|
|
发表于 2024-1-26 15:33
来自手机
|
显示全部楼层
pdvc 发表于 2024-1-25 23:20
ensemble是整合所有模型出结果，负载最大，没啥必要。
不带后缀是标准版，负载居中，然后是lite版。v2是 ...
模型在这里下载的：https://github.com/AmusementClub/vs-mlrt/releases/tag/external-models。我用v2版模型，训练时必然报错，不带后缀的就没问题，看了半天日志也没看出个所以然
[01/26/2024-15:05:25] [I] Start parsing network model
[01/26/2024-15:05:25] [I] [TRT] ----------------------------------------------------------------
[01/26/2024-15:05:25] [I] [TRT] Input filename: C:/Program Files (x86)/SVP 4/rife\models\rife_v2\rife_v4.9.onnx
[01/26/2024-15:05:25] [I] [TRT] ONNX IR version: 0.0.8
[01/26/2024-15:05:25] [I] [TRT] Opset version: 16
[01/26/2024-15:05:25] [I] [TRT] Producer name: pytorch
[01/26/2024-15:05:25] [I] [TRT] Producer version: 2.2.0
[01/26/2024-15:05:25] [I] [TRT] Domain:
[01/26/2024-15:05:25] [I] [TRT] Model version: 0
[01/26/2024-15:05:25] [I] [TRT] Doc string:
[01/26/2024-15:05:25] [I] [TRT] ----------------------------------------------------------------
[01/26/2024-15:05:25] [W] [TRT] onnx2trt_utils.cpp:377: Your ONNX model has been generated with INT64 weights, while TensorRT does not natively support INT64. Attempting to cast down to INT32.
[01/26/2024-15:05:25] [W] [TRT] onnx2trt_utils.cpp:403: One or more weights outside the range of INT32 was clamped
[01/26/2024-15:05:25] [W] [TRT] onnx2trt_utils.cpp:403: One or more weights outside the range of INT32 was clamped
[01/26/2024-15:05:25] [W] [TRT] onnx2trt_utils.cpp:403: One or more weights outside the range of INT32 was clamped
[01/26/2024-15:05:25] [W] [TRT] onnx2trt_utils.cpp:403: One or more weights outside the range of INT32 was clamped
[01/26/2024-15:05:26] [W] [TRT] onnx2trt_utils.cpp:403: One or more weights outside the range of INT32 was clamped
[01/26/2024-15:05:26] [I] Finish parsing network model
[01/26/2024-15:05:26] [W] Could not read timing cache from: C:\Users\hhhh\AppData\Roaming\SVP4\cache\Program Files (x86)/SVP 4/rife\models\rife_v2\rife_v4.9.onnx.min64x64_opt1920x1088_max2560x1440_fp16_trt-8502_cudnn_I-fp16_O-fp16_NVIDIA-GeForce-RTX-4090_65612893.engine.cache. A new timing cache will be generated and written.
[01/26/2024-15:05:26] [I] [TRT] [MemUsageChange] Init cuDNN: CPU +734, GPU +10, now: CPU 10048, GPU 1697 (MiB)
[01/26/2024-15:05:26] [I] [TRT] Global timing cache in use. Profiling results in this builder pass will be stored.
[01/26/2024-15:05:26] [W] [TRT] Using kFASTER_DYNAMIC_SHAPES_0805 preview feature.
[01/26/2024-15:05:55] [E] Error[2]: [deconvolutionV2Builder.cpp::nvinfer1::builder::CaskDeconvolutionBuilderBase<class nvinfer1::builder::CaskDeconvolutionV2Builder,class nvinfer1::rt::task::CaskDeconvolutionV2Runner,-2147483603>::createDeconvolution::759] Error Code 2: Internal Error (Assertion isOpConsistent(deconvolution.get()) failed. Cask deconvolution isConsistent check failed.)
[01/26/2024-15:05:55] [E] Error[2]: [builder.cpp::nvinfer1::builder::Builder::buildSerializedNetwork::751] Error Code 2: Internal Error (Assertion engine != nullptr failed. )
[01/26/2024-15:05:55] [E] Engine could not be created from network
[01/26/2024-15:05:55] [E] Building engine failed
[01/26/2024-15:05:55] [E] Failed to create engine from model or file.
[01/26/2024-15:05:55] [E] Engine set up failed
|
|
楼主: kongzheng3333
310112100042806
