* fix cast bug

* Move ITensor construction for Weight (persistable variable) from OpConvert to TensorRTEngine.

* Support matmul_v2 in Paddle-TensorRT converter.

* Support dynamic shape in multiclass_nms3 Plugin for Paddle-TensorRT.

* cherry pick delay tensorrt log
* Update trt_plugin.h

* [OpAttr]Refine Teller logic if encounter OpDesc with Variable type Attribute

* fix iterator

* fix typo

* fix lambda expr

* fix ptr

* first commit

* conver done

* correct format

* layernorm_shift_partition

* correct convert

* redefine plugin

* runable

* bug fix

* modify ShiftPartitionPattern

* correct

* add UT

* modify ut

* compile

* modify enforce

* modify UT

* add silu converter

* remove int8 fallback

* vit_384_opt

* just support trt8

* padding + unpadding

* fix:unit test

* refactor:padding

* fix: change the position of round_up

* refactor: delete workspace

* fix bug fix

* add shape size check

* polish code

* multi -1 shape fix

* code style improve

* bug fix

* code style fix

* fix bug fix

* fix layernormTrt meanVar alloc bug

* add group_norm dyanmic plugin

* convert multihead to oss

* fix:bug

* fix:delete const cast

* fix:don't support bias_qk

* add vit pass

* fix:convert bug and add preln_residual_bias

* support length=-1

* add UT for convert

* add no_bias_qk support for gpu_multihead_op

* delete infer_shape depends on bias_qk

* oss just can be used in T4 and A*

* fix:change api for ROCM CI

* trt engine input data type should be consistent with trt input bindings type

* fix some bugs

* fix some bugs

* fix some bugs

* disable_skip_layernorm_fp16

* clean tensor.h

* fix gather_nd

* update trt workspace size param

* update

* update

* update

* use int64_t

* use int64_t

* upate

* update

* add rnn

* remove cudaDeviceContext

* remove more template

* fix rocm compile

* remove alias name CUDADeviceContext

* fix compile

* fix tests

* revert changes

* add varlen_token_prune plugin, pass, convert

* skip cast trt convert when input dtype is bool