For Users
torch.compile
aiu-smi
aiu-trace-analyzer
clone
sendnn
torch.spyre.is_available()
torch.spyre.device_count()
torch.spyre.current_device()
torch.spyre.set_device()
torch.spyre.is_initialized()
torch.spyre.manual_seed()
torch.spyre.manual_seed_all()
torch.spyre.Stream
torch.spyre.stream()
torch.spyre.current_stream()
torch.spyre.default_stream()
torch.spyre.synchronize()
torch_spyre._C.SpyreTensorLayout
torch_spyre._C.DataFormats
torch_spyre._C.get_spyre_tensor_layout()
torch_spyre._C.set_spyre_tensor_layout()
torch_spyre._C.get_downcast_warning()
torch_spyre._C.set_downcast_warning()
torch_spyre.constants.DEVICE_NAME
For Developers
add
bundle.mlir
ir.Operation
Loops
coarse_tile()
CustomPreSchedulingPasses
insert_tiling_propagation
CountedLoopSchedulerNode
FusedSchedulerNode
LoopSpec
OpSpec.tiled_symbols
op_spec.py
loop_group_id
SuperDSCScheduling.codegen_node()
codegen_kernel()
ForLoopGraphInfo
span_reduction
work_distribution