代码拉取完成，页面将自动刷新
| name | about | labels |
|---|---|---|
| Bug Report | Use this template for reporting a bug | kind/bug |
DynamicRNN算子在910B环境上，当前在调用nn.LSTM ,且输入input=Tensor(np.ones([1, 1, 10])时，出现RuntimeError: Launch kernel failed, name:Default/DynamicRNN 问题
Ascend/GPU/CPU) / 硬件环境:Please delete the backend not involved / 请删除不涉及的后端:
/device ascend 910B
-- MindSpore version (e.g., 1.7.0.Bxxx) : 在910B上一直存在此问题
-- Python version (e.g., Python 3.7.5) :python3.9
-- OS platform and distribution (e.g., Linux Ubuntu 16.04):
-- GCC/Compiler version (if compiled from source):
PyNative/Graph):Please delete the mode not involved / 请删除不涉及的模式:
/mode pynative
/mode graph
import numpy as np
import mindspore.nn as nn
import mindspore
from mindspore import Tensor, context
input1 = Tensor(np.ones([1, 1, 10]).astype(np.float32))
lstm = nn.LSTM(input_size=10, hidden_size=20, num_layers=2,has_bias=True, batch_first=True, bidirectional=False)
h0 = Tensor(np.ones([1 * 2, 1, 20]).astype(np.float32))
c0 = Tensor(np.ones([1 * 2, 1, 20]).astype(np.float32))
ouput_1, output2 =lstm(input1, (h0, c0))
print(output2)
[ERROR] DEVICE(2571995,ffff2a0f0ff0,python):2024-04-26-14:34:45.626.777 [mindspore/ccsrc/plugin/device/ascend/hal/hardware/ge_kernel_executor.cc:951] LaunchKernel] Launch kernel failed, kernel full name: Default/DynamicRNN-op1
[INFO] PROFILING(2571995,python):2024-04-26-14:34:45.631.228 [msprofiler_impl.cpp:156] >>> (tid:2572757) ProfNotifySetDevice called, is open: 1, devId: 5
[CRITICAL] DEVICE(2571995,ffff2a0f0ff0,python):2024-04-26-14:34:45.634.878 [mindspore/ccsrc/runtime/pynative/op_runner.cc:623] LaunchKernels] Launch kernel failed, name:Default/DynamicRNN-op1
Traceback (most recent call last):
File "/home/miniconda3/envs/feature_39/lib/python3.9/site-packages/mindspore/nn/cell.py", line 712, in __call__
_pynative_executor.end_graph(self, output, *args, **kwargs)
File "/home/miniconda3/envs/feature_39/lib/python3.9/site-packages/mindspore/common/api.py", line 1302, in end_graph
self._executor.end_graph(obj, output, *args, *(kwargs.values()))
RuntimeError: Launch kernel failed, name:Default/DynamicRNN-op0
----------------------------------------------------
- C++ Call Stack: (For framework developers)
----------------------------------------------------
mindspore/ccsrc/runtime/pynative/op_runner.cc:623 LaunchKernels
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home/jenkins0/workspace/TDT_deployment/test_lstm1.py", line 11, in <module>
ouput_1, output2 =lstm(input1, (h0, c0))
File "/home/miniconda3/envs/feature_39/lib/python3.9/site-packages/mindspore/nn/cell.py", line 714, in __call__
_pynative_executor.clear_res()
File "/home/miniconda3/envs/feature_39/lib/python3.9/site-packages/mindspore/common/api.py", line 1345, in clear_res
return self._executor.clear_res()
RuntimeError: Launch kernel failed, name:Default/DynamicRNN-op1
----------------------------------------------------
- C++ Call Stack: (For framework developers)
----------------------------------------------------
mindspore/ccsrc/runtime/pynative/op_runner.cc:623 LaunchKernels
此处可能存在不合适展示的内容，页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容，可点击提交进行申诉，我们将尽快为您处理。
感谢您的提问，您可以评论//mindspore-assistant更快获取帮助：
cann算子问题，单： DTS2024042924488
登录 后才可以发表评论