specific method · filed under data curation
Three-stage pre-training curriculum
Structures pre-training into encoder-alignment, general, and long-context stages.
Also called three distinct stages.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 1
Documented in
Evidence
1 span quoted from the sources, strongest treatment first.
The pre-training of Qwen3.5-Omni is structured into three distinct stages.
usedtraining objectivein Qwen3.5-OmniQwen
Filed alongside
Other methods under data curation :: data mixture & curriculum.
AutoMixerDeterministic pre-splitting of ultra-long documentsVulnerability-discovery data inclusionAgent-centric multimodal data mixtureAttribution, hedging, and refusal data subsetsBlind pairwise bucket-boundary calibrationData SchedulerDomain-specific multimodal datasetsDynamic sampling for mixed-task reinforcement learningFiltering, deduplication, and difficulty calibrationGaussian-based data mixture and curriculum constructionInterleaved dataInterleaved multimodal trainingKL-regularized data-mixture optimizationLong-context data upsamplingPass-rate-based RL task samplingPrior-constrained Dirichlet mixture explorationProtocolQA-aligned RL training datasetsSample MixerScaling-ladder-guided data and model planningSmooth weighted round-robin source schedulingText-only pre-training corpusThree-stage data mixture strategyTwo-phase curriculum