Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix more tests #11

Merged
merged 1 commit into from
Oct 7, 2020
Merged

fix more tests #11

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 24 additions & 0 deletions tests/python/relay/test_json_runtime.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -108,18 +108,21 @@ def conv2d_direct():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data = relay.var("data", shape=(ishape), dtype=dtype)
weight = relay.var("weight", shape=(w1shape), dtype=dtype)
main_f = relay.Function([data, weight], glb_var(data, weight))
mod["main"] = main_f
mod = transform.InferType()(mod)

data0 = relay.var("data", shape=ishape, dtype=dtype)
weight0 = relay.var("weight", shape=w1shape, dtype=dtype)
out = relay.nn.conv2d(data0, weight0, kernel_size=(3, 3), padding=(1, 1))
main_f = relay.Function([data0, weight0], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

i_data = np.random.uniform(0, 1, ishape).astype(dtype)
w1_data = np.random.uniform(0, 1, w1shape).astype(dtype)
Expand All @@ -140,18 +143,21 @@ def group_conv2d():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data = relay.var("data", shape=(ishape), dtype=dtype)
weight = relay.var("weight", shape=(w2shape), dtype=dtype)
main_f = relay.Function([data, weight], glb_var(data, weight))
mod["main"] = main_f
mod = transform.InferType()(mod)

data0 = relay.var("data", shape=(ishape), dtype=dtype)
weight0 = relay.var("weight", shape=(w2shape), dtype=dtype)
out = relay.nn.conv2d(data0, weight0, kernel_size=(3, 3), padding=(1, 1), groups=32)
main_f = relay.Function([data0, weight0], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

i_data = np.random.uniform(0, 1, ishape).astype(dtype)
w_data = np.random.uniform(0, 1, w2shape).astype(dtype)
Expand Down Expand Up @@ -181,18 +187,21 @@ def gen_add():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data0 = relay.var("data0", shape=shape, dtype=dtype)
data1 = relay.var("data1", shape=shape, dtype=dtype)
main_f = relay.Function([data0, data1], glb_var(data0, data1))
mod["main"] = main_f
mod = transform.InferType()(mod)

data0 = relay.var("data0", shape=shape, dtype=dtype)
data1 = relay.var("data1", shape=shape, dtype=dtype)
out = relay.add(data0, data1)
main_f = relay.Function([data0, data1], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

return mod, ref_mod

Expand Down Expand Up @@ -221,16 +230,19 @@ def gen_relu():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data0 = relay.var("data0", shape=shape, dtype=dtype)
main_f = relay.Function([data0], glb_var(data0))
mod["main"] = main_f
mod = transform.InferType()(mod)

data0 = relay.var("data0", shape=shape, dtype=dtype)
out = relay.nn.relu(data0)
main_f = relay.Function([data0], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

return mod, ref_mod

Expand Down Expand Up @@ -268,18 +280,21 @@ def gen_dense():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

a = relay.var("A", shape=a_shape, dtype=dtype)
b = relay.var("B", shape=b_shape, dtype=dtype)
main_f = relay.Function([a, b], glb_var(a, b))
mod["main"] = main_f
mod = transform.InferType()(mod)

a = relay.var("A", shape=a_shape, dtype=dtype)
b = relay.var("B", shape=b_shape, dtype=dtype)
out = relay.nn.dense(a, b)
main_f = relay.Function([a, b], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

return mod, ref_mod

Expand Down Expand Up @@ -314,6 +329,7 @@ def gen_bn():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data = relay.var("data", shape=d_shape)
gamma = relay.var("gamma", shape=c_shape)
Expand All @@ -325,6 +341,7 @@ def gen_bn():
glb_var(data, gamma, beta, moving_mean, moving_var),
)
mod["main"] = main_f
mod = transform.InferType()(mod)

data = relay.var("data", shape=d_shape)
gamma = relay.var("gamma", shape=c_shape)
Expand All @@ -336,6 +353,7 @@ def gen_bn():
main_f = relay.Function([data, gamma, beta, moving_mean, moving_var], out)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_f
ref_mod = transform.InferType()(ref_mod)

return mod, ref_mod

Expand Down Expand Up @@ -457,12 +475,14 @@ def conv2d_relu():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = p_func
mod = transform.InferType()(mod)

# Main function
data = relay.var("data", shape=ishape, dtype=dtype)
weight = relay.var("weight", shape=w1shape, dtype=dtype)
main_func = relay.Function([data, weight], glb_var(data, weight))
mod["main"] = main_func
mod = transform.InferType()(mod)

# Reference module
data = relay.var("data", shape=ishape, dtype=dtype)
Expand All @@ -472,6 +492,7 @@ def conv2d_relu():
main_func = relay.Function([data, weight], relu)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_func
ref_mod = transform.InferType()(ref_mod)

i_data = np.random.uniform(0, 1, ishape).astype(dtype)
w1_data = np.random.uniform(0, 1, w1shape).astype(dtype)
Expand Down Expand Up @@ -504,13 +525,15 @@ def conv2d_bias_relu():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = p_func
mod = transform.InferType()(mod)

# Main function
data = relay.var("data", shape=ishape, dtype=dtype)
weight = relay.var("weight", shape=w1shape, dtype=dtype)
bias = relay.var("bias", shape=bshape, dtype=dtype)
main_func = relay.Function([data, weight, bias], glb_var(data, weight, bias))
mod["main"] = main_func
mod = transform.InferType()(mod)

# Reference module
data = relay.var("data", shape=ishape, dtype=dtype)
Expand All @@ -522,6 +545,7 @@ def conv2d_bias_relu():
main_func = relay.Function([data, weight, bias], relu)
ref_mod = tvm.IRModule()
ref_mod["main"] = main_func
ref_mod = transform.InferType()(ref_mod)

i_data = np.random.uniform(0, 1, ishape).astype(dtype)
w1_data = np.random.uniform(0, 1, w1shape).astype(dtype)
Expand Down
3 changes: 3 additions & 0 deletions tests/python/relay/test_pass_partition_graph.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -416,11 +416,13 @@ def expected():
glb_var = relay.GlobalVar("dnnl_0")
mod = tvm.IRModule()
mod[glb_var] = func
mod = transform.InferType()(mod)

data = relay.var("data", shape=(ishape), dtype=dtype)
weight = relay.var("input", shape=(w1shape), dtype=dtype)
main_f = relay.Function([data, weight], glb_var(data, weight))
mod["main"] = main_f
mod = transform.InferType()(mod)

return mod

Expand All @@ -440,6 +442,7 @@ def get_func():
mod = tvm.IRModule()
mod["main"] = WholeGraphAnnotator("dnnl").visit(get_func())
mod = transform.PartitionGraph()(mod)
mod = transform.InferType()(mod)

assert tvm.ir.structural_equal(mod, expected(), map_free_vars=True)

Expand Down