[TF][Op] Add TF op Where

python/tvm/relay/frontend/tensorflow.py

@@ -937,7 +937,10 @@ def _impl(inputs, attr, params):
 
 def _where():
     def _impl(inputs, attr, params):
-        return AttrCvt(op_name="where")(inputs, attr)
+        if len(inputs) == 1:
+            return AttrCvt(op_name="argwhere")(inputs, attr)
+        else:
+            return AttrCvt(op_name="where")(inputs, attr)
     return _impl
 
 def _clip_by_value():
@@ -1354,6 +1357,7 @@ def _impl(inputs, attr, params):
     'Transpose'                         : _transpose(),
     'TruncateMod'                       : _elemwise('mod'),
     'Unpack'                            : _unpack(),
+    'Where'                             : _where(),
     'ZerosLike'                         : AttrCvt('zeros_like'),
 
 }

tests/python/frontend/tensorflow/test_forward.py

@@ -46,8 +46,34 @@ def convert_to_list(x):
         x = [x]
     return x
 
+def vmobj_to_list(o):
+    if isinstance(o, tvm.relay.backend.vmobj.TensorObject):
+        return [o.asnumpy().tolist()]
+    elif isinstance(o, tvm.relay.backend.vmobj.DatatypeObject):
+        result = []
+        for f in o:
+            result.extend(vmobj_to_list(f))
+        return result
+    elif isinstance(o, tvm.relay.backend.interpreter.TupleValue):
+        result = []
+        for f in o.fields:
+            result.append(vmobj_to_list(f))
+        return result
+    elif isinstance(o, tvm.relay.backend.interpreter.ConstructorValue):
+        if o.constructor.name_hint == 'cons':
+            tl = vmobj_to_list(o.fields[1])
+            hd = vmobj_to_list(o.fields[0])
+            hd.extend(tl)
+            return hd
+        elif o.constructor.name_hint == 'nil':
+            return []
+    elif isinstance(o, tvm.relay.backend.interpreter.TensorValue):
+        return [o.data.asnumpy()]
+    else:
+        raise RuntimeError("Unknown object type: %s" % type(o))
+
 def run_tvm_graph(graph_def, input_data, input_node, num_output=1,
-                  target='llvm', out_names=None, opt_level=3):
+                  target='llvm', out_names=None, opt_level=3, mode='graph_runtime'):
     """ Generic function to compile on relay and execute on tvm """
     input_data = convert_to_list(input_data)
     input_node = convert_to_list(input_node)
@@ -63,24 +89,32 @@ def run_tvm_graph(graph_def, input_data, input_node, num_output=1,
                                                  layout=layout,
                                                  shape=shape_dict,
                                                  outputs=out_names)
-    with relay.build_config(opt_level=opt_level):
-        graph, lib, params = relay.build(mod, target, target_host, params)
-
-    ctx = tvm.context(target, 0)
-    from tvm.contrib import graph_runtime
-    m = graph_runtime.create(graph, lib, ctx)
-    # set inputs
-    for e, i in zip(input_node, input_data):
-        m.set_input(e, tvm.nd.array(i))
-
-    m.set_input(**params)
-    # execute
-    m.run()
-    # get outputs
-    assert out_names is None or num_output == len(out_names), (
-        "out_names: {} num_output: {}".format(out_names, num_output))
-    tvm_output_list = [m.get_output(i).asnumpy() for i in range(num_output)]
-    return tvm_output_list
+    if mode == 'interp':
+        ex = relay.create_executor("debug", mod=mod, ctx=tvm.cpu(), target="llvm")
+        inputs = []
+        for param in mod['main'].params:
+            inputs.append(tvm.nd.array(params[param.name_hint]))
+        result = ex.evaluate()(*inputs)
+        return vmobj_to_list(result)
+    else:
+        with relay.build_config(opt_level=opt_level):
+            graph, lib, params = relay.build(mod, target, target_host, params)
+
+        ctx = tvm.context(target, 0)
+        from tvm.contrib import graph_runtime
+        m = graph_runtime.create(graph, lib, ctx)
+        # set inputs
+        for e, i in zip(input_node, input_data):
+            m.set_input(e, tvm.nd.array(i))
+
+        m.set_input(**params)
+        # execute
+        m.run()
+        # get outputs
+        assert out_names is None or num_output == len(out_names), (
+            "out_names: {} num_output: {}".format(out_names, num_output))
+        tvm_output_list = [m.get_output(i).asnumpy() for i in range(num_output)]
+        return tvm_output_list
 
 def run_tf_graph(sess, input_data, input_node, output_node):
     """ Generic function to execute tensorflow """
@@ -97,7 +131,7 @@ def run_tf_graph(sess, input_data, input_node, output_node):
 
 
 def compare_tf_with_tvm(in_data, in_name, out_name, init_global_variables=False,
-                        no_gpu=False, opt_level=3):
+                        no_gpu=False, opt_level=3, mode='graph_runtime'):
     """Generic function to generate and compare tensorflow and TVM output"""
     def name_without_num(name):
         return name.split(':')[0] if ":" in name else name
@@ -128,7 +162,7 @@ def name_without_num(name):
 
             tvm_output = run_tvm_graph(final_graph_def, in_data, in_node,
                                        target=device, out_names=out_name,
-                                       num_output=len(out_name), opt_level=opt_level)
+                                       num_output=len(out_name), opt_level=opt_level, mode=mode)
             # since the names from tensorflow and relay runs are not exactly same,
             # first len(tf_output) will be compared
             for i in range(len(tf_output)):
@@ -325,6 +359,19 @@ def test_forward_biasadd():
     _test_biasadd([4, 17, 17, 19], 'NHWC')
     _test_biasadd([4, 3, 3, 124], 'NHWC')
 
+def _test_forward_where(input_shape):
+    with tf.Graph().as_default():
+        dtype =  tf.float32
+        t = tf.constant(np.random.choice([0, 1, 2, 3], size=input_shape).astype(dtype.name))
+        out = tf.where(t)
+        compare_tf_with_tvm([], [], out.name, mode='interp')
+
+def test_forward_argwhere():
+    _test_forward_where((5, 5))
+    _test_forward_where((5, 5, 5))
+    _test_forward_where((5, 5, 5, 5))
+    _test_forward_where((5, 5, 5, 5, 5))
+
 #######################################################################
 # SpaceToBatchND
 # --------------