|
关于“推理经验-Inference Experience”类别
|
|
0
|
31
|
June 5, 2025
|
|
MindSpore推理结果不稳定且出现报错[ERROR] RUNTIME: aicpu kernel execute failed, fault kernel_name=GetNext.
|
|
1
|
64
|
April 18, 2026
|
|
模型转换失败Convert failed. Ret: Common error code问题分析
|
|
0
|
37
|
April 18, 2026
|
|
MindSpore推理结果不稳定[ERROR] RUNTIME: aicpu kernel execute failed, fault kernel_name=GetNext.及解决
|
|
0
|
36
|
April 18, 2026
|
|
使用mindspore_hub将Hugging Face模型转换后，生成图像出现色偏的分析和解决
|
|
0
|
40
|
March 1, 2026
|
|
MindSpore模型推理时输出张量形状异常ValueError: shapes (32,10) and (32,10,1,1) not aligned
|
|
1
|
51
|
March 1, 2026
|
|
使用mindspore_hub将Hugging Face的模型转换后，推理生成图像出现色偏
|
|
1
|
68
|
March 1, 2026
|
|
将.ckpt模型转换为.om格式，运行转换脚本后报错Failed to allocate memory for tensor of shape [1, 64, 160, 160]
|
|
0
|
50
|
February 11, 2026
|
|
模型推理时TransData算子耗时多
|
|
1
|
61
|
January 30, 2026
|
|
模型推理时TransData算子耗时优化
|
|
0
|
68
|
January 30, 2026
|
|
MindSpore Lite输入输出数据免拷贝
|
|
0
|
44
|
January 9, 2026
|
|
MindSpore Lite问题定位——Dump图
|
|
0
|
55
|
January 7, 2026
|
|
利用昇思MindSpore图模式融合特性优化模型推理
|
|
0
|
67
|
December 10, 2025
|
|
使用MindSpore Lite端侧模型转换工具将YOLOv8.onnx转为.ms报错Convert failed. Ret: Common error code.
|
|
1
|
133
|
October 31, 2025
|
|
使用mindsporelite推理，出现data size not equal 错误，tensor size 0
|
|
2
|
45
|
October 30, 2025
|
|
qwen1.5-0.5b推理报错Launch kernel failed, kernel full name: Default/ScatterNdUpdate-op0
|
|
0
|
21
|
October 21, 2025
|
|
Mindspore在自回归推理时的精度对齐设置
|
|
0
|
23
|
September 27, 2025
|
|
qwen1.5_1.8B推理出现回答混乱问题及解决
|
|
0
|
26
|
September 25, 2025
|
|
MindSpore大模型在线推理速度慢及解决方案
|
|
0
|
55
|
September 24, 2025
|
|
MindSpore与CANN的协同优化：实现极致性能的深度学习训练与推理
|
|
0
|
187
|
September 24, 2025
|
|
MindSpore Lite模型加载报错RuntimeError: build from file failed! Error is Common error code.
|
|
0
|
62
|
September 9, 2025
|
|
昇腾910上yolov5推理比较慢
|
|
1
|
57
|
September 6, 2025
|
|
昇思MindSpore开源社区上线智谱GLM-4.5与GLM-4.5-Air大模型
|
|
1
|
192
|
August 18, 2025
|
|
使用.om格式模型结合gradio框架进行推理出现模型执行错误
|
|
0
|
35
|
September 1, 2025
|
|
mindformers推理qwen2.5-72b报显存不足及解决
|
|
0
|
68
|
September 1, 2025
|
|
LLama2_7b推理报错ValueError: For BatchMatMul, inputs shape cannot be broadcast on CPU/GPU.和解决
|
|
0
|
40
|
August 18, 2025
|
|
MindSpore论坛报错活动第三十期-报错：使用MindSpore将.ckpt转.air再转.om出现AttributeError: 'AclLiteModel' object has no attribute '_is_destroye
|
|
0
|
47
|
August 15, 2025
|
|
训练过程中推理精度不变问题定位思路
|
|
0
|
76
|
August 1, 2025
|
|
MindSpore网络推理时使用Matmul矩阵乘法算子计算速度较慢
|
|
0
|
47
|
August 1, 2025
|
|
mindspore推理报错NameError:The name 'LTM' is not defined, or not supported in graph mode.及解决
|
|
0
|
34
|
August 1, 2025
|