set flag to accelerate inference in paddle 3.0b2 (#2419)

paddlex/inference/utils/pp_option.py

@@ -12,6 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import os
+
 from ...utils.device import parse_device, set_env_for_device, get_default_device
 from ...utils import logging
 from .new_ir_blacklist import NEWIR_BLOCKLIST
@@ -122,6 +124,9 @@ class PaddlePredictorOption(object):
         if device_type not in ("cpu"):
             if device_ids is None or len(device_ids) > 1:
                 logging.debug(f"The device ID has been set to {device_id}.")
+        # XXX(gaotingquan): set flag to accelerate inference in paddle 3.0b2
+        if device_type in ("gpu", "cpu"):
+            os.environ["FLAGS_enable_pir_api"] = "1"
 
     @property
     def min_subgraph_size(self):