Skip to content

whisper model to onnx exception #20

@kendiyang

Description

@kendiyang

1.optimum-cli export onnx --model openai/whisper-small onnx_model/ --monolith
2.optimum-cli onnxruntime quantize --onnx_model onnx_model/ --arm64 -o quauntized_model/

E/OrtWhisper.kt: Error in getTranscription: Error code - ORT_NOT_IMPLEMENTED - message: Could not find an implementation for ConvInteger(10) node with name '/model/encoder/conv1/Conv_quant'
E/flutter ( 6184): [ERROR:flutter/runtime/dart_vm_initializer.cc(40)] Unhandled Exception: Exception: Transcript returned from platform code is null
E/flutter ( 6184): #0 WhisperNative._getTranscriptPlatformChannel (package:fonnx/models/whisper/whisper_native.dart:55:7)
E/flutter ( 6184):
E/flutter ( 6184): #1 SttService._whisperInferenceLoop.doIt (package:fonnx_example/stt_service.dart:270:23)
E/flutter ( 6184):
E/flutter ( 6184): #2 SttService._whisperInferenceLoop (package:fonnx_example/stt_service.dart:290:5)
E/flutter ( 6184):
E/flutter ( 6184):
Application finished.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions