[GitHub] [incubator-tvm] trevor-m commented on a change in pull request #6395: [BYOC][TensorRT] TensorRT BYOC integration

[GitBox <gi...@apache.org>]

trevor-m commented on a change in pull request #6395:
URL: https://github.com/apache/incubator-tvm/pull/6395#discussion_r488828728



##########
File path: tests/python/contrib/test_tensorrt.py
##########
@@ -0,0 +1,573 @@
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#   http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+import numpy as np
+import time
+import pytest
+
+import tvm
+import tvm.relay.testing
+from tvm import relay
+from tvm.relay.op.contrib import tensorrt
+from tvm.contrib import graph_runtime
+
+def should_skip():
+    if not tvm.runtime.enabled("cuda") or not tvm.gpu(0).exist:
+        print("skip because cuda is not enabled.")
+        return True
+    if not tensorrt.is_tensorrt_runtime_enabled():
+        print("skip because tensorrt runtime is not available")
+        return True
+    return False
+
+def test_tensorrt_simple():
+    if should_skip():
+        return
+    dtype = 'float32'
+    xshape = (1, 3, 2, 2)
+    yshape = (1, 3,  1,  1)
+    zshape = (1,  1,  1,  1)
+    x = relay.var('x', shape=(xshape), dtype=dtype)
+    y = relay.var('y', shape=(yshape), dtype=dtype)
+    z = relay.var('z', shape=(zshape), dtype=dtype)
+    w = z * (x + y)
+    out = relay.nn.relu(w)
+    f = relay.Function([x, y, z], out)
+
+    mod = tvm.IRModule()
+    mod['main'] = f
+    mod = tensorrt.partition_for_tensorrt(mod)
+    with relay.build_config(opt_level=3):
+        graph, lib, params = relay.build(mod, "cuda")
+    mod = graph_runtime.create(graph, lib, ctx=tvm.gpu(0))
+    x_data = np.random.uniform(-1, 1, xshape).astype(dtype)
+    y_data = np.random.uniform(-1, 1, yshape).astype(dtype)
+    z_data = np.random.uniform(-1, 1, zshape).astype(dtype)
+    mod.run(x=x_data, y=y_data, z=z_data)
+    results = [mod.get_output(i).asnumpy() for i in range(mod.get_num_outputs())]
+
+def test_tensorrt_not_compatible():
+    if should_skip():
+        return
+    dtype = 'float32'
+    xshape = (1, 32, 14, 14)
+    x = relay.var('x', shape=(xshape), dtype=dtype)
+    y = relay.add(x, x)
+    z = relay.erf(y)
+    out = relay.nn.relu(z)
+    f = relay.Function([x], out)
+    mod = tvm.IRModule()
+    mod['main'] = f
+    mod = tensorrt.partition_for_tensorrt(mod)
+    with relay.build_config(opt_level=3):
+        graph, lib, params = relay.build(mod, "cuda")
+    mod = graph_runtime.create(graph, lib, ctx=tvm.gpu(0))
+    x_data = np.random.uniform(-1, 1, xshape).astype(dtype)
+    mod.run(x=x_data)
+    results = [mod.get_output(i).asnumpy() for i in range(mod.get_num_outputs())]
+
+def test_tensorrt_ops():
+    if should_skip():
+        return
+    def run_and_verify(config):
+        f, input_shapes, is_param = config
+        params = {x: np.random.uniform(-1, 1, input_shapes[x]).astype(np.float32) for x in is_param}
+        input_dict = {k: np.random.uniform(-1, 1, v).astype(np.float32) for k, v in input_shapes.items() if k not in is_param}
+
+        # Run TRT 
+        mod = tvm.IRModule()
+        mod['main'] = f
+        mod = tensorrt.partition_for_tensorrt(mod, params)
+        with relay.build_config(opt_level=3):
+            graph, lib, graph_params = relay.build(mod, "cuda", params=params)
+        mod = graph_runtime.create(graph, lib, ctx=tvm.gpu(0))
+        mod.set_input(**graph_params)
+        mod.run(**input_dict)
+        results = [mod.get_output(i) for i in range(mod.get_num_outputs())]
+
+        # Run reference
+        mod = tvm.IRModule()
+        mod['main'] = f
+        with relay.build_config(opt_level=3):
+            graph, lib, graph_params = relay.build(mod, "cuda", params=params)

Review comment:
       I see, I can move all subtests from test_tensorrt_ops and test_tensorrt_integration into their own individual functions instead.




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org