10. Changelog

10.1. v1.0 (Poplar SDK 1.4)

10.1.1. New features

  • Added support for torch.nn.InstanceNorm1d, torch.nn.InstanceNorm2d and torch.nn.InstanceNorm3d

  • Fixed issue with torch.nn.GroupNorm where only 4-dimensional inputs could be used

  • Replaced Adam with AdamW optimizer.

  • Added support for the following loss functions:

    • torch.nn.KLDivLoss

    • torch.nn.PoissonNLLLoss

    • torch.nn.HingeEmbeddingLoss

    • torch.nn.BCEWithLogitsLoss

    • torch.nn.SmoothL1Loss

    • torch.nn.SoftMarginLoss

    • torch.nn.CosineEmbeddingLoss

    • torch.nn.MarginRankingLoss

    • torch.nn.TripletMarginLoss

    • torch.nn.NLLLoss for aten::nll_loss2d

  • Added support for torch.optim.RMSprop optimizer

  • Added support for bool inputs to models

  • Improved support for half type models and inputs.

    • Using a mix of float 16 and float 32 inputs is now supported. Please see the documentation for cases in which a model might use different types compared to when run natively with PyTorch.

  • Added support for serialized matrix multiplications (poptorch.serializedMatMul)

  • Add support for POPTORCH_IPU_MODEL_VERSION environment variable.

  • Added support for torch.cumsum

  • Add support for pipelined / phased / sharded execution.

  • Add PoplarExecutor.compile() to compile the model without executing it.

  • Use sphinx-build to generate the documentation.

  • Use Miniconda as build environment.

  • Added support for torch.meshgrid

  • Added support for torch.cartesian_prod

  • Optimized torch.matmul implementation with limitations

    • Fused its input 0’s batch dimensions with the row dimension to avoid ReduceSum in its backward pass, for performance purpose

  • Added partial support for torch.einsum

    • Diagonals and ellipsis notation is unsupported

  • Add support for executable caching: poptorch.Options.enableExecutableCaching()

  • Add optional title argument to poptorch.ipu_print_tensor

  • Add len() method to poptorch.AsynchronousDataLoader

  • Added support for LAMB optimizer

  • Add support for recomputationCheckpoint()

  • Added support for torch.tensordot

  • Add support for rounding up the number of IPU used to allow models which specify of number of IPUs which is not a power of 2: poptorch.Options.autoRoundNumIPUs(True) NB, this will reserve but not use IPUs and so it is preferable to specify the model to use a number of IPUs which is a power of two

  • Optimized torch.matmul implementation with limitations

    • Fused its input 0’s batch dimensions with the row dimension to avoid ReduceSum in its backward pass, for performance purpose

  • Added support for multi-convolutions with poptorch.MultiConv

  • Added support for PopART batch serialization settings

    • These can be set via poptorch.Options().Popart.set()

  • Support for PopVision System Analyser added: tracing can be enabled by setting PVTI_OPTIONS='{"enable":"true"}'

10.1.2. Known issues

  • Race condition in poptorch.DataLoader when using several workers resulting in the iteration sometimes finishing one element early.

    • Workaround: set num_workers to 0 or 1.

  • poptorch.custom_op() doesn’t allow the user to set attributes.

    • Workaround: hardcode the attributes in the custom operation or pass them as regular inputs.

  • Graphs containing block annotations (poptorch.Block or poptorch.BeginBlock) cannot be exported using torch.save()

    • Workaround: Make a soft copy of the model that doesn’t contain Blocks and use it to save /load the weights. (The weights should be shared between the two models).

  • Lists of tensors are not supported as inputs.

    • Workaround: Use tuples instead.

      # Use a tuple
assert inference_model((t1, t2)) # instead of [t1, t2]

10.2. v0.1 (Poplar SDK 1.3)

10.2.1. New features

  • PopTorch now exposes PopART anchor options to choose how much data to return from a model. These are passed into the model wrapper via anchor_mode. options are Sum, All, Final and EveryN.

  • Add support for batched LSTM and batch first

  • An Options object can now be passed to poptorch.trainingModel / poptorch.inferenceModel to configure the session and select IPUs

  • The ‘profile’ option has been removed, instead profiling can be enabled by setting the environment variable POPLAR_ENGINE_OPTIONS='{autoReport.all:true, autoReport.directory:.}'

  • Add support for POPTORCH_IPU_MODEL and POPTORCH_WAIT_FOR_IPU environment variables.

  • Adds support for the torch comparisons operations:

    • torch.eq

    • torch.ge

    • torch.gt

    • torch.le

    • torch.lt

    • torch.max

    • torch.min

    • torch.ne

    • torch.isnan

    • torch.topk

    • torch.min and torch.max only support (tensor, tensor) and (tensor) overloads. They do not support the (tensor, dim=, keepdim=) overload.

    • torch.topk only supports sorted=False and Largest=True

  • Automatically synchronise the weights back to the Host after using the IPU for training. (i.e no need to explicitly call copyWeightsToHost() anymore)

  • Adds support for non-linear activations torch.nn.PReLU and torch.nn.Hardtanh

  • Adds support for Adam optimizer.

  • Adds support for half type models and inputs.

    • Models that require operations on input tensors of mixed precision are not currently supported. For example:

      def forward(self, x, y):
  x // Half
  y // Float32
  return x + y // Not supported.

  • Support for tensor.fill_, torch.full, torch.full_like

  • Adds support for user provided custom operations. See PopART documentation for information on how to write them. They are exposed by poptorch.custom_op this takes in a list of input tensors, strings for the PopART op name and domain, the domain version, and a list of tensors the same shape and size as the expected output tensors. This is to ensure the pytorch trace remains valid as it traces on CPU so won’t actually execute the operation when building the graph.

  • Adds support for torch.nn.Conv1D / torch.nn.Conv2D / torch.nn.Conv3D

  • Adds support for torch.nn.Upsample (‘nearest’ mode only)

  • Adds support for tensor.size

  • Adds support for the following random sampling operations.

    • torch.rand

    • torch.uniform_

    • torch.distributions.Uniform

    • torch.randn

    • torch.normal

    • torch.normal_

    For repeatable random number generation use the randomSeed method of poptorch.Options

  • Adds support for torch.clamp

  • Adds poptorch.DataLoader

  • Adds optimized poptorch.AsynchronousDataAccessor which allows for a dataloader to be offloaded to a background thread asynchronously.

  • Adds support for torch.norm

  • Upgraded from torch 1.5.0 to torch 1.6.0

  • Added experimental support for single host distributed execution

  • Added torch.where and tensor.masked_fill