CUDNN_STATUS_BAD_PARAM
cuda
runtime_error
ai_generated
true
运行时错误：cuDNN 错误：调用 cudnnBatchNormalizationForwardTraining 时出现 CUDNN_STATUS_BAD_PARAM
RuntimeError: cuDNN error: CUDNN_STATUS_BAD_PARAM when calling cudnnBatchNormalizationForwardTraining
ID: cuda/cudnn-batch-norm-bad-param
78%修复率
83%置信度
1证据数
2024-01-18首次发现
版本兼容性
| 版本 | 状态 | 引入 | 弃用 | 备注 |
|---|---|---|---|---|
| cuDNN 8.9.5 | active | — | — | — |
| CUDA 12.1 | active | — | — | — |
| PyTorch 2.2.0 | active | — | — | — |
| TensorFlow 2.15 | active | — | — | — |
根因分析
向 cuDNN 批归一化传递了无效参数，例如 epsilon 为零、动量为负值，或输入、缩放和偏置张量的形状不匹配。
English
An invalid parameter was passed to cuDNN batch normalization, such as a zero epsilon value, negative momentum, or mismatched tensor shapes between input, scale, and bias.
官方文档
https://docs.nvidia.com/deeplearning/cudnn/api/index.html#cudnnBatchNormalizationForwardTraining解决方案
-
import torch; bn = torch.nn.BatchNorm2d(64, eps=1e-5, momentum=0.1)
-
torch.backends.cudnn.enabled = False
无效尝试
常见但无效的做法:
-
95% 失败
Batch size does not affect parameter validation; the bad parameter error occurs regardless of batch size.
-
90% 失败
Deterministic mode only affects algorithm selection, not parameter checking; the bad parameter is still detected.