qwen-1.5

3rdparty/backend-r22.12/examples/backends/recommended/cmake/TutorialRecommendedBackendConfig.cmake.in

+# Copyright 2021, 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.
+
+include(CMakeFindDependencyMacro)
+
+get_filename_component(
+  TUTORIALRECOMMENDEDBACKEND_CMAKE_DIR "${CMAKE_CURRENT_LIST_FILE}" PATH
+)
+
+list(APPEND CMAKE_MODULE_PATH ${TUTORIALRECOMMENDEDBACKEND_CMAKE_DIR})
+
+if(NOT TARGET TutorialRecommendedBackend::triton-recommended-backend)
+  include("${TUTORIALRECOMMENDEDBACKEND_CMAKE_DIR}/TutorialRecommendedBackendTargets.cmake")
+endif()
+
+set(TUTORIALRECOMMENDEDBACKEND_LIBRARIES TutorialRecommendedBackend::triton-recommended-backend)

3rdparty/backend-r22.12/examples/backends/recommended/src/libtriton_recommended.ldscript

+# Copyright 2021, 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.
+{
+  global:
+    TRITONBACKEND_*;
+  local: *;
+};

3rdparty/backend-r22.12/examples/clients/bls_client

+#!/usr/bin/python
+# Copyright 2022, 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.
+
+import sys
+import argparse
+import numpy as np
+import tritonhttpclient as httpclient
+from tritonclientutils import np_to_triton_dtype
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('-u',
+                        '--url',
+                        type=str,
+                        required=False,
+                        default='localhost:8000',
+                        help='Inference server URL. Default is localhost:8000.')
+    FLAGS = parser.parse_args()
+
+    model_name = "bls_fp32"
+    shape = [16]
+    with httpclient.InferenceServerClient(url=FLAGS.url) as client:
+        input0_data = np.random.rand(*shape).astype(np.float32)
+        input1_data = np.random.rand(*shape).astype(np.float32)
+        inputs = [
+            httpclient.InferInput("INPUT0", input0_data.shape,
+                                    np_to_triton_dtype(input0_data.dtype)),
+            httpclient.InferInput("INPUT1", input1_data.shape,
+                                    np_to_triton_dtype(input1_data.dtype)),
+        ]
+
+        inputs[0].set_data_from_numpy(input0_data)
+        inputs[1].set_data_from_numpy(input1_data)
+
+        outputs = [
+            httpclient.InferRequestedOutput("OUTPUT0"),
+            httpclient.InferRequestedOutput("OUTPUT1"),
+        ]
+        response = client.infer(model_name,
+                                inputs,
+                                request_id=str(1),
+                                outputs=outputs)
+
+        result = response.get_response()
+        output0_data = response.as_numpy("OUTPUT0")
+        output1_data = response.as_numpy("OUTPUT1")
+
+        print("INPUT0 ({}) + INPUT1 ({}) = OUTPUT0 ({})".format(
+            input0_data, input1_data, output0_data))
+        print("INPUT0 ({}) - INPUT1 ({}) = OUTPUT1 ({})".format(
+            input0_data, input1_data, output1_data))
+
+        if not np.allclose(input0_data + input1_data, output0_data):
+            print("error: incorrect sum")
+            sys.exit(1)
+
+        if not np.allclose(input0_data - input1_data, output1_data):
+            print("error: incorrect difference")
+            sys.exit(1)
+
+    print('\nPASS')
+    sys.exit(0)

3rdparty/backend-r22.12/examples/clients/minimal_client

+#!/usr/bin/env python
+# Copyright 2021, 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.
+
+import argparse
+import numpy as np
+
+import tritonclient.http as httpclient
+from tritonclient.utils import InferenceServerException
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('-u',
+                        '--url',
+                        type=str,
+                        required=False,
+                        default='localhost:8000',
+                        help='Inference server URL. Default is localhost:8000.')
+    FLAGS = parser.parse_args()
+
+    # For the HTTP client, need to specify large enough concurrency to
+    # issue all the inference requests to the server in parallel. For
+    # this example we want to be able to send 2 requests concurrently.
+    try:
+        concurrent_request_count = 2
+        triton_client = httpclient.InferenceServerClient(
+            url=FLAGS.url, concurrency=concurrent_request_count)
+    except Exception as e:
+        print("channel creation failed: " + str(e))
+        sys.exit(1)
+
+    # First send a single request to the nonbatching model.
+    print('=========')
+    input0_data = np.array([ 1, 2, 3, 4 ], dtype=np.int32)
+    print('Sending request to nonbatching model: IN0 = {}'.format(input0_data))
+
+    inputs = [ httpclient.InferInput('IN0', [4], "INT32") ]
+    inputs[0].set_data_from_numpy(input0_data)
+    result = triton_client.infer('nonbatching', inputs)
+
+    print('Response: {}'.format(result.get_response()))
+    print('OUT0 = {}'.format(result.as_numpy('OUT0')))
+
+    # Send 2 requests to the batching model. Because these are sent
+    # asynchronously and Triton's dynamic batcher is configured to
+    # delay up to 5 seconds when forming a batch for this model, we
+    # expect these 2 requests to be batched within Triton and sent to
+    # the minimal backend as a single batch.
+    print('\n=========')
+    async_requests = []
+
+    input0_data = np.array([[ 10, 11, 12, 13 ]], dtype=np.int32)
+    print('Sending request to batching model: IN0 = {}'.format(input0_data))
+    inputs = [ httpclient.InferInput('IN0', [1, 4], "INT32") ]
+    inputs[0].set_data_from_numpy(input0_data)
+    async_requests.append(triton_client.async_infer('batching', inputs))
+
+    input0_data = np.array([[ 20, 21, 22, 23 ]], dtype=np.int32)
+    print('Sending request to batching model: IN0 = {}'.format(input0_data))
+    inputs = [ httpclient.InferInput('IN0', [1, 4], "INT32") ]
+    inputs[0].set_data_from_numpy(input0_data)
+    async_requests.append(triton_client.async_infer('batching', inputs))
+
+    for async_request in async_requests:
+        # Get the result from the initiated asynchronous inference
+        # request. This call will block till the server responds.
+        result = async_request.get_result()
+        print('Response: {}'.format(result.get_response()))
+        print('OUT0 = {}'.format(result.as_numpy('OUT0')))

3rdparty/backend-r22.12/examples/clients/recommended_client

+#!/usr/bin/env python
+# Copyright 2021, 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.
+
+import argparse
+import numpy as np
+
+import tritonclient.http as httpclient
+from tritonclient.utils import InferenceServerException
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+    parser.add_argument('-u',
+                        '--url',
+                        type=str,
+                        required=False,
+                        default='localhost:8000',
+                        help='Inference server URL. Default is localhost:8000.')
+    FLAGS = parser.parse_args()
+
+    # For the HTTP client, need to specify large enough concurrency to
+    # issue all the inference requests to the server in parallel. For
+    # this example we want to be able to send 2 requests concurrently.
+    try:
+        concurrent_request_count = 2
+        triton_client = httpclient.InferenceServerClient(
+            url=FLAGS.url, concurrency=concurrent_request_count)
+    except Exception as e:
+        print("channel creation failed: " + str(e))
+        sys.exit(1)
+
+    # Send 2 requests to the batching model. Because these are sent
+    # asynchronously and Triton's dynamic batcher is configured to
+    # delay up to 5 seconds when forming a batch for this model, we
+    # expect these 2 requests to be batched within Triton and sent to
+    # the backend as a single batch.
+    #
+    # The recommended backend can handle any model with 1 input and 1
+    # output as long as the input and output datatype and shape are
+    # the same. The batching model uses datatype FP32 and shape
+    # [ 4, 4 ].
+    print('\n=========')
+    async_requests = []
+
+    input0_data = np.array([[[ 1.0, 1.1, 1.2, 1.3 ],
+                             [ 2.0, 2.1, 2.2, 2.3 ],
+                             [ 3.0, 3.1, 3.2, 3.3 ],
+                             [ 4.0, 4.1, 4.2, 4.3 ]]], dtype=np.float32)
+    print('Sending request to batching model: input = {}'.format(input0_data))
+    inputs = [ httpclient.InferInput('INPUT', [1, 4, 4], "FP32") ]
+    inputs[0].set_data_from_numpy(input0_data)
+    async_requests.append(triton_client.async_infer('batching', inputs))
+
+    input0_data = np.array([[[ 10.0, 10.1, 10.2, 10.3 ],
+                             [ 20.0, 20.1, 20.2, 20.3 ],
+                             [ 30.0, 30.1, 30.2, 30.3 ],
+                             [ 40.0, 40.1, 40.2, 40.3 ]]], dtype=np.float32)
+    print('Sending request to batching model: input = {}'.format(input0_data))
+    inputs = [ httpclient.InferInput('INPUT', [1, 4, 4], "FP32") ]
+    inputs[0].set_data_from_numpy(input0_data)
+    async_requests.append(triton_client.async_infer('batching', inputs))
+
+    for async_request in async_requests:
+        # Get the result from the initiated asynchronous inference
+        # request. This call will block till the server responds.
+        result = async_request.get_result()
+        print('Response: {}'.format(result.get_response()))
+        print('OUTPUT = {}'.format(result.as_numpy('OUTPUT')))

3rdparty/backend-r22.12/examples/model_repos/bls_models/addsub_python/config.pbtxt

+# Copyright 2022, 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: "addsub_python"
+backend: "python"
+max_batch_size: 0
+
+input [
+  {
+    name: "INPUT0"
+    data_type: TYPE_FP32
+    dims: [ 16 ]
+  }
+]
+input [
+  {
+    name: "INPUT1"
+    data_type: TYPE_FP32
+    dims: [ 16 ]
+  }
+]
+output [
+  {
+    name: "OUTPUT0"
+    data_type: TYPE_FP32
+    dims: [ 16 ]
+  }
+]
+output [
+  {
+    name: "OUTPUT1"
+    data_type: TYPE_FP32
+    dims: [ 16 ]
+  }
+]