tvm-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [incubator-tvm] mbaret commented on a change in pull request #5479: [Relay-TFLite] FP32 and Quantized Object Detection Model
Date Wed, 06 May 2020 16:41:56 GMT

mbaret commented on a change in pull request #5479:
URL: https://github.com/apache/incubator-tvm/pull/5479#discussion_r420935075



##########
File path: python/tvm/relay/frontend/tflite.py
##########
@@ -320,6 +321,45 @@ def dequantize(self, expr, tensor):
                                          input_zero_point=tensor.qnn_params['zero_point'])
         return dequantized
 
+
+    def convert_qnn_fused_activation_function(self, expr, fused_activation_fn,
+                                              scale, zero_point, dtype):
+        """Convert TFLite fused activation function. The expr is an input quantized tensor
with
+        scale and zero point """

Review comment:
       It's more from a position of having the history clearly reflect when features were
added. Supporting fused qnn functions is needed for this model, but it's also necessary for
a wide range of other models so I'd view this as a feature not specific to object detection.




----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
users@infra.apache.org



Mime
View raw message