Renames package to nvtripy to avoid PyPI naming collisions

tripy/CONTRIBUTING.md

@@ -25,24 +25,24 @@ Thank you for considering contributing to Tripy!
         docker login ghcr.io/nvidia/tensorrt-incubator
         ```
 
-        Next, pull and launch the container. From the [`tripy` root directory](.), run:
+        Next, pull and launch the container. From the [`nvtripy` root directory](.), run:
 
         ```bash
-        docker run --pull always --gpus all -it -p 8080:8080 -v $(pwd):/tripy/ --rm ghcr.io/nvidia/tensorrt-incubator/tripy
+        docker run --pull always --gpus all -it -p 8080:8080 -v $(pwd):/nvtripy/ --rm ghcr.io/nvidia/tensorrt-incubator/nvtripy
         ```
 
     - Otherwise, you can build the container locally and launch it.
-        From the [`tripy` root directory](.), run:
+        From the [`nvtripy` root directory](.), run:
 
         ```bash
-        docker build -t tripy .
-        docker run --gpus all -it -p 8080:8080 -v $(pwd):/tripy/ --rm tripy:latest
+        docker build -t nvtripy .
+        docker run --gpus all -it -p 8080:8080 -v $(pwd):/nvtripy/ --rm nvtripy:latest
         ```
 
-3. You should now be able to use `tripy` in the container. To test it out, you can run a quick sanity check:
+3. You should now be able to use `nvtripy` in the container. To test it out, you can run a quick sanity check:
 
     ```bash
-    python3 -c "import tripy as tp; print(tp.ones((2, 3)))"
+    python3 -c "import nvtripy as tp; print(tp.ones((2, 3)))"
     ```
 
     This should give you some output like:
@@ -64,7 +64,7 @@ You only need to do this once.
 We suggest you do this *outside* the container and also use `git` from
 outside the container.
 
-From the [`tripy` root directory](.), run:
+From the [`nvtripy` root directory](.), run:
 ```bash
 python3 -m pip install pre-commit
 pre-commit install
@@ -87,7 +87,7 @@ If you're intersted in adding a new operator to Tripy, refer to [this guide](./d
     - Create a PR to merge changes from fork to the main repo
 
 2. Managing PRs
-    - Label your PR correctly (e.g., use `tripy` for changes to `tripy`).
+    - Label your PR correctly (e.g., use `nvtripy` for changes to `nvtripy`).
     - Add a brief description explaining the purpose of the change.
     - Each functional change should include an update to an existing test or a new test.
     - Ensure any commits you make are signed. See [this page](https://docs.github.com/en/authentication/managing-commit-signature-verification/about-commit-signature-verification#ssh-commit-signature-verification)
@@ -153,9 +153,9 @@ The Tripy container includes a build of MLIR-TensorRT, but in some cases, you ma
 
 1. Build MLIR-TensorRT as per the instructions in the [README](../mlir-tensorrt/README.md).
 
-2. Launch the container with mlir-tensorrt repository mapped for accessing wheels files; from the [`tripy` root directory](.), run:
+2. Launch the container with mlir-tensorrt repository mapped for accessing wheels files; from the [`nvtripy` root directory](.), run:
     ```bash
-    docker run --gpus all -it -p 8080:8080 -v $(pwd):/tripy/ -v $(pwd)/../mlir-tensorrt:/mlir-tensorrt  --rm tripy:latest
+    docker run --gpus all -it -p 8080:8080 -v $(pwd):/nvtripy/ -v $(pwd)/../mlir-tensorrt:/mlir-tensorrt  --rm nvtripy:latest
     ```
 
 3. Install MLIR-TensorRT wheels
@@ -180,5 +180,5 @@ The Tripy container includes a build of MLIR-TensorRT, but in some cases, you ma
 
 4. Verify everything works:
     ```bash
-    python3 -c "import tripy as tp; print(tp.ones((2, 3)))"
+    python3 -c "import nvtripy as tp; print(tp.ones((2, 3)))"
     ```

tripy/Dockerfile

@@ -2,7 +2,7 @@ FROM ubuntu:22.04
 
 LABEL org.opencontainers.image.description="Tripy development container"
 
-WORKDIR /tripy
+WORKDIR /nvtripy
 
 SHELL ["/bin/bash", "-c"]
 
@@ -23,7 +23,7 @@ RUN apt-get update && \
     python3 -m pip install --upgrade pip
 
 COPY .lldbinit /root/
-COPY pyproject.toml /tripy/pyproject.toml
+COPY pyproject.toml /nvtripy/pyproject.toml
 
 RUN pip install build .[docs,dev,test,build] \
     -f https://nvidia.github.io/TensorRT-Incubator/packages.html \
@@ -46,5 +46,5 @@ RUN echo "deb http://apt.llvm.org/jammy/ llvm-toolchain-jammy-$LLVM_VERSION main
     rm -rf /var/lib/apt/lists/* && \
     ln -s /usr/bin/lldb-17 /usr/bin/lldb
 
-# Export tripy into the PYTHONPATH so it doesn't need to be installed after making changes
-ENV PYTHONPATH=/tripy
+# Export nvtripy into the PYTHONPATH so it doesn't need to be installed after making changes
+ENV PYTHONPATH=/nvtripy

tripy/README.md

@@ -29,11 +29,11 @@ an excellent user experience without compromising performance. Some of the goals
 <!-- Tripy: DOC: OMIT End -->
 
 ```bash
-python3 -m pip install --no-index -f https://nvidia.github.io/TensorRT-Incubator/packages.html tripy --no-deps
-python3 -m pip install -f https://nvidia.github.io/TensorRT-Incubator/packages.html tripy
+python3 -m pip install --no-index -f https://nvidia.github.io/TensorRT-Incubator/packages.html nvtripy --no-deps
+python3 -m pip install -f https://nvidia.github.io/TensorRT-Incubator/packages.html nvtripy
 ```
 
-***Important:** There is another package named `tripy` on PyPI.*
+***Important:** There is another package named `nvtripy` on PyPI.*
 *Note that it is **not** the package from this repository.*
 *Please use the instructions above to ensure you install the correct package.*
 
@@ -48,23 +48,23 @@ To get the latest changes in the repository, you can build Tripy wheels from sou
     python3 -m pip install build
     ```
 
-2. From the [`tripy` root directory](.), run:
+2. From the [`nvtripy` root directory](.), run:
 
     ```bash
     python3 -m build . -w
     ```
 
 3. Install the wheel, which should have been created in the `dist/` directory.
-    From the [`tripy` root directory](.), run:
+    From the [`nvtripy` root directory](.), run:
 
     ```bash
-    python3 -m pip install -f https://nvidia.github.io/TensorRT-Incubator/packages.html dist/tripy-*.whl
+    python3 -m pip install -f https://nvidia.github.io/TensorRT-Incubator/packages.html dist/nvtripy-*.whl
     ```
 
 4. **[Optional]** To ensure that Tripy was installed correctly, you can run a sanity check:
 
     ```bash
-    python3 -c "import tripy as tp; x = tp.ones((5,), dtype=tp.int32); assert x.tolist() == [1] * 5"
+    python3 -c "import nvtripy as tp; x = tp.ones((5,), dtype=tp.int32); assert x.tolist() == [1] * 5"
     ```
 
 <!-- Tripy: DOC: OMIT End -->
@@ -73,7 +73,7 @@ To get the latest changes in the repository, you can build Tripy wheels from sou
 
 We've included several guides in Tripy to make it easy to get started.
 We recommend starting with the
-[Introduction To Tripy](https://nvidia.github.io/TensorRT-Incubator/pre0_user_guides/00-introduction-to-tripy.html)
+[Introduction To Tripy](https://nvidia.github.io/TensorRT-Incubator/pre0_user_guides/00-introduction-to-nvtripy.html)
 guide.
 
 Other features covered in our guides include:

tripy/RELEASE.md

@@ -3,7 +3,7 @@
 This document explains how to release a new version of Tripy.
 
 1. Update version numbers in [`pyproject.toml`](./pyproject.toml) and
-    [`__init__.py`](./tripy/__init__.py) (make sure they match!).
+    [`__init__.py`](./nvtripy/__init__.py) (make sure they match!).
 
     Often, updates to Tripy will also require updates to dependencies,
     like MLIR-TRT, so make sure to update those version numbers as well.
@@ -21,7 +21,7 @@ This document explains how to release a new version of Tripy.
 
     Once the post-merge pipelines have succeeded, create a new tag with:
     ```bash
-    git tag tripy-vX.Y.Z
+    git tag nvtripy-vX.Y.Z
     ```
     replacing `X.Y.Z` with the version number and push it to the repository.
 

tripy/docs/README.md

@@ -21,7 +21,7 @@ generation, such as where in the documentation hierarchy the API should be docum
 The `generate_rsts.py` script uses this information to automatically generate a directory
 structure and populate it with `.rst` files.
 
-For more information, see the docstring for [`export.public_api()`](../tripy/export.py).
+For more information, see the docstring for [`export.public_api()`](../nvtripy/export.py).
 
 ### Docstrings
 
@@ -70,7 +70,7 @@ This means we need to make some special considerations:
     Myst can replace them with URLs to our remote repository. Otherwise, the links will
     cause the relevant file to be downloaded. For example:
     ```
-    [Fill operation](source:/tripy/frontend/trace/ops/fill.py)
+    [Fill operation](source:/nvtripy/frontend/trace/ops/fill.py)
     ```
 
     Links to markdown files are an exception; if a markdown file is part of the *rendered*
@@ -85,7 +85,7 @@ This means we need to make some special considerations:
     For example:
 
     ```md
-    {class}`tripy.Tensor`
+    {class}`nvtripy.Tensor`
     ```
 
     `<api_kind>` can take on any value that is a valid role provided by
@@ -101,11 +101,11 @@ Code examples in public facing docstrings and guides are preprocessed before
 documentation is generated. Specifically:
 
 - Any code examples are executed so that their output can be
-    displayed after the code block. Several modules, including `tripy` (as `tp`),
+    displayed after the code block. Several modules, including `nvtripy` (as `tp`),
     `numpy` (as `np`), `cupy` (as `cp`), and `torch` are automatically imported
     and can be used in code examples.
 
-- The values of any `tripy` type local variables are appended to the output.
+- The values of any `nvtripy` type local variables are appended to the output.
     You can customize this behavior:
 
     - To only display certain variables, add `# doc: print-locals` followed by a space

tripy/docs/conf.py

@@ -26,9 +26,9 @@
 
 from tests import helper
 
-import tripy as tp
-from tripy.common.datatype import DATA_TYPES
-from tripy.wrappers import TYPE_VERIFICATION
+import nvtripy as tp
+from nvtripy.common.datatype import DATA_TYPES
+from nvtripy.wrappers import TYPE_VERIFICATION
 
 PARAM_PAT = re.compile(":param .*?:")
 
@@ -52,8 +52,8 @@
 python_use_unqualified_type_names = True
 
 nitpick_ignore = {
-    ("py:class", "tripy.types.ShapeLike"),
-    ("py:class", "tripy.types.TensorLike"),
+    ("py:class", "nvtripy.types.ShapeLike"),
+    ("py:class", "nvtripy.types.TensorLike"),
     ("py:class", "Tensor"),
 }
 nitpick_ignore_regex = {
@@ -140,7 +140,7 @@
 myst_url_schemes = {
     "http": None,
     "https": None,
-    "source": "https://github.com/NVIDIA/TensorRT-Incubator/tree/main/tripy/{{path}}",
+    "source": "https://github.com/NVIDIA/TensorRT-Incubator/tree/main/nvtripy/{{path}}",
 }
 myst_number_code_blocks = ["py", "rst"]
 
@@ -155,7 +155,7 @@
 def process_docstring_impl(app, what, name, obj, options, lines):
     doc = "\n".join(lines).strip()
     blocks = helper.consolidate_code_blocks(doc)
-    name = name.lstrip("tripy.")
+    name = name.lstrip("nvtripy.")
 
     # Check signature for functions/methods and class constructors.
     if what in {"function", "method"} or (what == "class" and name in seen_classes):
@@ -265,9 +265,9 @@ def allow_no_example():
         # `tp.Module`s include examples in their constructors, so their __call__ methods don't require examples.
         is_tripy_module_call_method = False
         if what == "method" and obj.__name__ == "__call__":
-            class_name = "tripy." + name.rpartition(".")[0]
-            # Class names are prefixed with tripy.<...>, so we need to import it here to make eval() work.
-            import tripy
+            class_name = "nvtripy." + name.rpartition(".")[0]
+            # Class names are prefixed with nvtripy.<...>, so we need to import it here to make eval() work.
+            import nvtripy
 
             is_tripy_module_call_method = issubclass(eval(class_name), tp.Module)
 
@@ -335,7 +335,7 @@ def setup(app):
     # A note on aliases: if you rename a class via an import statement, e.g. `import X as Y`,
     # the documentation generated for `Y` will just be: "Alias of X"
     # To get the real documentation, you can make Sphinx think that `Y` is not an alias but instead a real
-    # class/function. To do so, you just need to define the __name__ attribute in this function (*not* in tripy code!):
+    # class/function. To do so, you just need to define the __name__ attribute in this function (*not* in nvtripy code!):
     #   Y.__name__ = "Y"
 
     app.connect("autodoc-process-docstring", process_docstring)

tripy/docs/generate_rsts.py

@@ -27,9 +27,9 @@
 from textwrap import dedent, indent
 from typing import Dict, List, Set
 
-import tripy as tp
+import nvtripy as tp
 from tests import helper
-from tripy.export import PUBLIC_APIS
+from nvtripy.export import PUBLIC_APIS
 
 
 @dataclass

tripy/docs/packages.html

@@ -10,22 +10,22 @@
 <body>
     <h1>Package Index</h1>
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.6/tripy-0.0.6-py3-none-any.whl">tripy-0.0.6-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.6/nvtripy-0.0.6-py3-none-any.whl">nvtripy-0.0.6-py3-none-any.whl</a><br>
 
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.5/tripy-0.0.5-py3-none-any.whl">tripy-0.0.5-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.5/nvtripy-0.0.5-py3-none-any.whl">nvtripy-0.0.5-py3-none-any.whl</a><br>
 
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.4/tripy-0.0.4-py3-none-any.whl">tripy-0.0.4-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.4/nvtripy-0.0.4-py3-none-any.whl">nvtripy-0.0.4-py3-none-any.whl</a><br>
 
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.3/tripy-0.0.3-py3-none-any.whl">tripy-0.0.3-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.3/nvtripy-0.0.3-py3-none-any.whl">nvtripy-0.0.3-py3-none-any.whl</a><br>
 
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.2/tripy-0.0.2-py3-none-any.whl">tripy-0.0.2-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.2/nvtripy-0.0.2-py3-none-any.whl">nvtripy-0.0.2-py3-none-any.whl</a><br>
 
     <a
-        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/tripy-v0.0.1/tripy-0.0.1-py3-none-any.whl">tripy-0.0.1-py3-none-any.whl</a><br>
+        href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/nvtripy-v0.0.1/nvtripy-0.0.1-py3-none-any.whl">nvtripy-0.0.1-py3-none-any.whl</a><br>
 
     <a
         href="https://github.com/NVIDIA/TensorRT-Incubator/releases/download/mlir-tensorrt-v0.1.29/mlir_tensorrt_compiler-0.1.29+cuda12.trt102-cp310-cp310-linux_x86_64.whl">mlir_tensorrt_compiler-0.1.29+cuda12.trt102-cp310-cp310-linux_x86_64.whl</a><br>

tripy/docs/post0_developer_guides/architecture.md

@@ -80,12 +80,12 @@ inp = tp.full((2, 3), value=0.5)
 The `tp.full()` and `tp.tanh()` APIs are part of the frontend and like other frontend functions, map to one or more
 (just one in this case) `Trace` operations. For frontend functions that map to exactly one `Trace` operation,
 we define the function directly alongside the corresponding `Trace` operation.
-In this case, the [`Fill` operation](source:/tripy/frontend/trace/ops/fill.py) provides `tp.full()` and
-the [`UnaryElementwise` operation](source:/tripy/frontend/trace/ops/unary_elementwise.py) provides `tp.tanh()`.
+In this case, the [`Fill` operation](source:/nvtripy/frontend/trace/ops/fill.py) provides `tp.full()` and
+the [`UnaryElementwise` operation](source:/nvtripy/frontend/trace/ops/unary_elementwise.py) provides `tp.tanh()`.
 
 *We organize it this way to reduce the number of files that need to be touched when adding new ops.*
     *If an operation is composed of multiple `Trace` operations, the frontend function can be*
-    *defined under the [`frontend/ops`](source:/tripy/frontend/ops) submodule instead.*
+    *defined under the [`frontend/ops`](source:/nvtripy/frontend/ops) submodule instead.*
 
 #### What Does It Do?
 
@@ -145,7 +145,7 @@ Here's the textual representation for the `Trace` from our example:
 <!-- Tripy: DOC: OMIT Start -->
 ```py
 # doc: no-print-locals
-from tripy.frontend.trace import Trace
+from nvtripy.frontend.trace import Trace
 # Output has been eval'd already, so we'll construct a new one
 new_out = tp.tanh(inp)
 trace = Trace([new_out])
@@ -169,7 +169,7 @@ but this is good enough for a conceptual understanding):
 
 ```py
 def to_flat_ir(self, inputs, outputs):
-    from tripy.flat_ir.ops import TanhOp
+    from nvtripy.flat_ir.ops import TanhOp
 
     TanhOp.build(inputs, outputs)
 ```
@@ -178,9 +178,9 @@ Wait a second - what's happening here? The function doesn't return anything; in
 anything at all!
 
 The way this works is as follows: when we call `to_flat_ir()` we provide input and output
-[`FlatIRTensor`](source:/tripy/flat_ir/tensor.py)s. `to_flat_ir()` is responsible for generating a
+[`FlatIRTensor`](source:/nvtripy/flat_ir/tensor.py)s. `to_flat_ir()` is responsible for generating a
 subgraph of `FlatIR` operations that bind to these inputs and outputs. The
-[`BaseFlatIROp` build function](source:/tripy/flat_ir/ops/base.py) updates the producer of the output tensors,
+[`BaseFlatIROp` build function](source:/nvtripy/flat_ir/ops/base.py) updates the producer of the output tensors,
 meaning that *just building a `FlatIR` operation is enough to add it to the subgraph*. Once this binding
 is done, we take the resulting subgraph and inline it into the `FlatIR`, remapping the I/O tensors to those
 that already exist in the `FlatIR`.
@@ -203,7 +203,7 @@ Our final translation step is to go from `FlatIR` into MLIR.
 Similar to `Trace` operations, `FlatIR` operations implement `to_mlir()` which generates MLIR operations.
 Unlike `Trace` operations, this is always a 1:1 mapping.
 
-Here's a snippet for how [`tanh()` is implemented](source:/tripy/flat_ir/ops/tanh.py):
+Here's a snippet for how [`tanh()` is implemented](source:/nvtripy/flat_ir/ops/tanh.py):
 ```py
 def to_mlir(self, operands):
     return [stablehlo.TanhOp(*operands)]
@@ -227,4 +227,4 @@ Once we have the complete MLIR representation, we then compile it to an executab
 
 Finally, we use the MLIR-TRT executor to launch the executable and retrieve the output tensors.
 The executable returns [`memref`s](https://mlir.llvm.org/docs/Dialects/MemRef/) which we then
-wrap in Tripy frontend {class}`tripy.Tensor`s.
+wrap in Tripy frontend {class}`nvtripy.Tensor`s.

tripy/docs/post0_developer_guides/debugging.md

@@ -29,7 +29,7 @@ Alternatively, you can use [LLDB](https://lldb.llvm.org/) if you launch the cont
 ```bash
 docker run --gpus all --cap-add=SYS_PTRACE \
     --security-opt seccomp=unconfined --security-opt apparmor=unconfined \
-    -p 8080:8080 -v $(pwd):/tripy/ -it --rm tripy:latest
+    -p 8080:8080 -v $(pwd):/nvtripy/ -it --rm nvtripy:latest
 ```
 <!-- Tripy: DOC: NO_EVAL End -->