5.9. PopART

3.4.0

Add createHostTransferableTensorWithOffset session option (experimental). This can prevent out of memory errors for models with a large number of small input tensors.
Enable custom ops created using the Python API to accept a list of float values as attributes.

None

Reduce the compilation time for the following:
- when using overlapped I/O.
- for models with a large number of inputs.
Remove the limit on the number of I/O tiles. The number of I/O tiles was previously limited to 192.

None

None

Added support for the largest=false option in the TopK operation.
Added the NormalizeImage op that normalises the data and pads to four channels on-device after the data has been transferred to the IPU. This can enable better performance for subsequent convolution operations.
Added the stashAllTensorsInferencePipeline option to the SessionOptions class to enable all tensors to be stashed when doing inference. This may improve performance for certain use cases.
Added the ability to fine-tune the auto virtual graph result with the virtualGraphSplitRatios option in the SessionOptions class.
Added the SplineBasis and SplineWeighting ops to support the PyTorch Geometric SplineConv operator.
Switched use of the deprecated deviceIteration to deviceIterations.

Fixed the Loop and Scan ops so that any constant tensors are implicitly added into the subgraphs of these ops as inputs.
Fixed the overflow in the Clip operator.

None

None