mindspore-ai/mindspore

Issues

• Batch_size不一致导致相同数据在同一模型上输出不一致

  #328 opened 10 days ago by Let-the-bullet-fly
  0

• 在 GRAPH_MODE 下 construct 中调用 Parameter.set_data 引发 RuntimeError

  #327 opened 2 months ago by huadaox
  1

• 什么时候mindspore可以支持一下AMD显卡

  #326 opened 2 months ago by MuDaMuDaMuDa2
  1

• Encounter ImportError When Runing Code in Jupyter: libmindspore_backend.so: undefined symbol: dnnl_threadpool_interop_stream_create

  #325 opened 3 months ago by 641217688tly
  0

• DeepSeek V3适配

  #324 opened 3 months ago by xinxin2015
  0

• has no attribute 'device_context' in [2.5.0.dev20241208]

  #321 opened 3 months ago by suzhelearning
  1

• How to convert mindspore.common._stub_tensor.StubTensor to mindspore.common.tensor.Tensor?

  #322 opened 3 months ago by psychedelicosisyphus
  1

• AttributeError: module 'mindspore.dataset.vision' has no attribute 'RandomCropDecodeResize'

  #323 opened 3 months ago by bebopoy
  0

• 在4090上安装失败 No module named 'mindspore._c_expression'

  #320 opened 4 months ago by Ethereal-sakura
  2

• CPU算子bug

  #315 opened 5 months ago by ResDream
  1

• 如何降低gpu的初始化时间

  #318 opened 4 months ago by keke444
  1

• 华为910B显卡上跑MindYolo训练报错 EZ9999: Inner Error!

  #319 opened 4 months ago by MuDaMuDaMuDa2
  2

• The slack link in readme is not available

  #317 opened 5 months ago by longtsing
  1

• Mindspore 是否支持Elastic Training

  #316 opened 5 months ago by Ind1x1
  1

• MindSpore cannot convert the data types of model parameters

  #313 opened 5 months ago by PhyllisJi
  1

• 910B机器

  #314 opened 5 months ago by Jayc-Z
  1

• [ascend910b] does not support opType [RmsNorm]

  #312 opened 5 months ago by Jayc-Z
  1

• 如何在910B的麒麟操作系统上进行多lora的推理部署？

  #311 opened 5 months ago by Jayc-Z
  3

• 如何在910B的麒麟操作系统上进行多lora的推理部署

  #310 opened 5 months ago by Jayc-Z
  0

• AdaptiveAvgPool2D's exception checking gives relevant code in the kernel code that are prone to security issues

  #285 opened a year ago by PhyllisJi
  2

• The same input runs mindspore.ops.reciprocal on both CPU and GPU outputting one as INF and one as 0

  #293 opened 10 months ago by PhyllisJi
  4

• The mindspore preserves the gradients of the layers that are not involved in forward propagation

  #309 opened 5 months ago by PhyllisJi
  3

• mindspore.nn.MaxPool2d throws exception but shows MaxPool3D

  #308 opened 5 months ago by PhyllisJi
  2

• Mindspore.nn.Conv2d lacks checks on convolutional kernel shapes

  #291 opened 10 months ago by PhyllisJi
  2

• fail to remove dataset feature from huggingface (unable to align with DatasetDict.map parameters remove_columns in transformers)

  #307 opened 6 months ago by EdwinWang37
  1

• mindspore.nn.Cell.get_inputs() 返回值为 null

  #306 opened 6 months ago by Hsiayukoo
  0

• 使用c_api 如何获取模型运行内存？

  #305 opened 6 months ago by keke444
  0

• mindspore.ops.reciprocal会导致完全一致的代码在CPU和GPU上的运行结果存在巨大差异

  #282 opened a year ago by PhyllisJi
  1

• lack of output shape checking in mindspore.nn.Conv2d

  #288 opened 10 months ago by PhyllisJi
  2

• mindspore.nn.Conv1d and mindspore.nn.Conv1dtranspose 's shapes are four-dimensional

  #284 opened a year ago by PhyllisJi
  1

• mindspore_lite通过 java 调用 npu 进行推理问题

  #299 opened 8 months ago by 983183947
  1

• 相同 cann环境下进行onnx 转换，在 310B 上执行推理报错

  #300 opened 8 months ago by 983183947
  2

• mindspore训练模型时显存无限制增加

  #302 opened 8 months ago by Veteranback
  0

• mindnlp.core.nn.init.uniform函数生成异常

  #301 opened 8 months ago by ShangJingLi
  2

• Lack of description of the range of values for in_channels and out_channels in the mindspore.nn.Dense documentation

  #289 opened 10 months ago by PhyllisJi
  1

• 通过 mindspore lite，执行 onnx 到 mindir 转换报错

  #295 opened 8 months ago by 983183947
  1

• mindspore lite 中的benchmark报错

  #298 opened 8 months ago by 983183947
  1

• mindspore.nn.Sigmoid() causes the code to run successfully on the CPU but fails on the GPU

  #283 opened a year ago by PhyllisJi
  1

• Wrong parameter name in error message for BatchNorm1D

  #290 opened 10 months ago by PhyllisJi
  1

• 集成TensorRT相关问题

  #287 opened 10 months ago by BaiStone2017
  1

• 官网案例--快速入门，RNN，CycleGAN，ResNet50迁移学习，ResNet50图像分类 ，ViT，ShuffleNet，Pix2Pix，GAN，SSD，和Diffusion在pynative_mode下 报错RunimeError，具体日志信息见附件

  #296 opened 9 months ago by cui0523
  1

• 官网案例--FCN, LSTM+CRF和DCGAN在pynative_mode下 报错[ERROR]KERNEL:Launch] Kernel launch failed,，具体日志信息见附件

  #297 opened 9 months ago by cui0523
  1

• 官网案例--快速入门报错RunimeError: the kernel ArgMaxExt unregistered

  #286 opened 9 months ago by cui0523
  0

• MultiHeadAttention.construct does not accept named args

  #294 opened 10 months ago by Kahsolt
  0

• Request for output_padding Parameter Support in Conv1dTranspose and Conv2dTranspose Layers

  #292 opened 10 months ago by PhyllisJi
  0

• 希望能支持ParameterDIct

  #281 opened 10 months ago by moyanxinxu
  1

• mindspore框架同步问题

  #278 opened a year ago by cqulilujia
  1

• Pynative Train，OOM！

  #280 opened a year ago by Enlion91
  2

• data sink mode does not work

  #279 opened a year ago by Enlion91
  4

• [Bug] mindspore.ops.argsort/sort perform differently between cpu and gpu.

  #277 opened a year ago by Redempt1onzzZZ
  6