unsupported ops

[znsoftm]

There's 5 ops are not supported yet =========== top_k =========== =========== fake_quantize_moving_average_abs_max =========== =========== ctc_align =========== =========== lstm =========== =========== fake_channel_wise_dequantize_max_abs ===========

There's 2 ops are not supported yet =========== fake_channel_wise_dequantize_max_abs =========== =========== fake_quantize_moving_average_abs_max ===========

[Channingss]

@znsoftm

• paddle2onnx don't support quanted model.
• if your model is contruct by paddlepaddle>2.0.0rc1，lstm is supported. Old version of paddlepaddle is not supported.
• ctc_align is not support now.

your model is from paddleocr?

[znsoftm]

@znsoftm

• paddle2onnx don't support quanted model.
• if your model is contruct by paddlepaddle>2.0.0rc1，lstm is supported. Old version of paddlepaddle is not supported.
• ctc_align is not support now.

your model is from paddleocr?

Sure, I'm working on inferencing paddleOCR with onnx. So when will you guys support them?

[Channingss]

@znsoftm can you share your paddle model(saved) to us, ctc_align is a operator with LodTensor Inputs, which may support by onnx when the inputs of ctc_align in your model is equal length sequence, we need to check it.

[znsoftm]

The requirement is to transfer paddleocr mobile model V2.0 to onnx without input limitation.

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.