* Paddle-Trt: Replace fc mul matmul matmul_v2 with matrix_multiply

* Support two inputs of multihead attention named qk_multihead

* split pass

* fix compile

* fix ut

* more time

* modify ut

* reduce dim

* fix compile

* reshape weight

* tensor

* remove enforce

* static shape ut

* batchsize

* reorder pass

* minus test cases

* windows timeout

* windows time out

* remove test for windows

* correct

* sssss

* xxx

* mkldnn->onednn

* fused softplus op + kernel

* remove extra attributes

* add missing handler

* change var name

[XPU] add multi_encoder_xpu_slice_fuse_pass, generate_sequence_xpu_fuse_pass, generate_sequence_xpu kernel (#50570)

* eleadd_trans first version

log fix

* refine code for linear format, add pass check

* linear format refine and ut fix

* fix ut

* windows ut

* windows ut 2

* move tensorMeta and alloc to configure

* trans_layernorm

* QuantTranpose pattern is being found by pass

* quant + transpose fuse

* code style changes

* UT written, reorder fixed

* Dequantize + transpose2 fuse  added

* pass name changed

* UT added & shift corrected

* got rid of redundancy

* review changes

* AsIntermediate corrected

* compat added

* disable conv2d_fusion_layout_transfer_pass temporarily

* disable conv2d_fusion_layout_transfer_pass temporarily

* gn_silu

* add ut

* set TIMEOUT

* correct comments

* comments

* disable windows ut

* rename parameter

* add trt_support_nhwc_pass

* add gpu_cpu_map_matmul_to_mul_pass to kGpuLowerPrecisionPasses

* disable fc_elementwise_layernorm_fuse_pass in mixed precision

* add fmha_flashattention oss plugin

* skip_groupnorm

* init

* preln

* add ut

* more assert

* set timeout

* fix windows ci issue

* Implement conv2d_fusion NHWC format using CUTLASS
* Add unit testing for CUTLASS Conv in inference
* Add experimental API for CUTLASS.