[Doc] Remove MxNet related tutorials

apps/benchmark/adreno/adreno_gpu_bench_clml.py

@@ -84,17 +84,6 @@ def get_network(name, batch_size, dtype="float32"):
         net, params = testing.squeezenet.get_workload(
             batch_size=batch_size, version=version, dtype=dtype
         )
-    elif name == "mxnet":
-        # an example for mxnet model
-        from mxnet.gluon.model_zoo.vision import get_model
-
-        block = get_model("resnet18_v1", pretrained=True)
-        net, params = relay.frontend.from_mxnet(block, shape={"data": input_shape}, dtype=dtype)
-        net = net["main"]
-        net = relay.Function(
-            net.params, relay.nn.softmax(net.body), None, net.type_params, net.attrs
-        )
-        net = tvm.IRModule.from_expr(net)
     else:
         raise ValueError("Unsupported network: " + name)
 

apps/benchmark/adreno/adreno_gpu_bench_texture.py

@@ -83,17 +83,6 @@ def get_network(name, batch_size, dtype="float32"):
         net, params = testing.squeezenet.get_workload(
             batch_size=batch_size, version=version, dtype=dtype
         )
-    elif name == "mxnet":
-        # an example for mxnet model
-        from mxnet.gluon.model_zoo.vision import get_model
-
-        block = get_model("resnet18_v1", pretrained=True)
-        net, params = relay.frontend.from_mxnet(block, shape={"data": input_shape}, dtype=dtype)
-        net = net["main"]
-        net = relay.Function(
-            net.params, relay.nn.softmax(net.body), None, net.type_params, net.attrs
-        )
-        net = tvm.IRModule.from_expr(net)
     else:
         raise ValueError("Unsupported network: " + name)
 

apps/benchmark/util.py

@@ -72,17 +72,6 @@ def get_network(name, batch_size, dtype="float32"):
         net, params = testing.squeezenet.get_workload(
             batch_size=batch_size, version=version, dtype=dtype
         )
-    elif name == "mxnet":
-        # an example for mxnet model
-        from mxnet.gluon.model_zoo.vision import get_model
-
-        block = get_model("resnet18_v1", pretrained=True)
-        net, params = relay.frontend.from_mxnet(block, shape={"data": input_shape}, dtype=dtype)
-        net = net["main"]
-        net = relay.Function(
-            net.params, relay.nn.softmax(net.body), None, net.type_params, net.attrs
-        )
-        net = tvm.IRModule.from_expr(net)
     else:
         raise ValueError("Unsupported network: " + name)
 

docs/conf.py

@@ -441,7 +441,6 @@ def jupyter_notebook(script_blocks, gallery_conf, target_dir, real_func):
     for p in [
         tvm_path / "vta" / "tutorials" / "frontend",
         tvm_path / "vta" / "tutorials" / "optimize",
-        tvm_path / "vta" / "tutorials" / "autotvm",
     ]
 )
 

gallery/how_to/deploy_models/deploy_model_on_nano.py

@@ -102,22 +102,22 @@
 # -----------------------------
 # Back to the host machine, which should have a full TVM installed (with LLVM).
 #
-# We will use pre-trained model from
-# `MXNet Gluon model zoo <https://mxnet.apache.org/api/python/gluon/model_zoo.html>`_.
-# You can found more details about this part at tutorial :ref:`tutorial-from-mxnet`.
+# We will use pre-trained model from torchvision
 
-import sys
-
-from mxnet.gluon.model_zoo.vision import get_model
+import torch
+import torchvision
 from PIL import Image
 import numpy as np
 
 # one line to get the model
-try:
-    block = get_model("resnet18_v1", pretrained=True)
-except RuntimeError:
-    print("Downloads from mxnet no longer supported", file=sys.stderr)
-    sys.exit(0)
+model_name = "resnet18"
+model = getattr(torchvision.models, model_name)(pretrained=True)
+model = model.eval()
+
+# We grab the TorchScripted model via tracing
+input_shape = [1, 3, 224, 224]
+input_data = torch.randn(input_shape)
+scripted_model = torch.jit.trace(model, input_data).eval()
 
 ######################################################################
 # In order to test our model, here we download an image of cat and
@@ -158,9 +158,9 @@ def transform_image(image):
 # Now we would like to port the Gluon model to a portable computational graph.
 # It's as easy as several lines.
 
-# We support MXNet static graph(symbol) and HybridBlock in mxnet.gluon
-shape_dict = {"data": x.shape}
-mod, params = relay.frontend.from_mxnet(block, shape_dict)
+input_name = "input0"
+shape_list = [(input_name, x.shape)]
+mod, params = relay.frontend.from_pytorch(scripted_model, shape_list)
 # we want a probability so add a softmax operator
 func = mod["main"]
 func = relay.Function(func.params, relay.nn.softmax(func.body), None, func.type_params, func.attrs)
@@ -241,7 +241,7 @@ def transform_image(image):
 
 module = runtime.GraphModule(rlib["default"](dev))
 # set input data
-module.set_input("data", tvm.nd.array(x.astype("float32")))
+module.set_input(input_name, tvm.nd.array(x.astype("float32")))
 # run
 module.run()
 # get output

gallery/how_to/deploy_models/deploy_model_on_rasp.py

@@ -95,22 +95,22 @@
 # -----------------------------
 # Back to the host machine, which should have a full TVM installed (with LLVM).
 #
-# We will use pre-trained model from
-# `MXNet Gluon model zoo <https://mxnet.apache.org/api/python/gluon/model_zoo.html>`_.
-# You can found more details about this part at tutorial :ref:`tutorial-from-mxnet`.
+# We will use pre-trained model from torchvision
 
-import sys
-
-from mxnet.gluon.model_zoo.vision import get_model
+import torch
+import torchvision
 from PIL import Image
 import numpy as np
 
 # one line to get the model
-try:
-    block = get_model("resnet18_v1", pretrained=True)
-except RuntimeError:
-    print("Downloads from mxnet no longer supported", file=sys.stderr)
-    sys.exit(0)
+model_name = "resnet18"
+model = getattr(torchvision.models, model_name)(pretrained=True)
+model = model.eval()
+
+# We grab the TorchScripted model via tracing
+input_shape = [1, 3, 224, 224]
+input_data = torch.randn(input_shape)
+scripted_model = torch.jit.trace(model, input_data).eval()
 
 ######################################################################
 # In order to test our model, here we download an image of cat and
@@ -148,12 +148,12 @@ def transform_image(image):
     synset = eval(f.read())
 
 ######################################################################
-# Now we would like to port the Gluon model to a portable computational graph.
+# Now we would like to port the PyTorch model to a portable computational graph.
 # It's as easy as several lines.
 
-# We support MXNet static graph(symbol) and HybridBlock in mxnet.gluon
-shape_dict = {"data": x.shape}
-mod, params = relay.frontend.from_mxnet(block, shape_dict)
+input_name = "input0"
+shape_list = [(input_name, x.shape)]
+mod, params = relay.frontend.from_pytorch(scripted_model, shape_list)
 # we want a probability so add a softmax operator
 func = mod["main"]
 func = relay.Function(func.params, relay.nn.softmax(func.body), None, func.type_params, func.attrs)
@@ -226,7 +226,7 @@ def transform_image(image):
 dev = remote.cpu(0)
 module = runtime.GraphModule(rlib["default"](dev))
 # set input data
-module.set_input("data", tvm.nd.array(x.astype("float32")))
+module.set_input(input_name, tvm.nd.array(x.astype("float32")))
 # run
 module.run()
 # get output