pytorch/ort

Issues

• python -m torch_ort.configure fail

  #177 opened 10 months ago by kimkeithvn
  2

• Why should I be forced to have a CUDA or ROCm machine when wanting to run OpenVino on Intel?

  #175 opened a year ago by rmast
  2

• Does it support TensorRT backend?

  #185 opened 7 months ago by jinluyang
  0

• Warning: Checker does not support models with experimental ops: ATen

  #150 opened 10 months ago by msaroufim
  2

• topKgate loss issues

  #178 opened 10 months ago by powermano
  0

• Clarify installation requirements for CUDA vs ROCm

  #171 opened a year ago by natke
  1

• Compatibility between ORTModule and DeepSpeed

  #108 opened a year ago by JingyaHuang
  6

• torch-ort cannot be installed on windows: onnxruntime-training not found

  #119 opened 2 years ago by alpatots
  5

• RuntimeError: Error in execution: At least one output should be requested.

  #145 opened 2 years ago by StochasticRomanAgeev
  0

• [torch-ort-infer] Aten fallback doesn't work

  #139 opened 2 years ago by saipj
  6

• CUDA error cudaErrorInvalidConfiguration:invalid configuration argument

  #116 opened 2 years ago by Suparjie
  3

• Will there be new nightly builds with version 1.13.0.dev?

  #133 opened 2 years ago by jambayk
  2

• What does ORT stands for?

  #130 opened 2 years ago by oneengineer
  1

• Where operator export error when performing fp16 quantization

  #117 opened 2 years ago by prathikr
  0

• Turn off fallback to torch by default

  #114 opened 2 years ago by justinchuby
  3

• `python -m torch_ort.configure` fails with protobuf errors

  #115 opened 2 years ago by justinchuby
  1

• [Question] PyTorch 1.11

  #112 opened 2 years ago by austinmw
  2

• Running ORTModule with other EPs from ORT

  #78 opened 2 years ago by chethanpk
  11

• Seg fault while training model with maxpool op

  #107 opened 2 years ago by chethanpk
  0

• ONNXRuntimeError after enabled fp16 mixed precision training

  #104 opened 2 years ago by JingyaHuang
  8

• MaxPool op resolved as Aten OP

  #105 opened 2 years ago by chethanpk
  6

• no speedup using ort

  #103 opened 2 years ago by housebaby
  4

• MKL_THREADING_LAYER=INTEL is incompatible with libgomp.so.1 library

  #102 opened 2 years ago by TaehwanKwon
  0

• Pytorch lightning

  #101 opened 2 years ago by javier-alvarez
  2

• Torch unable to use ORT because of opset version issue

  #100 opened 2 years ago by chethanpk
  2

• tests/bert_for_sequence_classification.py reports "This is an invalid model"

  #99 opened 2 years ago by xysmlx
  4

• Question about supported device

  #98 opened 3 years ago by mengniwang95
  2

• Fallback not kicking in with non-contiguous tensors

  #92 opened 3 years ago by natke
  4

• torch_ort configure fails

  #94 opened 3 years ago by wschin
  3

• ORT support for CUDA 11.0

  #93 opened 3 years ago by karthikrangasai
  2

• Sample does not work with torch-ort:1.8.1

  #68 opened 3 years ago by natke
  1

• Add `ninja` in requirements.txt

  #30 opened 3 years ago by ryotatomioka
  2

• RecursionError: maximum recursion depth exceeded in comparison

  #34 opened 3 years ago by DuinoDu
  7

• Lack of speed improvement when using custom GPT model with ORT

  #62 opened 3 years ago by SeanNaren
  6

• AttributeError: 'ORTModule' object has no attribute 'resize_token_embeddings'

  #53 opened 3 years ago by bmedishe
  6

• PyTorch Lightning Integration

  #56 opened 3 years ago by SeanNaren
  6

• Should / How can we deal with unsupported operator warning?

  #57 opened 3 years ago by richarddwang
  2