rm:TensorFlow - fix to 'L0_server_status' (#8100)

Author: mc-nv

qa/L0_server_status/server_status_test.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 
-# Copyright 2018-2024, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# Copyright 2018-2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -45,7 +45,7 @@ class ServerMetadataTest(tu.TestResultCollector):
     def test_basic(self):
         try:
             for pair in [("localhost:8000", "http"), ("localhost:8001", "grpc")]:
-                model_name = "graphdef_int32_int8_int8"
+                model_name = "libtorch_int32_int8_int8"
                 extensions = [
                     "classification",
                     "sequence",
@@ -129,7 +129,7 @@ def test_unknown_model(self):
     def test_unknown_model_version(self):
         try:
             for pair in [("localhost:8000", "http"), ("localhost:8001", "grpc")]:
-                model_name = "graphdef_int32_int8_int8"
+                model_name = "onnx_int32_int8_int8"
                 if pair[1] == "http":
                     triton_client = httpclient.InferenceServerClient(
                         url=pair[0], verbose=True
@@ -149,18 +149,18 @@ def test_unknown_model_version(self):
         except InferenceServerException as ex:
             self.assertTrue(
                 ex.message().startswith(
-                    "Request for unknown model: 'graphdef_int32_int8_int8' version 99 is not found"
+                    "Request for unknown model: 'onnx_int32_int8_int8' version 99 is not found"
                 )
             )
 
     def test_model_latest_infer(self):
         input_size = 16
         tensor_shape = (1, input_size)
-        platform_name = {"graphdef": "tensorflow_graphdef", "onnx": "onnxruntime_onnx"}
+        platform_name = {"plan": "tensorrt_plan", "onnx": "onnxruntime_onnx"}
 
         # There are 3 versions of *_int32_int32_int32 and all
         # should be available.
-        for platform in ("graphdef", "onnx"):
+        for platform in ("plan", "onnx"):
             model_name = platform + "_int32_int32_int32"
 
             # Initially there should be no version stats..
@@ -316,7 +316,7 @@ def test_model_specific_infer(self):
 
         # There are 3 versions of *_float32_float32_float32 but only
         # versions 1 and 3 should be available.
-        for platform in ("graphdef", "onnx", "plan"):
+        for platform in ("libtorch", "onnx", "plan"):
             tensor_shape = (1, input_size)
             model_name = platform + "_float32_float32_float32"
 
@@ -439,7 +439,7 @@ def test_model_versions_deleted(self):
         # version 3 was executed once. Version 2 and 3 models were
         # deleted from the model repository so now only expect version 1 to
         # be ready and show stats.
-        for platform in ("graphdef", "onnx"):
+        for platform in ("libtorch", "onnx"):
             model_name = platform + "_int32_int32_int32"
 
             try:
@@ -513,7 +513,7 @@ def test_model_versions_added(self):
         # Originally There was version 1 of *_float16_float32_float32.
         # Version 7 was added so now expect just version 7 to be ready
         # and provide infer stats.
-        for platform in ("graphdef",):
+        for platform in ("plan",):
             model_name = platform + "_float16_float32_float32"
 
             try:
@@ -615,7 +615,7 @@ def test_infer_stats_no_model_version(self):
         # version 3 was executed once. Version 2 and 3 models were
         # deleted from the model repository so now only expect version 1 to
         # be ready and show infer stats.
-        for platform in ("graphdef", "onnx"):
+        for platform in ("libtorch", "onnx"):
             model_name = platform + "_int32_int32_int32"
 
             try:
@@ -723,8 +723,8 @@ def test_infer_stats_no_model(self):
                     stats = infer_stats.model_stats
                 self.assertEqual(
                     len(stats),
-                    221,
-                    "expected 221 infer stats for all ready versions of all model",
+                    125,
+                    "expected 125 infer stats for all ready versions of all model",
                 )
 
         except InferenceServerException as ex:

qa/L0_server_status/test.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Copyright (c) 2018-2024, NVIDIA CORPORATION. All rights reserved.
+# Copyright (c) 2018-2025, NVIDIA CORPORATION. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -84,9 +84,9 @@ fi
 
 set -e
 
-rm -fr models/graphdef_int32_int32_int32/2 models/graphdef_int32_int32_int32/3
+rm -fr models/libtorch_int32_int32_int32/2 models/libtorch_int32_int32_int32/3
 rm -fr models/onnx_int32_int32_int32/2 models/onnx_int32_int32_int32/3
-cp -r models/graphdef_float16_float32_float32/1 models/graphdef_float16_float32_float32/7
+cp -r models/plan_float16_float32_float32/1 models/plan_float16_float32_float32/7
 sleep 3
 
 # Dumping the contents of the models that are currently loaded for debugging purposes