Fix some of the errors I've encountered while following tutorials

Conceptual_Guide/Part_1-model_deployment/client.py

@@ -207,6 +207,6 @@ def recognition_postprocessing(scores: np.ndarray) -> str:
     )
 
     # Process response from recognition model
-    final_text = recognition_postprocessing(recognition_response.as_numpy("308"))
+    final_text = recognition_postprocessing(recognition_response.as_numpy("307"))
 
     print(final_text)

Conceptual_Guide/Part_1-model_deployment/model_repository/text_recognition/config.pbtxt

@@ -36,7 +36,7 @@ input [
 ]
 output [
   {
-    name: "308"
+    name: "307"
     data_type: TYPE_FP32
     dims: [ 1, 26, 37 ]
   }

Conceptual_Guide/Part_2-improving_resource_utilization/README.md

@@ -110,7 +110,7 @@ model.load_state_dict(state)
 
 # Create ONNX file by tracing model
 trace_input = torch.randn(1, 1, 32, 100)
-torch.onnx.export(model, trace_input, "str.onnx", verbose=True, dynamic_axes={'input.1':[0],'308':[0]})
+torch.onnx.export(model, trace_input, "str.onnx", verbose=True, dynamic_axes={'input.1':[0],'307':[0]})
 ```
 
 ### Launching the server
@@ -231,7 +231,7 @@ Request concurrency: 16
 ```
 As each of the requests had a batch size (of 2), while the maximum batch size of the model was 8, dynamically batching these requests resulted in considerably improved throughput. Another consequence is a reduction in the latency. This reduction can be primarily attributed to reduced wait time in queue wait time. As the requests are batched together, multiple requests can be processed in parallel.
 
-* **Dynamic Batching with multiple model instances**: To set up the Triton Server in this configuration, add `instance_group` in `config.pbtxt` and make sure to include `--gpus=1` and make sure to include `--gpus=1` in the `docker run` command to set up the server. Include `dynamic_batching` per instructions of the previous section in the model configuration. A point to note is that peak GPU utilization on the GPU shot up to 74% (A100 in this case) while just using a single model instance with dynamic batching. Adding one more instance will definitely improve performance but linear perf scaling will not be achieved in this case.
+* **Dynamic Batching with multiple model instances**: To set up the Triton Server in this configuration, add `instance_group` in `config.pbtxt` and make sure to include `--gpus=1` in the `docker run` command to set up the server. Include `dynamic_batching` per instructions of the previous section in the model configuration. A point to note is that peak GPU utilization on the GPU shot up to 74% (A100 in this case) while just using a single model instance with dynamic batching. Adding one more instance will definitely improve performance but linear perf scaling will not be achieved in this case.
 
 ```
 # Query

Conceptual_Guide/Part_2-improving_resource_utilization/model_repository/text_recognition/config.pbtxt

@@ -36,7 +36,7 @@ input [
 ]
 output [
   {
-    name: "308"
+    name: "307"
     data_type: TYPE_FP32
     dims: [ 26, 37 ]
   }

Conceptual_Guide/Part_5-Model_Ensembles/README.md

@@ -316,10 +316,10 @@ ensemble_scheduling {
 We'll again be launching Triton using docker containers. This time, we'll start an interactive session within the container instead of directly launching the triton server.
 
 ```bash
-docker run --gpus=all -it --shm-size=256m --rm  \
+docker run --gpus=all -it --shm-size=512m --rm  \
   -p8000:8000 -p8001:8001 -p8002:8002 \
   -v ${PWD}:/workspace/ -v ${PWD}/model_repository:/models \
-  nvcr.io/nvidia/tritonserver:22.12-py3
+  nvcr.io/nvidia/tritonserver:yy.mm-py3
 ```
 
 We'll need to install a couple of dependencies for our Python backend scripts.

Conceptual_Guide/Part_5-Model_Ensembles/model_repository/ensemble_model/config.pbtxt

@@ -100,7 +100,7 @@ ensemble_scheduling {
         value: "cropped_images"
       }
       output_map {
-        key: "308"
+        key: "307"
         value: "recognition_output"
       }
     },

Conceptual_Guide/Part_5-Model_Ensembles/model_repository/text_recognition/config.pbtxt

@@ -36,7 +36,7 @@ input [
 ]
 output [
   {
-    name: "308"
+    name: "307"
     data_type: TYPE_FP32
     dims: [ 26, 37 ]
   }

Conceptual_Guide/Part_5-Model_Ensembles/utils/export_text_recognition.py

@@ -45,5 +45,5 @@
     trace_input,
     model_directory / "model.onnx",
     verbose=True,
-    dynamic_axes={"input.1": [0], "308": [0]},
+    dynamic_axes={"input.1": [0], "307": [0]},
 )

HuggingFace/ensemble_model_repository/ensemble_model/config.pbtxt

@@ -40,11 +40,6 @@ output [
         name: "last_hidden_state"
         data_type: TYPE_FP32
         dims: [-1, -1]
-    },
-    {
-        name: "1519"
-        data_type: TYPE_FP32
-        dims: [768]
     }
 ]
 ensemble_scheduling {
@@ -72,10 +67,6 @@ ensemble_scheduling {
                 key: "last_hidden_state"
                 value: "last_hidden_state"
             }
-            output_map {
-                key: "1519"
-                value: "1519"
-            }
         }
     ]
 }

HuggingFace/python_model_repository/python_vit/1/model.py

@@ -30,22 +30,22 @@
 
 class TritonPythonModel:
     def initialize(self, args):
-        self.feature_extractor = ViTFeatureExtractor.from_pretrained('google/vit-base-patch16-224-in21k').to("cuda")
+        self.feature_extractor = ViTFeatureExtractor.from_pretrained('google/vit-base-patch16-224-in21k')#.to("cuda")
         self.model = ViTModel.from_pretrained("google/vit-base-patch16-224-in21k").to("cuda")
 
     def execute(self, requests):
         responses = []
         for request in requests:
             inp = pb_utils.get_input_tensor_by_name(request, "image")
             input_image = np.squeeze(inp.as_numpy()).transpose((2,0,1))
-            inputs = self.feature_extractor(images=input_image, return_tensors="pt")
+            inputs = self.feature_extractor(images=input_image, return_tensors="pt").to("cuda")
 
             outputs = self.model(**inputs)
 
             inference_response = pb_utils.InferenceResponse(output_tensors=[
                 pb_utils.Tensor(
-                    "label",
-                    outputs.last_hidden_state.numpy()
+                    "last_hidden_state",
+                    outputs.last_hidden_state.detach().cpu().numpy()
                 )
             ])
             responses.append(inference_response)

Quick_Deploy/ONNX/client.py

@@ -39,7 +39,7 @@ def rn50_preprocess(img_path="img1.jpg"):
         transforms.ToTensor(),
         transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
     ])
-    return np.expand_dims(preprocess(img).numpy(),axis=0)
+    return np.expand_dims(preprocess(img).numpy(), axis=0).squeeze()
 
 transformed_img = rn50_preprocess()
 

Quick_Deploy/ONNX/model_repository/densenet_onnx/config.pbtxt

@@ -0,0 +1,44 @@
+# Copyright 2023, 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
+# are met:
+#  * Redistributions of source code must retain the above copyright
+#    notice, this list of conditions and the following disclaimer.
+#  * Redistributions in binary form must reproduce the above copyright
+#    notice, this list of conditions and the following disclaimer in the
+#    documentation and/or other materials provided with the distribution.
+#  * Neither the name of NVIDIA CORPORATION nor the names of its
+#    contributors may be used to endorse or promote products derived
+#    from this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
+# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+# PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT OWNER OR
+# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
+# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
+# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
+# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
+# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+name: "densenet_onnx"
+platform: "onnxruntime_onnx"
+max_batch_size : 0
+input [
+  {
+    name: "data_0"
+    data_type: TYPE_FP32
+    dims: [ 3, 224, 224 ]
+    reshape { shape: [ 1, 3, 224, 224 ] }
+  }
+]
+output [
+  {
+    name: "fc6_1"
+    data_type: TYPE_FP32
+    dims: [ 1, 1000 ,1, 1]
+  }
+]