Merge pull request !26235 from danishfarid/autotune

Merge pull request !24080 from Rainfor/compute_detlas

Merge pull request !21648 from 杨旭华/YesNoDataset

Merge pull request !26234 from markuskunej/offload_config

slice recompute activation

Merge pull request !26494 from Xiaoda/102-adjusting-sharding-propagation

Merge pull request !24547 from yangwm/autoaugment

Merge pull request !26126 from cathwong/ckw_ut_prof_startstop

1) when propagate sharding strategy from one op to another, try to find the strategy with zero communication cost;
2) if there is no such strategy, find the strategy with minimum communication cost, and raise a warning;

Merge pull request !26350 from yangzhenzhang/add-check-for-resize-op

Merge pull request !26453 from Margaret_wangrui/fallback_tests

Merge pull request !26345 from chenzhuo/jvp

Merge pull request !24519 from Rainfor/SlidingWindowCmn

Merge pull request !26279 from TuDouNi/neighborexchangev2_bug

Merge pull request !20576 from 杨旭华/LJSpeechDataset

Merge pull request !25488 from wenkai/wk1027setdump1

Merge pull request !26236 from Mohammad Motallebi/add_profiler_save

Merge pull request !25909 from huangbingjian/update_global_params

Merge pull request !25703 from 孟权令/LpNorm

Merge pull request !22193 from 杨旭华/DBpedia

Merge pull request !25918 from chenzhuo/jvp

Merge pull request !26135 from LiangZhibo/broaden3

Merge pull request !26036 from yangzhenzhang/add-output-strategy-for-op-init

Merge pull request !26049 from Xiaoda/99-ad-reshape-adjusting-in-sharding-propagation

Merge pull request !26131 from alashkari/hw_accel_move_copy

Merge pull request !19986 from 杨旭华/AGNEWS

1) using 'swc index of strategy_cost_' as reshape's selected strategy;
2) when encountering reshape in BFS, select the 'swc index' with zero communication cost;
3) when encountering a reshape that is already visited, check whether there exists communication between reshape and current operator. It is OK if communication happens between two configured operators;
4) currently, two consecutive reshapes are not supported;
5) adjusting BFS structure in graph_costmodel.cc;
6) adjusting some code in step_auto_parallel.cc to avoid cyclomatic complexity.