asynchronous training (Async RL), offloading massive sampling computing power to unstable GPU instances and Parallax-based heterogeneous GPUs. Coupled with breakthroughs in bounded staleness, instance fault-tolerant scheduling, and self-developed Lattica communication protocol, it significantly improves training efficiency while ensuring model accuracy. Alongside the framework release, Gradient is also set to launch the RLaaS platform Logits, promoting a shift in AI research from "capital accumulation" to "efficiency iteration." Logits is now open for reservation to students and researchers worldwide (logits.dev).

About Gradient

Gradient is an AI laboratory dedicated to building distributed infrastructure, focusing on distributed training, servicing, and deployment of cutting-edge large models. Gradient has received support from top investment institutions and is building an open and efficient future intelligent era.

免责声明:本文章仅代表作者个人观点,不代表本平台的立场和观点。本文章仅供信息分享,不构成对任何人的任何投资建议。用户与作者之间的任何争议,与本平台无关。如网页中刊载的文章或图片涉及侵权,请提供相关的权利证明和身份证明发送邮件到support@aicoin.com,本平台相关工作人员将会进行核查。

OKX 活期简单赚币,让你的链上黄金生生不息
广告
|
|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

|
|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

Selected Articles by PANews

3 days ago
The Hidden Winners of the FTX Finale: The "Bankruptcy Arbitrage Feast" Behind the Distribution of 2.2 Billion Dollars and the Bloodletting Effect on TradFi
3 days ago
PA Illustration | One image to understand the major Web3 events in April 2026
3 days ago
Airdrops cannot make you rich, and edgeX does not need a community.
View More

Table of Contents

|
|
APP
Windows
Mac
Share To

X

Telegram

Facebook

Reddit

CopyLink

Related Articles

avatar
avatar律动BlockBeats
27 minutes ago
Dialogue with Pantera Founder: Bitcoin has reached escape velocity, traditional assets are being left behind.
avatar
avatarOdaily星球日报
1 hour ago
Weekly Editor's Picks (0328-0403)
avatar
avatar链捕手
2 hours ago
Is Circle still worth buying?
avatar
avatarOdaily星球日报
15 hours ago
Gate Institutional Weekly Report: BTC Funding Rate Turns Positive, CEX TradFi Trading Volume Soars (March 23, 2026 - March 29, 2026)
avatar
avatarOdaily星球日报
16 hours ago
CoinGlass: 2026 Q1 Cryptocurrency Market Share Research Report
APP
Windows
Mac

X

Telegram

Facebook

Reddit

CopyLink