fix: update slice, correct typo

Signed-off-by: inocsin <[email protected]>

Author: inocsin

core/conversion/converters/converter_util.cpp

@@ -243,7 +243,7 @@ nvinfer1::ITensor* clamp_to_input_dim(ConversionCtx* ctx, nvinfer1::ITensor* x,
 }
 
 // return indices < 0 ? inputDims + indices : indices
-nvinfer1::ITensor* bump_if_negtive(ConversionCtx* ctx, nvinfer1::ITensor* input_dim, nvinfer1::ITensor* indices, int nbdims) {
+nvinfer1::ITensor* bump_if_negative(ConversionCtx* ctx, nvinfer1::ITensor* input_dim, nvinfer1::ITensor* indices, int nbdims) {
   auto zero = torch::zeros({nbdims}).to(torch::kI32);
   auto neg = -torch::ones({nbdims}).to(torch::kI32);
   auto zero_itensor = tensor_to_const(ctx, zero);
@@ -253,14 +253,14 @@ nvinfer1::ITensor* bump_if_negtive(ConversionCtx* ctx, nvinfer1::ITensor* input_
 
   // get the inputDim value where indices == -1, else 0
   auto mul = ctx->net->addElementWise(*signs, *input_dim, nvinfer1::ElementWiseOperation::kPROD);
-  TORCHTRT_CHECK(mul, "Unable to create mul layer in bump_if_negtive");
-  LOG_DEBUG(ctx->logger, "Create " << mul->getName() << " for bump_if_negtive");
+  TORCHTRT_CHECK(mul, "Unable to create mul layer in bump_if_negative");
+  LOG_DEBUG(ctx->logger, "Create " << mul->getName() << " for bump_if_negative");
   auto mul_itensor = mul->getOutput(0);
 
   // add the inputDim value to indices where indices == -1
   auto sub = ctx->net->addElementWise(*indices, *mul_itensor, nvinfer1::ElementWiseOperation::kSUB);
-  TORCHTRT_CHECK(sub, "Unable to create sub layer in bump_if_negtive");
-  LOG_DEBUG(ctx->logger, "Create " << sub->getName() << " for bump_if_negtive");
+  TORCHTRT_CHECK(sub, "Unable to create sub layer in bump_if_negative");
+  LOG_DEBUG(ctx->logger, "Create " << sub->getName() << " for bump_if_negative");
   auto sub_itensor = sub->getOutput(0);
   return sub_itensor;
 }
@@ -271,9 +271,9 @@ std::vector<nvinfer1::ITensor*> update_start_and_end(
     nvinfer1::ITensor* in_start,
     nvinfer1::ITensor* in_end,
     int nbdims) {
-  auto start = bump_if_negtive(ctx, in_shape, in_start, nbdims);
+  auto start = bump_if_negative(ctx, in_shape, in_start, nbdims);
   auto out_start = clamp_to_input_dim(ctx, start, in_shape, nbdims);
-  auto end = bump_if_negtive(ctx, in_shape, in_end, nbdims);
+  auto end = bump_if_negative(ctx, in_shape, in_end, nbdims);
   auto out_end = clamp_to_input_dim(ctx, end, in_shape, nbdims);
   std::vector<nvinfer1::ITensor*> outputs;
   outputs.push_back(out_start);
@@ -309,17 +309,6 @@ nvinfer1::ITensor* calculate_output_size(
   return size_itensor;
 }
 
-bool is_dynamic_shape(nvinfer1::ITensor* tensor) {
-  auto dim = tensor->getDimensions();
-  auto ndims = dim.nbDims;
-  for (int i = 0; i < ndims; i++) {
-    if (dim.d[i] == -1) {
-      return true;
-    }
-  }
-  return false;
-}
-
 } // namespace converters
 } // namespace conversion
 } // namespace core

core/conversion/converters/converter_util.h

@@ -57,7 +57,7 @@ nvinfer1::ITensor* clamp(
     nvinfer1::ITensor* lower_bound,
     nvinfer1::ITensor* upper_bound);
 
-nvinfer1::ITensor* bump_if_negtive(ConversionCtx* ctx, nvinfer1::ITensor* input_dim, nvinfer1::ITensor* indices);
+nvinfer1::ITensor* bump_if_negative(ConversionCtx* ctx, nvinfer1::ITensor* input_dim, nvinfer1::ITensor* indices);
 
 std::vector<nvinfer1::ITensor*> update_start_and_end(
     ConversionCtx* ctx,
@@ -73,8 +73,6 @@ nvinfer1::ITensor* calculate_output_size(
     nvinfer1::ITensor* stride,
     int nbdims);
 
-bool is_dynamic_shape(nvinfer1::ITensor* tensor);
-
 } // namespace converters
 } // namespace conversion
 } // namespace core

core/conversion/converters/impl/select.cpp

@@ -309,7 +309,7 @@ auto select_registrations TORCHTRT_UNUSED =
                auto in = args[0].ITensorOrFreeze(ctx);
                int axis = args[1].unwrapToInt();
                int maxDim = static_cast<int32_t>(in->getDimensions().d[axis]);
-               bool dynamic_shape = is_dynamic_shape(in);
+               bool dynamic_shape = ctx->input_is_dynamic;
                auto input_dim = in->getDimensions();
                // add Shape Tensor
                auto ishape_layer = ctx->net->addShape(*in);