+ jTN^RIHt^RIt^RIt^RItRRltRRltRRltR#) )SequenceNcV^8dQhR\\P,R\\\ ,\\ ,3,/#tensorsreturn)rtorchTensortuplelistintobject)formats"X/home/luca/screen-checking/.venv/lib/python3.14/site-packages/torch/_lazy/computation.py __annotate__rs; U U ell # U 49d6l "# Ucf\PPP\ V44#)zReturn tensor ids and eager tensors for DeviceData nodes in the IR for the passed in lazy tensors. TODO: This API is currently ts backend specific. We are working on generalizing it to all backends including XLA. )r _C_lazy_ts_backend _get_tensors_ts_device_data_noder rs&rget_tensors_ts_device_data_noders# 88 $ $ E Ed7m TTrcZV^8dQhR\\P,R\/#r)rr r str)rs"rrrs#99HU\\29s9rcf\PPP\ V44#)z4Return the graph hash for the passed in lazy tensors)r r_lazy_get_graph_hashr rs&rget_graph_hashrs 88>> ) )$w- 88rc|V^8dQhR\R\\,R\\P ,/#)rhash_str graph_inputsr)rrrr r r )rs"rrrs7UUU!)&!1U %,,Urch\PPPV\ V44#)zRunning the cached computation graph with the given inputs TODO: This API is currently ts backend specific. We are working on generalizing it to all backends including XLA. )r rr_run_cached_graphr )r r!s&&rrun_cached_graphr$s& 88 $ $ 6 6xlAS TTr)collections.abcrr torch._C._lazytorch._C._lazy_ts_backendrrr$rrr)s"$   U9 Ur