[Frontend]Make onnx gemm tensor C optional

python/tvm/relay/frontend/onnx.py

@@ -513,7 +513,9 @@ class Gemm(OnnxOpConverter):
 
     @classmethod
     def _impl_v1(cls, inputs, attr, params):
-        assert len(inputs) == 3, "Gemm op take 3 inputs, {} given".format(len(inputs))
+        assert len(inputs) == 3 or len(inputs) == 2, "Gemm op take 2 or 3 inputs, {} given".format(
+            len(inputs)
+        )
         # Y = alpha * A * B + beta * C
         alpha = float(attr.get("alpha", 1.0))
         beta = float(attr.get("beta", 1.0))
@@ -531,9 +533,12 @@ def _impl_v1(cls, inputs, attr, params):
             inputs[0] *= _expr.const(alpha)
         out = _op.nn.dense(inputs[0], inputs[1], units=channels)
 
-        # skip (beta * C) if zero
-        C_array = params[inputs[2].name_hint].asnumpy()
-        if (beta == 0.0) or np.array_equal(C_array, np.array([0])):
+        if len(inputs) == 3:
+            # skip (beta * C) if zero
+            C_array = params[inputs[2].name_hint].asnumpy()
+            if (beta == 0.0) or np.array_equal(C_array, np.array([0])):
+                return out
+        else:
             return out
         return _op.nn.bias_add(out, _expr.const(beta) * inputs[2])
 

tests/python/frontend/onnx/test_forward.py

@@ -1008,6 +1008,31 @@ def test_onehot():
         tvm.testing.assert_allclose(out_np, tvm_out, rtol=1e-5, atol=1e-5)
 
 
+@tvm.testing.uses_gpu
+def test_gemm():
+    a_shape = (4, 3)
+    b_shape = (3, 4)
+    out_shape = [a_shape[0], b_shape[1]]
+
+    a_array = np.random.uniform(size=a_shape).astype("float32")
+    b_array = np.random.uniform(size=b_shape).astype("float32")
+
+    gemm_node = helper.make_node("Gemm", ["a", "b"], ["out"])
+
+    graph = helper.make_graph(
+        [gemm_node],
+        "gemm_test",
+        inputs=[
+            helper.make_tensor_value_info("a", TensorProto.FLOAT, list(a_shape)),
+            helper.make_tensor_value_info("b", TensorProto.FLOAT, list(b_shape)),
+        ],
+        outputs=[helper.make_tensor_value_info("out", TensorProto.FLOAT, list(out_shape))],
+    )
+
+    model = helper.make_model(graph, producer_name="gemm_test")
+    verify_with_ort_with_inputs(model, [a_array, b_array])
+
+
 @tvm.testing.uses_gpu
 def test_matmul():
     a_shape = (4, 3)
@@ -4065,6 +4090,7 @@ def verify_cumsum(indata, axis, exclusive=0, reverse=0, type="float32"):
     test_clip()
     test_clip_min_max_as_inputs()
     test_onehot()
+    test_gemm()
     test_matmul()
     test_gather()
     test_gatherelements()