Segmentation Fault on M1 Mac.

Running the sample, _python dlrm_spytorch.py --mini-batch-size=2 --data-size=6 --debug-mode

Outputs:

Using CPU... model arch: mlp top arch 3 layers, with input to output dimensions: [8 4 2 1]

of interactions

8 mlp bot arch 2 layers, with input to output dimensions: [4 3 2]

of features (sparse and dense)

4 dense feature size 4 sparse feature size 2

of embeddings (= # of sparse features) 3, with dimensions 2x:

[4 3 2] data (inputs and targets): mini-batch: 0 tensor([[0.6965, 0.2861, 0.2269, 0.5513], [0.7195, 0.4231, 0.9808, 0.6848]]) tensor([[1, 2], [1, 1], [1, 1]], dtype=torch.int32) [tensor([1, 0, 1]), tensor([0, 1]), tensor([1, 0])] tensor([[0.3618], [0.2283]]) mini-batch: 1 tensor([[0.2937, 0.6310, 0.0921, 0.4337], [0.4309, 0.4937, 0.4258, 0.3123]]) tensor([[1, 2], [1, 2], [1, 1]], dtype=torch.int32) [tensor([3, 0, 2]), tensor([1, 1, 2]), tensor([1, 1])] tensor([[0.6031], [0.5451]]) mini-batch: 2 tensor([[0.3428, 0.3041, 0.4170, 0.6813], [0.8755, 0.5104, 0.6693, 0.5859]]) tensor([[2, 1], [1, 2], [1, 1]], dtype=torch.int32) [tensor([2, 3, 2]), tensor([0, 0, 2]), tensor([1, 1])] tensor([[0.5568], [0.1590]]) initial parameters (weights and bias): [[-0.34693 0.19553] [-0.18123 0.19197] [ 0.05438 -0.11105] [ 0.42513 0.34167]] [[-0.16466 -0.52702] [-0.22543 -0.11757] [ 0.23667 0.57199]] [[-0.20377 0.3713 ] [ 0.13177 0.27111]] [[-0.16825 -0.58044 -0.39152 -0.64812] [ 1.11561 0.0879 0.61481 -0.67743] [ 0.09677 0.62959 -0.17907 0.55115]] [-0.62618 -0.7872 0.21905] [[-0.23981 0.40607 -1.25093] [ 0.45048 1.64331 -0.01557]] [0.02414 0.12696] [[-0.76015 0.17397 -0.65541 -0.1746 0.5074 -0.30015 0.20463 0.41345] [ 0.1138 -0.55969 -0.13573 0.79993 -0.82672 -0.11259 -0.2254 0.04929] [ 0.30546 0.65675 -0.11032 0.33164 0.20402 0.19365 -0.23022 -0.40715] [-0.44909 -0.30881 0.13133 0.31066 0.13206 -0.22411 0.73728 0.62007]] [-0.177 -0.41172 0.06511 0.63365] [[ 0.19212 0.32132 -0.12244 0.26343] [ 0.89174 -0.13837 0.08274 0.14654]] [ 0.20062 -0.99836] [[-1.53246 -0.83254]] [0.16794] time/loss/accuracy (if enabled): zsh: segmentation fault python dlrm_s_pytorch.py --mini-batch-size=2 --data-size=6 --debug-mode

facebookresearch / dlrm

Segmentation Fault on M1 Mac. #371

of interactions

of features (sparse and dense)

of embeddings (= # of sparse features) 3, with dimensions 2x: