* Shift up fluid models * Update nlp submodules * Add redirect guide in org readme * Move HiNAS_models to AutoDL