refactor: Last trtorch references

Signed-off-by: Naren Dasan <[email protected]>
Signed-off-by: Naren Dasan <[email protected]>

Author: narendasan

examples/custom_converters/README.md

@@ -16,13 +16,13 @@ Note that ELU converter is now supported in our library. If you want to get abov
 error and run the example in this document, you can either:
 1. get the source code, go to root directory, then run: <br />
   `git apply ./examples/custom_converters/elu_converter/disable_core_elu.patch`
-2. If you are using a pre-downloaded release of TRTorch, you need to make sure that
-it doesn't support elu operator in default. (TRTorch <= v0.1.0)
+2. If you are using a pre-downloaded release of Torch-TensorRT, you need to make sure that
+it doesn't support elu operator in default. (Torch-TensorRT <= v0.1.0)
 
 ## Writing Converter in C++
 We can register a converter for this operator in our application. You can find more
 information on all the details of writing converters in the contributors documentation
-([Writing Converters](https://nvidia.github.io/TRTorch/contributors/writing_converters.html)).
+([Writing Converters](https://nvidia.github.io/Torch-TensorRT/contributors/writing_converters.html)).
 Once we are clear about these rules and writing patterns, we can create a seperate new C++ source file as:
 
 ```c++
@@ -66,7 +66,7 @@ from torch.utils import cpp_extension
 
 
 # library_dirs should point to the libtorch_tensorrt.so, include_dirs should point to the dir that include the headers
-# 1) download the latest package from https://github.com/NVIDIA/TRTorch/releases/
+# 1) download the latest package from https://github.com/NVIDIA/Torch-TensorRT/releases/
 # 2) Extract the file from downloaded package, we will get the "torch_tensorrt" directory
 # 3) Set torch_tensorrt_path to that directory
 torch_tensorrt_path = <PATH TO TRTORCH>
@@ -87,7 +87,7 @@ setup(
 ```
 Make sure to include the path for header files in `include_dirs` and the path
 for dependent libraries in `library_dirs`. Generally speaking, you should download
-the latest package from [here](https://github.com/NVIDIA/TRTorch/releases), extract
+the latest package from [here](https://github.com/NVIDIA/Torch-TensorRT/releases), extract
 the files, and the set the `torch_tensorrt_path` to it. You could also add other compilation
 flags in cpp_extension if you need. Then, run above python scripts as:
 ```shell
@@ -99,7 +99,7 @@ by the command above. In build folder, you can find the generated `.so` library,
 which could be loaded in our Python application.
 
 ## Load `.so` in Python Application
-With the new generated library, TRTorch now support the new developed converter.
+With the new generated library, Torch-TensorRT now support the new developed converter.
 We use `torch.ops.load_library` to load `.so`. For example, we could load the ELU
 converter and use it in our application:
 ```python
@@ -124,7 +124,7 @@ def cal_max_diff(pytorch_out, torch_tensorrt_out):
     diff = torch.sub(pytorch_out, torch_tensorrt_out)
     abs_diff = torch.abs(diff)
     max_diff = torch.max(abs_diff)
-    print("Maximum differnce between TRTorch and PyTorch: \n", max_diff)
+    print("Maximum differnce between Torch-TensorRT and PyTorch: \n", max_diff)
 
 
 def main():
@@ -146,12 +146,12 @@ def main():
 
     torch_tensorrt_out = trt_ts_module(input_data)
     print('PyTorch output: \n', pytorch_out[0, :, :, 0])
-    print('TRTorch output: \n', torch_tensorrt_out[0, :, :, 0])
+    print('Torch-TensorRT output: \n', torch_tensorrt_out[0, :, :, 0])
     cal_max_diff(pytorch_out, torch_tensorrt_out)
 
 
 if __name__ == "__main__":
     main()
 
 ```
-Run this script, we can get the different outputs from PyTorch and TRTorch.
+Run this script, we can get the different outputs from PyTorch and Torch-TensorRT.

examples/int8/ptq/README.md

@@ -161,7 +161,7 @@ export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$(pwd)/deps/torch_tensorrt/lib:$(pwd)/de
 
 2) Build and run `ptq`
 
-We import header files `cifar10.h` and `benchmark.h` from `ROOT_DIR`. `ROOT_DIR` should point to the path where Torch-TensorRT is located `<path_to_TRTORCH>`.
+We import header files `cifar10.h` and `benchmark.h` from `ROOT_DIR`. `ROOT_DIR` should point to the path where Torch-TensorRT is located `<path_to_torch_tensorrt>`.
 
 By default it is set to `../../../`. If your Torch-TensorRT directory structure is different, please set `ROOT_DIR` accordingly.
 

examples/int8/training/vgg16/test_qat.py

@@ -79,13 +79,12 @@ def test(model, dataloader, crit):
 print("[JIT] Test Loss: {:.5f} Test Acc: {:.2f}%".format(test_loss, 100 * test_acc))
 
 import torch_tensorrt as torchtrt
-# trtorch.logging.set_reportable_log_level(trtorch.logging.Level.Debug)
 compile_settings = {
-"inputs": [torchtrt.Input([1, 3, 32, 32])],
-"enabled_precisions": {torch.float, torch.half, torch.int8} # Run with FP16
+    "inputs": [torchtrt.Input([1, 3, 32, 32])],
+    "enabled_precisions": {torch.float, torch.half, torch.int8} # Run with FP16
 }
 new_mod = torch.jit.load('trained_vgg16_qat.jit.pt')
-trt_ts_module = torchtrt.compile(new_mod, compile_settings)
+trt_ts_module = torchtrt.compile(new_mod, **compile_settings)
 testing_dataloader = torch.utils.data.DataLoader(testing_dataset, batch_size=1, shuffle=False, num_workers=2)
 test_loss, test_acc = test(trt_ts_module, testing_dataloader, crit)
 print("[TRTorch] Test Loss: {:.5f} Test Acc: {:.2f}%".format(test_loss, 100 * test_acc))

examples/torchtrt_runtime_example/network.py

@@ -3,7 +3,7 @@
 import torch_tensorrt as torchtrt
 
 # create a simple norm layer.
-# This norm layer uses NormalizePlugin from TRTorch
+# This norm layer uses NormalizePlugin from Torch-TensorRT
 class Norm(torch.nn.Module):
     def __init__(self):
         super(Norm, self).__init__()
@@ -12,7 +12,7 @@ def forward(self, x):
         return torch.norm(x, 2, None, False)
 
 # Create a sample network with a conv and gelu node.
-# Gelu layer in TRTorch is converted to CustomGeluPluginDynamic from TensorRT plugin registry.
+# Gelu layer in Torch-TensorRT is converted to CustomGeluPluginDynamic from TensorRT plugin registry.
 class ConvGelu(torch.nn.Module):
     def __init__(self):
         super(ConvGelu, self).__init__()

py/BUILD

@@ -1,23 +1,23 @@
 package(default_visibility = ["//visibility:public"])
 
-load("@trtorch_py_deps//:requirements.bzl", "requirement")
+load("@torch_tensorrt_py_deps//:requirements.bzl", "requirement")
 
 # Exposes the library for testing
 py_library(
-    name = "trtorch",
+    name = "torch_tensorrt",
     srcs = [
-        "trtorch/__init__.py",
-        "trtorch/_compile_spec.py",
-        "trtorch/_compiler.py",
-        "trtorch/_types.py",
-        "trtorch/_version.py",
-        "trtorch/logging.py",
-        "trtorch/ptq.py",
+        "torch_tensorrt/__init__.py",
+        "torch_tensorrt/_compile_spec.py",
+        "torch_tensorrt/_compiler.py",
+        "torch_tensorrt/_types.py",
+        "torch_tensorrt/_version.py",
+        "torch_tensorrt/logging.py",
+        "torch_tensorrt/ptq.py",
     ],
     data = [
-        "trtorch/lib/libtrtorch.so",
+        "torch_tensorrt/lib/libtrtorch.so",
     ] + glob([
-        "trtorch/_C.cpython*.so",
+        "torch_tensorrt/_C.cpython*.so",
     ]),
     deps = [
         requirement("torch"),