1 系统环境
- 硬件环境(Ascend/GPU/CPU): Ascend
- MindSpore版本: mindspore=2.6.0
- 执行模式（PyNative/ Graph）: 不限
- Python版本: Python=3.9
- 操作系统平台: Linux
2 报错信息
2.1 问题描述
代码1和代码2执行都能运行出正确结果，但是使用模型转换工具converter_lite转换时，代码1能够转换成功，代码2会转换失败。
2.2 脚本信息
- 代码1
def construct(self, t):
t = t.reshape(self.iASampleLen, t.size//self.iASampleLen)
t = self.slice(t, [0, self.iHeaderCPLX], [self.iASampleLen, self.iRcLength*self.iChannelNum])
t = t.reshape(self.line, self.iRcLength)
t = self.slice(t, [0,0], [self.line, self.iRSampleLen])
t = ops.cat((t, self.zero), 1)
t3 = t[]
for i in range(self.line):
t1 = self.fft(t[i])
t1 = self.mul(t1, self.pcGSMRcRef[i])
t1 = self.ifft(t1)
t3.append(self.slice(t1, [0,], [self.iRSampleLen,]))
t3 = ops.cat(t3)
return t3
- 代码2
def construct(self, t):
t = t.reshape(self.iASampleLen, t.size//self.iASampleLen)
t = self.slice(t, [0, self.iHeaderCPLX], [self.iASampleLen, self.iRcLength*self.iChannelNum])
t = t.reshape(self.line, self.iRcLength)
t = self.slice(t, [0,0], [self.line, self.iRSampleLen])
t = ops.cat((t, self.zero), 1)
t2 = t
for i in range(self.line):
t1 = self.fft(t[i])
t1 = self.mul(t1, self.pcGSMRcRef[i])
t2[i] = self.ifft(t1)
t = self.slice(t2, [0,0], [self.line, self.iRSampleLen])
t = t.reshape(t.size)
return t
2.3 报错信息
[ERROR] LITE(43234,7flcle7cbf40, converter_lite):2026-01-27-17:21:35.979.843 [mindspore/lite/tools/converter/legacy_optimizer/graph/dropout_node_remove_pass.cc:35] IsolateDropoutNode] Only support node who hasno more than one input and two output
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite):2026-01-27-17:21:35.979.883 [mindspore/lite/tools/converter/legacy_optimizer/graph/dropout_node_remove_pass.cc:106] Run] IsolateDropout Node failed, subGraph: , node: Default/TensorScatterUpdate-opO, error: -1
[ERROR] LITE(43234,7f1cle7cbf40,converter_lite) :2026-01-27-17:21:35.979.889 [mindspore/lite/tools/converter/optimizer.cc:78] Run] Run GraphPass failed
[ERROR] LITE(43234,7f1cle7cbf40,converter_lite):2026-01-27-17:21:35.979.894 [mindspore/lite/tools/converter/graphdef_transform.cc:93] Transform] Run unused_op_remove_optimizer graphPasses Failed
[ERROR] LITE(43234,7f1c1e7cbf40, converter_lite):2026-01-27-17:21:35.979.904 [mindspore/lite/tools/converter/converter_metagraph.cc:102] Build] Transform meta graph failed -1
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite):2026-01-27-17:21:35.983.657 [mindspore/lite/tools/converter/converter.cc:1316] SaveGraph] Convert to meta graph failed
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite):2026-01-27-17:21:35.983.703 [mindspore/lite/tools/converter/converter.cc:1269] HandleGraphCommon] Savegraph failed: -1 Common error code,
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite) :2026-01-27-17:21:35.987.997 [mindspore/lite/tools/converter/converter.cc:1136] Convert] Handle graph failed: -1 Common error code.
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite):2026-01-27-17:21:35.988.045 [mindspore/lite/tools/converter/converter.cc:1401] RunConverter] Convert model failed
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite) :2026-01-27-17:21:35.988.051 [mindspore/lite/tools/converter/converter_context.h:60] PrintOps]
[ERROR] LITE(43234,7f1cle7cbf40, converter lite):2026-01-27-17:21:35.988.056 [mindspore/lite/tools/converter/converter_context.h:61] Printops] UNSUPPORTED OP LIST:
[ERROR] LITE(43234,7f1cle7cbf40, converter_lite):2026-01-27-17:21:35.988.060 [mindspore/lite/tools/converter/converter_context.h:63] Printops] FMКTYPЕ: , ОР TYPЕ: ТensorScatterUpdate
[ERROR] LITE(43234,7f1c1e7cbf40, converter_lite) :2026-01-27-17:21:35.988.065 [mindspore/lite/tools/converter/converter_context.h:65] PrintOps]
[ERROR] LITE(43234,7f1c1e7cbf40, converter_lite):2026-01-27-17:21:35.988.073 [mindspore/lite/tools/converter/cxx_api/converter.cc:374] Convert] Convert modelfailed, ret=Common error c
ode.
ERROR [mindspore/lite/tools/converter/converter_lite/main.cc:104] main] Convert failed. Ret: Common error code.
Convert failed. Ret: Common error code.
3 根因分析
- 此处由用户填写。
4 解决方案
- 此处由用户填写；
- 包含文字方案和最终脚本代码 请将正确的脚本打包并上传附件。