Buy Crypto
Pay with
USD
USD
Buy & Sell
Hot
Supports Visa, Mastercard, SEPA & more
P2P
0 Fees
Flexible trading, zero fees
Gate Card
Use your crypto for payments worldwide
Markets
Trade
Basic
Advanced
Futures
Futures
Access hundreds of perpetual contracts
CFD
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Earn
Launch
CandyDrop
tag
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Pre-IPOs
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Investment
Square
Square
Post, share, and explore crypto trends
Live
moments live
Live crypto market analysis
Chat
Chat with crypto traders
News
What is happening in crypto
flower_head
More
Promotions
AI
Others
TradFi
Gate Card
Rewards
Square
Latest
Hot
News
Profile

SenseNova-U1 full training code open source, supporting multimodal multi-task training

MeNews
robot
Abstract generation in progress
ME News Report, May 26 (UTC+8), OpenSenseNova has open-sourced the complete training codebase for SenseNova-U1, supporting its 8B dense model and A3B MoE architecture. The codebase uses a unified framework that can train multiple multimodal tasks simultaneously, including text-to-image generation, image editing, interleaved generation, and text and visual understanding. Engineered for large-scale training, it supports mixed parallelism, streaming recoverable data pipelines, environment variable-driven configuration, and scalability from 1×8 GPUs to multi-node clusters. The code has been open-sourced on GitHub under the Apache-2.0 license. (Source: AiHot)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • 5
  • 1
  • Share
Comment
Add a comment
Add a comment
NeonMeltsIceCream
· 5h ago
Hybrid parallelism + streaming recovery, large-scale training finally no longer fears checkpoints
View OriginalReply0
StardustRouter
· 5h ago
Environment variable-driven configuration, CI/CD integration should be much smoother.
View OriginalReply0
AuroraStone
· 5h ago
Comprehensive package for text understanding, image generation, and editing—one framework to handle the multimodal all-in-one suite
View OriginalReply0
GateUser-46c777d0
· 5h ago
Expanding from a 1×8 card to multiple nodes, this flexible design is very friendly for small and medium teams.
View OriginalReply0
NftsOutsideTheTidalLine
· 5h ago
Apache-2.0, a genuinely generous release — 8B dense + A3B MoE dual architecture for maximum flexibility
View OriginalReply0

  • Pinned

Sitemap