Changes to `ingest-qonnx`#461

Merged

jmitrevs merged 14 commits intoingest-qonnxfrom

ingest-qonnx-thesps

Dec 2, 2021

Commits on Oct 11, 2021

fix batched multiple inputs
jmduarte
committed

Commits on Oct 12, 2021

Merge pull request #414 from fastmachinelearning/multiple_inputs_fixes
Thea Aarrestad
authored

Commits on Nov 3, 2021

Fixed 'qkeras_mnist_dense' example build problem #423

siorpaes
authored and
thesps
committed

Commits on Nov 4, 2021

Update for pyyaml 6.0 (#435)
thesps
authored

Commits on Nov 5, 2021

`axi_stream_driver` update (#420)
nghielme
authored

Commits on Nov 9, 2021

Reshape fixes: don't repack stream for flatten; remove final reshape (#443)

jmduarte
and
thesps
authored

Commits on Nov 10, 2021

Reorder loops in im2col_2d_cl given resource strategy issue. Reenable relevant test. Use 5000 MNIST samples rather than full dataset for faster testing
thesps
committed

Commits on Nov 11, 2021

Support applying Softmax over multidimensional tensors (#384)

vloncar
and
nghielme
authored
Disable some unsupported layers

thesps
authored and
vloncar
committed

Commits on Nov 12, 2021

Set appropriate data type for quantized_relu activations

thesps
authored and
vloncar
committed
Display unsigned types properly in profiling

thesps
authored and
vloncar
committed

Commits on Nov 25, 2021

Commits on Nov 26, 2021

Don't convert Quant to BatchNorm. Convert weight-Quant to Constant, and activation-quant to Activation
thesps
committed