int8 quantization is not working for LSTM op in TensorFLow Lite

[Niranjan Yadla]

We are using LSTM op in our model and could verify accuracy using TF, TFLite float32 models.
Please see below Colab for more information.
https://colab.research.google.com/drive/1LIqNXYszOB4eUHFhqnbGLESdZs4Ipu_n?usp=sharing
However when I generate TFLite int8 model using post training quantization and int8 model doesn't work with my test data.
Note: If I use uint8 post training quantization it works fine.
For some reason int8 quantization is not working with LSTM op.

[Geoffrey Martin-Noble]

Hi Niranjan,

Maybe I'm missing something, but it seems like this is not connected to IREE. Maybe you want to direct your question to the TensorFlow and TFLite folks?

Best,

Geoffrey

--
You received this message because you are subscribed to the Google Groups "iree-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to iree-discuss...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/iree-discuss/83490e20-2304-4d2f-ab9f-830a3d7bf7c5n%40googlegroups.com.

[Niranjan Yadla]

Thanks I will post to TF and TFLite folks.

Thanks,

Niranjan