Common Issues

The list below outlines common issues that may arise while using LaunchX. Please feel free to contact us at any time if you encounter any undefined exception cases or require technical support.

Upload

FrameworkException

This occurs when the uploaded model file is not supported by the framework. Please check the scope of support.

Note

This doesn’t seem to be a supported framework model. Compatibility range of Conversion | Compatibility range of Benchmark

InputShapeLengthException

This occurs when the input shape length is not 4.

Note

It seems that the input shape of the model is not compatible. Compatibility range of Conversion | Compatibility range of Benchmark

InspectFail

This occurs when there is an issue with the model file.

Note

It seems to be an incompatible file. Compatibility range of Conversion | Compatibility range of Benchmark

KerasError

This occurs when there is an issue with the Keras model file.

Note

It seems to be an incompatible file.

MakeInputsError

This occurs when there is no subgraph in the TensorFlow model file.

Note

Please check if the model file contains a subgraph.

ModelFileInspectException

If there are issues during the process of inspecting the uploaded model file.

Note

It may not be a model file, or it may be a model file from an unsupported. Compatibility range of Conversion | Compatibility range of Benchmark

ModelTypeError

This occurs when there is an issue with the model file.

Note

It seems to be an incompatible file. Compatibility range of Conversion | Compatibility range of Benchmark

MultiLayerException

This occurs when the input layer is multi-layered.

Note

Multi-input layer model is not supported. Compatibility range of Conversion | Compatibility range of Benchmark

OpenVINOError

This occurs when there is an issue with the OpenVINO model file.

Note

Please check if the first layer of the OpenVINO XML file contains ‘element_type’ and ‘shape’ information.

TensorRTDeserializerError

This occurs when there is an issue with the TensorRT model file.

Note

The model file cannot be deserialized. Please check the model file.

TooLargeBatchModelException

This occurs when the batch size of the model being uploaded exceeds 4.

Note

The maximum batch size is 4.

TooLargeChannelModelException

This occurs when the number of channels in the model being uploaded exceeds 3.

Note

Channel exceeds the supported range. We only support channels 1 to 4. Compatibility range of Conversion | Compatibility range of Benchmark

Benchmark

BenchmarkFailedException

This occurs during benchmark execution. Detailed investigation through log analysis is required to determine the specific cause.

Note

If you want to know the specific cause, please request a technical team investigation. [Contact support]

DynamicBatchModelException

This occurs when the uploaded model is a dynamic batch model. Currently, only static batch models are supported.

Note

Dynamic batch model is not supported for benchmark. Read related content

DynamicInputShapeModelException

This occurs when the uploaded model has dynamic width and height, whereas only static input sizes are currently supported.

Note

Dynamic input shape model is not supported for benchmark. Read related content

StartBenchmarkFail

This occurs when there is a failure to execute a benchmark in the device farm.

Note

A temporary issue occurred during the execution process. Please try again later. If the problem persists, please contact us. [Contact support]

Convert

ConvertFailedException

This occurs during the conversion process. Detailed investigation through log analysis is required to determine the specific cause.

Note

If you want to know the specific cause, please request a technical team investigation. [Contact support]

NeedDeviceException

This occurs when no device is specified during the conversion process to TensorRT. Since the conversion to TensorRT runs on a device, you need to specify both the device and Jetpack versions.

Note

To convert to TensorRT, a device must be specified. Read related content

NeedInputShapeException

This occurs during the conversion of a model with dynamic shape when the input shape value of the model is missing.

Note

Please input the ‘input shape’ to convert.

Common cases

DeviceFarmUploadError

This occurs during the process of uploading a model to the Device farm.

Note

There was a temporary issue during the file upload process. Please try again later. If the problem persists, please contact us. [Contact support]

DeviceNotMatchedWithFrameworkException

This occurs when the target framework and target device do not match.

Note

{target_device_name} is not supported for this models target framework {target_framework}. Compatibility range of Conversion | Compatibility range of Benchmark

InternalException

In case of a temporary error

Note

A temporary error has occurred. Please try again later. If the problem persists, please contact us. [Contact support]

LatestJobIsNotFinished

This occurs when attempting to execute a new task while there are ongoing convert or benchmark operations.

Note

There is a current ongoing task. Please try again after this task completes.

NoAvailableDevice

This occurs when there are no available devices in the NetsPresso device farm.

Note

There is not available device. It is a temporary issue, please try again later. If the problem persists, please contact us. [Contact support]

NoCredit

This occurs when there are not enough credits available to execute the task.

Note

Insufficient credits. Please contact us to recharge your credits. [Contact support]

NoModelUuidException

This occurs when requesting a model_uuid that does not exist.

Note

The model UUID does not exist. Please check the model UUID and try again.

NotSupportedDeviceException

This occurs when the input target device is not supported by the selected framework.

Note

{target_device} is not available device for this model. Compatibility range of Conversion | Compatibility range of Benchmark

AccountException

If there’s an issue during the server’s communication relay process, a related message will be displayed on the service. For undefined exception cases, log investigation is necessary to identify the detailed cause.

Note

If you want to know the specific cause, please request a technical team investigation. [Contact support]

Dataset file

Prepare dataset file for INT8 quantization with LaunchX

DatasetFileBrokenException

This occurs when there is a failure to load the uploaded NumPy file for quantization calibration. There is a high possibility that there is an error in the NumPy file.

Note

Please check the NumPy file.

DatasetFileSuffixException

This occurs when the extension of the uploaded file is not .npy for quantization calibration.

Note

Please upload a .npy file extension. Read related content

InvalidTypeException

This occurs when the dataset file is not a readable NumPy array.

Note

The dataset file has an invalid type NumPy array. Read related content

NumpyArrayHasWrongShapeException

This occurs when the shape of the npy file for quantization calibration does not match the input shape of the model being converted.

Note

The NumPy file has a different input shape than the model’s. Read related content