PulseAugur
EN
LIVE 11:29:39

Huawei open-sources openPangu-2.0 training code for Ascend stack

By PulseAugur Editorial · [1 sources] ·

Huawei has released the training code for its openPangu-2.0 model, which is designed for its Ascend-native Mixture-of-Experts (MoE) stack. This release includes code for pre-training, supervised fine-tuning (SFT), and reinforcement learning (RL) from human feedback. The open-sourced code complements previous releases of the model's weights, specifically the Pro (505B) and Flash (92B) versions. AI

IMPACT This release provides researchers and developers with the tools to train and fine-tune large-scale MoE models on Huawei's Ascend hardware, potentially fostering innovation in AI development.

RANK_REASON Open-source release of a large language model's training code. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Pandaily →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

Huawei open-sources openPangu-2.0 training code for Ascend stack

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Open-source release of a large language model's training code. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
model release, infra
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
2 days old
Coverage has settled into its steady-state source set.

Full methodology in our editorial standards.

COVERAGE [1]

  1. Pandaily TIER_1 English(EN) · [email protected] (Pandaily) ·

    Huawei Open-Sources openPangu-2.0 Pretrain, SFT and RL Training Code on Ascend

    On September 28, 2026 Huawei open-sourced openPangu-2.0 pretrain, SFT and post-training RL code for its Ascend-native MoE stack, complementing earlier Pro (505B) and Flash (92B) weight releases.

RELATED ENTITIES

RELATED TOPICS