set paddle-cpu backend's mkldnn=True (#2085)

* set paddle-cpu backend's mkldnn=True

* rerun CI

paddlex/utils/hpi_configs/PP-OCRv4_mobile_det.yaml

@@ -7,7 +7,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
     paddle_tensorrt:
       dynamic_shapes:
         x:

paddlex/utils/hpi_configs/PP-OCRv4_mobile_rec.yaml

@@ -5,7 +5,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
     paddle_tensorrt:
       dynamic_shapes:
         x:

paddlex/utils/hpi_configs/PP-OCRv4_server_det.yaml

@@ -7,7 +7,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
     paddle_tensorrt:
       dynamic_shapes:
         x:

paddlex/utils/hpi_configs/PP-OCRv4_server_rec.yaml

@@ -5,7 +5,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
     paddle_tensorrt:
       dynamic_shapes:
         x:

paddlex/utils/hpi_configs/ch_RepSVTR_rec.yaml

@@ -5,7 +5,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
     paddle_tensorrt:
       dynamic_shapes:
         x:

paddlex/utils/hpi_configs/ch_SVTRv2_rec.yaml

@@ -5,7 +5,6 @@ Hpi:
     paddle_infer:
       cpu_num_threads: 8
       enable_log_info: false
-      enable_mkldnn: false
   selected_backends:
     cpu: openvino
     gpu: paddle_infer