feat(aten::flatten): Adds a converter for aten flatten since MM is the

preferred path now Signed-off-by: Naren Dasan <[email protected]> Signed-off-by: Naren Dasan <[email protected]>
pytorch · May 1, 2020 · d945eb9 · d945eb9
1 parent 4acc3fd
commit d945eb9
Show file tree

Hide file tree

Showing 2 changed files with 85 additions and 17 deletions.
diff --git a/core/conversion/converters/impl/shuffle.cpp b/core/conversion/converters/impl/shuffle.cpp
@@ -1,5 +1,7 @@
 #include "core/conversion/converters/converters.h"
 
+#include "torch/torch.h"
+
 namespace trtorch {
 namespace core {
 namespace conversion {
@@ -8,23 +10,41 @@ namespace impl {
 namespace {
 
 static auto shuffle_registrations = RegisterNodeConversionPatterns()
-    .pattern({
-        "aten::reshape(Tensor self, int[] shape) -> (Tensor)",
-        [](ConversionCtx* ctx, const torch::jit::Node* n, args& args) -> bool {
-            auto in = args[0].ITensor();
-            auto new_shape = util::toDimsPad(args[1].unwrapToIntList(), 2);
-
-            auto shuffle = ctx->net->addShuffle(*in);
-            TRTORCH_CHECK(shuffle, "Unable to create shuffle layer from node: " << *n);
-            shuffle->setReshapeDimensions(new_shape);
-            shuffle->setName(util::node_info(n).c_str());
-
-            auto out_tensor = ctx->AssociateValueAndTensor(n->outputs()[0], shuffle->getOutput(0));
-            LOG_DEBUG("Output tensor shape: " << out_tensor->getDimensions());
-
-            return true;
-        }
-    });
+  .pattern({
+    "aten::flatten.using_ints(Tensor self, int start_dim=0, int end_dim=-1) -> (Tensor)",
+    [](ConversionCtx* ctx, const torch::jit::Node* n, args& args) -> bool {
+      auto in = args[0].ITensor();
+      auto start_dim = args[1].unwrapToInt();
+      auto end_dim = args[2].unwrapToInt();
+      auto in_shape = util::toVec(in->getDimensions());
+      auto out_shape = torch::flatten(torch::rand(in_shape), start_dim, end_dim).sizes();
+
+      auto shuffle = ctx->net->addShuffle(*in);
+      TRTORCH_CHECK(shuffle, "Unable to create shuffle layer from node: " << *n);
+      shuffle->setReshapeDimensions(util::toDims(out_shape));
+      shuffle->setName(util::node_info(n).c_str());
+
+      auto out_tensor = ctx->AssociateValueAndTensor(n->outputs()[0], shuffle->getOutput(0));
+      LOG_DEBUG("Output tensor shape: " << out_tensor->getDimensions());
+      return true;
+    }
+  }).pattern({
+    "aten::reshape(Tensor self, int[] shape) -> (Tensor)",
+    [](ConversionCtx* ctx, const torch::jit::Node* n, args& args) -> bool {
+      auto in = args[0].ITensor();
+      auto new_shape = util::toDimsPad(args[1].unwrapToIntList(), 2);
+
+      auto shuffle = ctx->net->addShuffle(*in);
+      TRTORCH_CHECK(shuffle, "Unable to create shuffle layer from node: " << *n);
+      shuffle->setReshapeDimensions(new_shape);
+      shuffle->setName(util::node_info(n).c_str());
+
+      auto out_tensor = ctx->AssociateValueAndTensor(n->outputs()[0], shuffle->getOutput(0));
+      LOG_DEBUG("Output tensor shape: " << out_tensor->getDimensions());
+
+      return true;
+    }
+  });
 } // namespace
 } // namespace impl
 } // namespace converters

diff --git a/tests/core/converters/test_shuffle.cpp b/tests/core/converters/test_shuffle.cpp
@@ -4,6 +4,54 @@
 #include "tests/util/util.h"
 #include "core/compiler.h"
 
+// TODO: IR Parser doesnt work well with neg numbers
+TEST(Converters, ATenFlattenConvertsCorrectly) {
+    const auto graph = R"IR(
+      graph(%0 : Tensor):
+        %1 : int = prim::Constant[value=0]()
+        %2 : int = prim::Constant[value=1]()
+        %3 : Tensor = aten::flatten(%0, %1, %2)
+        return (%3))IR";
+
+    auto g = std::make_shared<torch::jit::Graph>();
+    torch::jit::parseIR(graph, &*g);
+
+    auto in = at::randint(0, 5, {2, 3}, {at::kCUDA});
+    auto params = trtorch::core::conversion::get_named_params(g->inputs(), {});
+    auto jit_results = trtorch::tests::util::RunGraph(g, params, {in});
+
+    in = at::clone(in);
+    params = trtorch::core::conversion::get_named_params(g->inputs(), {});
+    auto trt_results = trtorch::tests::util::RunGraphEngine(g, params, {in});
+    auto trt = trt_results[0].reshape_as(jit_results[0]);
+
+    ASSERT_TRUE(trtorch::tests::util::almostEqual(jit_results[0], trt, 2e-6));
+}
+
+// TODO: IR Parser doesnt work well with neg numbers
+TEST(Converters, ATenFlattenOtherDimsConvertsCorrectly) {
+    const auto graph = R"IR(
+      graph(%0 : Tensor):
+        %1 : int = prim::Constant[value=1]()
+        %2 : int = prim::Constant[value=2]()
+        %3 : Tensor = aten::flatten(%0, %1, %2)
+        return (%3))IR";
+
+    auto g = std::make_shared<torch::jit::Graph>();
+    torch::jit::parseIR(graph, &*g);
+
+    auto in = at::randint(0, 5, {2, 3, 3}, {at::kCUDA});
+    auto params = trtorch::core::conversion::get_named_params(g->inputs(), {});
+    auto jit_results = trtorch::tests::util::RunGraph(g, params, {in});
+
+    in = at::clone(in);
+    params = trtorch::core::conversion::get_named_params(g->inputs(), {});
+    auto trt_results = trtorch::tests::util::RunGraphEngine(g, params, {in});
+    auto trt = trt_results[0].reshape_as(jit_results[0]);
+
+    ASSERT_TRUE(trtorch::tests::util::almostEqual(jit_results[0], trt, 2e-6));
+}
+
 TEST(Converters, ATenReshapeConvertsCorrectly) {
     const auto graph = R"IR(
       graph(%0 : Tensor):