enhance benchmark (#604)

Signed-off-by: Xin He <[email protected]>

Author: xin3he

neural_compressor/benchmark.py

@@ -37,6 +37,7 @@
 from .utils import logger
 from .conf.pythonic_config import Config
 from .config import BenchmarkConfig
+from .utils.utility import Statistics
 
 
 def set_env_var(env_var, value, overwrite_existing=False):
@@ -162,13 +163,11 @@ def __init__(self, conf):
         if self.conf.usr_cfg.model.framework != 'NA':
             self.framework = self.conf.usr_cfg.model.framework.lower()
 
-    def __call__(self):
+    def __call__(self, raw_cmd=None):
         """Directly call a Benchmark object.
 
         Args:
-            model: Get the model
-            b_dataloader: Set dataloader for benchmarking
-            b_func: Eval function for benchmark
+            raw_cmd: raw command used for benchmark
         """
         cfg = self.conf.usr_cfg
         assert cfg.evaluation is not None, 'benchmark evaluation filed should not be None...'
@@ -181,7 +180,9 @@ def __call__(self):
         logger.info("Start to run Benchmark.")
         if os.environ.get('NC_ENV_CONF') == 'True':
             return self.run_instance()
-        self.config_instance()
+        if raw_cmd is None:
+            raw_cmd = sys.executable + ' ' + ' '.join(sys.argv)
+        self.config_instance(raw_cmd)
         self.summary_benchmark()
         return None
 
@@ -204,16 +205,26 @@ def summary_benchmark(self):
                         throughput_l.append(float(throughput.group(1))) if throughput and throughput.group(1) else None
             assert len(latency_l)==len(throughput_l)==num_of_instance, \
                 "Multiple instance benchmark failed with some instance!"
-            logger.info("\n\nMultiple instance benchmark summary: ")
-            logger.info("Latency average: {:.3f} ms".format(sum(latency_l)/len(latency_l)))
-            logger.info("Throughput sum: {:.3f} images/sec".format(sum(throughput_l)))
+
+            output_data = [
+                ["Latency average [second/sample]", "{:.3f}".format(sum(latency_l)/len(latency_l))],
+                ["Throughput sum [samples/second]", "{:.3f}".format(sum(throughput_l))]
+            ]
+            logger.info("********************************************")
+            Statistics(
+                output_data, 
+                header='Multiple Instance Benchmark Summary',
+                field_names=["Items", "Result"]).print_stat()
         else:
             # (TODO) should add summary after win32 benchmark has log
             pass
 
-    def config_instance(self):
-        """Configure the multi-instance commands and trigger benchmark with sub process."""
-        raw_cmd = sys.executable + ' ' + ' '.join(sys.argv)
+    def config_instance(self, raw_cmd):
+        """Configure the multi-instance commands and trigger benchmark with sub process.
+
+        Args:
+            raw_cmd: raw command used for benchmark
+        """
         multi_instance_cmd = ''
         num_of_instance = int(os.environ.get('NUM_OF_INSTANCE'))
         cores_per_instance = int(os.environ.get('CORES_PER_INSTANCE'))