projects / platform / upstream / tvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 5248c14) | patch
[TFLite] QNN support for TFLite 2.1.0 quantized models (#5848)
authorAnimesh Jain <anijain@umich.edu>
Fri, 3 Jul 2020 01:13:56 +0000 (18:13 -0700)
committerGitHub <noreply@github.com>
Fri, 3 Jul 2020 01:13:56 +0000 (06:43 +0530)
commit575a3835315a533a19e871ee913f01142befebab
tree926756982888f7a340887cdbb4378ecc9d124022tree | snapshot
parent5248c141a151760a162f2afbba1b18993d530bbfcommit | diff
[TFLite] QNN support for TFLite 2.1.0 quantized models (#5848)

* [TFLite] TFLite 2.x parser quantization support.

* Address comments. Fix a bug for depthwise conv

* Added tests for relu, conv, quantize. Address comments.

* Using web-data. Minor refactoring.

* Removing TF hub package

* Trigger CI.

* Handle TFLite input layer naming.

* Addressing reviews.

* Retrigger CI.
python/tvm/relay/frontend/tflite.py diff | blob | history
src/relay/qnn/op/convolution.cc diff | blob | history
tests/python/frontend/tflite/test_forward.py diff | blob | history
Domain: Machine Learning / ML Framework;