| commit | b3cd60a12f8f1c593a900b61d407eff138c443e5 | [log] [tgz] |
|---|---|---|
| author | Scott Todd <scotttodd@google.com> | Thu Oct 12 09:40:30 2023 -0700 |
| committer | GitHub <noreply@github.com> | Thu Oct 12 09:40:30 2023 -0700 |
| tree | 3ed2dd5a7e2f39fc401c7e393d9bfee04e8b1ea5 | |
| parent | f782069ecde3774553c0c15e1b953fea40b4a7c2 [diff] [blame] |
Add pytorch_jit sample Colab notebook using SHARK-Turbine. (#15146) Progress on https://github.com/openxla/iree/issues/15117 This notebook shows how to use [SHARK-Turbine](https://github.com/nod-ai/SHARK-Turbine) for eager execution within a PyTorch session using IREE and [torch-mlir](https://github.com/llvm/torch-mlir) under the covers. I'm starting simple to get the concepts across, with minimal API usage and a tiny `nn.Module` sourced from https://pytorch.org/docs/stable/notes/modules.html. My expectation is that this notebook will evolve alongside other notebooks (e.g. `pytorch_aot.ipynb`), documentation (https://github.com/openxla/iree/issues/15114), and the SHARK-Turbine project itself. Preview URL for review: https://colab.research.google.com/github/openxla/iree/blob/scotttodd-pytorch-samples-1/samples/colab/pytorch_jit.ipynb skip-ci: no-op
diff --git a/samples/colab/test_notebooks.py b/samples/colab/test_notebooks.py index 984002c..17ca31e 100755 --- a/samples/colab/test_notebooks.py +++ b/samples/colab/test_notebooks.py
@@ -15,6 +15,8 @@ # tflite_runtime requires some deps ("version `GLIBC_2.29' not found") that # samples.Dockerfile does not currently include. "tflite_text_classification.ipynb", + # Requires Python 3.10+ in our Docker image. + "pytorch_jit.ipynb", ] NOTEBOOKS_EXPECTED_TO_FAIL = [