BestHub
Sign in
Tagged articles
8 articles
Page 1 of 1
Baidu Geek Talk
Baidu Geek Talk
May 25, 2026 · Artificial Intelligence

Accelerating Multimodal Model Training: LoongForge's DP Load‑Balancing Optimization Explained

The article analyzes how data‑parallel (DP) load imbalance hampers large‑scale multimodal model training, details LoongForge's two‑stage adaptive data‑reallocation method that builds a precise compute‑cost model and dynamically redistributes samples, and presents experimental results showing up to 10% throughput gains on massive DP clusters.

DP load balancingData ParallelDistributed Training
0 likes · 16 min read
Accelerating Multimodal Model Training: LoongForge's DP Load‑Balancing Optimization Explained
Baidu Geek Talk
Baidu Geek Talk
May 13, 2026 · Artificial Intelligence

LoongForge Boosts Multimodal Training Speed by 45% on GPU and Kunlun XPU

LoongForge, Baidu Baige’s open‑source full‑modal training framework, unifies LLM, VLM and VLA workloads, runs unchanged on NVIDIA GPUs and Kunlun XPU, and delivers 15‑45% end‑to‑end speedups with up to 90% linear scaling on 5,000‑plus card clusters, while simplifying model integration via YAML.

AI InfrastructureGPUKunlun XPU
0 likes · 23 min read
LoongForge Boosts Multimodal Training Speed by 45% on GPU and Kunlun XPU
Baidu Intelligent Cloud Tech Hub
Baidu Intelligent Cloud Tech Hub
Apr 24, 2026 · Artificial Intelligence

LoongForge: Open‑Source Multimodal Training Framework Runs on GPU and Kunlun XPU with 45% Speedup

LoongForge is an open‑source, Megatron‑based multimodal training framework that unifies LLM, VLM, VLA and diffusion models, runs seamlessly on NVIDIA GPUs and Baidu Kunlun XPU, and delivers 15%‑45% end‑to‑end training acceleration while scaling linearly on thousands of cards.

GPUKunlun XPULoongForge
0 likes · 23 min read
LoongForge: Open‑Source Multimodal Training Framework Runs on GPU and Kunlun XPU with 45% Speedup
Baidu Intelligent Cloud Tech Hub
Baidu Intelligent Cloud Tech Hub
Nov 20, 2025 · Artificial Intelligence

Boost Multimodal Model Training Efficiency with Offline Sequence Packing and Mixed‑Modality Data

Baidu's Baige team introduces an extended multimodal data loader, automated ShareGPT format conversion, and offline sequence packing techniques that together double token throughput, cut SFT training time by up to six times, and improve GPU utilization and stability for large vision‑language models.

AI InfrastructureAIAKGPU efficiency
0 likes · 7 min read
Boost Multimodal Model Training Efficiency with Offline Sequence Packing and Mixed‑Modality Data
Baidu Intelligent Cloud Tech Hub
Baidu Intelligent Cloud Tech Hub
Nov 4, 2025 · Artificial Intelligence

How Baidu’s Baige Accelerates Multimodal Video Training with Context Parallelism

Baidu Baige’s enhanced veRL framework dramatically boosts video frame rates and resolution limits, cuts training time, reduces memory usage, and improves model accuracy by leveraging context parallelism and optimized attention on Ampere GPUs for multimodal mixed‑training scenarios.

AI accelerationContext ParallelismVideo Processing
0 likes · 6 min read
How Baidu’s Baige Accelerates Multimodal Video Training with Context Parallelism
HyperAI Super Neural
HyperAI Super Neural
Oct 27, 2025 · Artificial Intelligence

MIT’s Open‑Source BoltzGen Achieves nM‑Level Affinity for 66% of Targets Across Molecular Types

BoltzGen, an all‑atom generative model released by MIT and collaborators, unifies protein folding and binder design with a geometric continuous representation and a flexible design language, training on multimodal datasets and demonstrating nM‑level affinity for 66% of 26 diverse targets including proteins, nanobodies, peptides and small molecules.

BoltzGenGenerative AIdiffusion model
0 likes · 12 min read
MIT’s Open‑Source BoltzGen Achieves nM‑Level Affinity for 66% of Targets Across Molecular Types
Amap Tech
Amap Tech
Oct 3, 2025 · Artificial Intelligence

How OmniNav Unifies Multi‑Task Embodied Navigation with a Fast‑Slow Dual System

OmniNav introduces a unified framework for embodied navigation that simultaneously handles instruction‑goal, object‑goal, point‑goal, and frontier‑based exploration tasks using a fast visual‑language‑driven policy and a slow memory‑augmented planner, achieving state‑of‑the‑art performance and real‑world 5 Hz deployment.

continuous controlembodied navigationfast-slow architecture
0 likes · 9 min read
How OmniNav Unifies Multi‑Task Embodied Navigation with a Fast‑Slow Dual System