{
  "id": 215317,
  "title": "Pytorch TPU Model Inference in GPU got error.",
  "url": "/competitions/cassava-leaf-disease-classification/discussion/215317",
  "author_name": "Md. Masud Rana",
  "post_date": "2021-01-29T12:08:26.569000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>When I tried to inference the TPU model in GPU, it's produced this error.</p>\n<p><code>RuntimeError: Could not run 'aten::empty_strided' with arguments from the 'XLA' backend. 'aten::empty_strided' is only available for these backends: [CPU, CUDA, BackendSelect, Autograd, Profiler, Tracer].</code></p>\n<p>Anyone facing this error. If yes, how did you solve this?</p>\n<p>Someone can share the TPU PyTorch model inference. </p>",
  "messages": [
    {
      "id": 1175926,
      "postDate": "2021-01-29T12:08:26.570Z",
      "content": "<p>When I tried to inference the TPU model in GPU, it's produced this error.</p>\n<p><code>RuntimeError: Could not run 'aten::empty_strided' with arguments from the 'XLA' backend. 'aten::empty_strided' is only available for these backends: [CPU, CUDA, BackendSelect, Autograd, Profiler, Tracer].</code></p>\n<p>Anyone facing this error. If yes, how did you solve this?</p>\n<p>Someone can share the TPU PyTorch model inference. </p>",
      "rawMarkdown": "When I tried to inference the TPU model in GPU, it's produced this error.\n\n`RuntimeError: Could not run 'aten::empty_strided' with arguments from the 'XLA' backend. 'aten::empty_strided' is only available for these backends: [CPU, CUDA, BackendSelect, Autograd, Profiler, Tracer].`\n\nAnyone facing this error. If yes, how did you solve this?\n\nSomeone can share the TPU PyTorch model inference. ",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1175926": "When I tried to inference the TPU model in GPU, it's produced this error.\n\n`RuntimeError: Could not run 'aten::empty_strided' with arguments from the 'XLA' backend. 'aten::empty_strided' is only available for these backends: [CPU, CUDA, BackendSelect, Autograd, Profiler, Tracer].`\n\nAnyone facing this error. If yes, how did you solve this?\n\nSomeone can share the TPU PyTorch model inference. "
  }
}