Minor refactor and cleanup

triton-inference-server · Oct 18, 2023 · fdd4d58 · fdd4d58
1 parent bca684c
commit fdd4d58
Show file tree

Hide file tree

Showing 5 changed files with 13 additions and 4 deletions.
diff --git a/input-data-10.json b/input-data-10.json
@@ -0,0 +1 @@
+{"data": [{"PROMPT": ["Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello"], "STREAM": [true]}]}
diff --git a/input-data-100.json b/input-data-100.json
@@ -0,0 +1 @@
+{"data": [{"PROMPT": ["Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello"], "STREAM": [true]}]}
diff --git a/input-data-50.json b/input-data-50.json
@@ -0,0 +1 @@
+{"data": [{"PROMPT": ["Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello"], "STREAM": [true]}]}
diff --git a/model_analyzer/config/generate/perf_analyzer_config_generator.py b/model_analyzer/config/generate/perf_analyzer_config_generator.py
@@ -379,15 +379,22 @@ def _update_perf_config_based_on_parameter_combination(
     ) -> None:
         if "request-parameter" in parameter_combination:
             request_parameter = parameter_combination["request-parameter"]
-            max_token_start = request_parameter.find(":")
-            max_token_stop = request_parameter.find(":", max_token_start + 1)
-            max_token = int(request_parameter[max_token_start + 1 : max_token_stop])
+            max_token = self._extract_max_token_from_request_parameter(
+                request_parameter
+            )
             parameter_combination["request-period"] = (
                 max_token if max_token < 10 else 10
             )
 
         perf_config.update_config(parameter_combination)
 
+    def _extract_max_token_from_request_parameter(self, request_parameter: str) -> int:
+        max_token_start = request_parameter.find(":")
+        max_token_stop = request_parameter.find(":", max_token_start + 1)
+        max_token = int(request_parameter[max_token_start + 1 : max_token_stop])
+
+        return max_token
+
     def _update_perf_config_based_on_perf_analyzer_flags(
         self, perf_config: PerfAnalyzerConfig
     ) -> None:

diff --git a/tests/test_perf_analyzer_config_generator.py b/tests/test_perf_analyzer_config_generator.py
@@ -1007,7 +1007,6 @@ def _run_and_test_perf_analyzer_config_generator(
             self.assertEqual(
                 expected_configs[i]._options, perf_analyzer_configs[i]._options
             )
-            self.maxDiff = None
             self.assertEqual(expected_configs[i]._args, perf_analyzer_configs[i]._args)
             self.assertEqual(
                 expected_configs[i]._additive_args,
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		{"data": [{"PROMPT": ["Hello Hello Hello Hello Hello Hello Hello Hello Hello Hello"], "STREAM": [true]}]}