Pull Requests pytorch/xla

Implement torch.linspace and torch.logspace

#8533 opened 2025-01-06 18:50 by vyom1611

Support S32 index operations for copy and fill

#8463 opened 2024-12-06 00:31 by rpsilva-aws

update test_replace_xla_tensor

#8458 opened 2024-12-05 22:40 by laithsakka

Remove the constraint of pages_per_sequence%num_kv_pages_per_compute_…

#8457 opened 2024-12-05 21:55 by vanbasten23

[WIP] Update test_ops.py

#8449 opened 2024-12-04 00:05 by ManfeiBai

[torch_xla2] Create jtorchvision_roi_align.py

#8394 opened 2024-11-19 07:53 by ManfeiBai

norm and pca_lowrank op info test from issue #7528

#8384 opened 2024-11-14 23:16 by vyom1611

WIP implement lowering for torchvision::roi_align

#8362 opened 2024-11-07 00:08 by barney-s

custom_kernel: fix shape mismatch by sharding segment_ids in flash attn. pallas

#8333 opened 2024-10-29 22:26 by dudulightricks

[release] add permission of iam.auto.tfvars

#8326 opened 2024-10-25 21:32 by ManfeiBai

debug DO_NOT_MERGE

#8293 opened 2024-10-21 07:14 by zpcore

[DO NOT REVIEW] Experiment the pl.debug_print in Pallas.

#8284 opened 2024-10-18 18:27 by vanbasten23

[DO NOT REVIEW YET] Extend paged attention

#8237 opened 2024-10-08 20:53 by vanbasten23

Check warning msg in second captured warning.

#8069 opened 2024-09-25 08:48 by zhuhaozhe

Support FSDPv2 compute dtype

#8056 opened 2024-09-24 10:20 by lausannel

Part 2. Introduce multi-node SPMD initialization for Neuron

#8046 opened 2024-09-19 23:34 by rpsilva-aws

Forward `XLATensorImpl::is_contiguous_custom` to `TensorImpl`.

#8032 opened 2024-09-17 17:51 by ysiraichi

[benchmarks] Add docs file.

#8023 opened 2024-09-16 16:53 by ysiraichi

support dist.broadcast usability

#7956 opened 2024-09-05 01:53 by zpcore

Update dedup logic to use dedup_save_plans

#7929 opened 2024-08-29 19:56 by jonb377

[ROCm] Add initial ROCm PJRT support xla:gpu runtime rocm

#7896 opened 2024-08-21 12:14 by mmakevic-amd

[WIP] Skip AOTAutograd view-reconstruction for inference runs. xla:gpu

#7834 opened 2024-08-12 15:31 by ysiraichi

API_GUIDE.md: fix execution => exception typo + rewording

#7808 opened 2024-08-06 09:04 by houeland

Support SplashAttention in PyTorch/XLA

#7798 opened 2024-08-01 18:28 by wonjoo-wj

usability

#7728 opened 2024-07-23 17:54 by zpcore

Support of AOT compilation (refine #6992)

#7581 opened 2024-06-26 20:24 by zpcore

Update HLO instructions as per issue #7188 docathon-h1-2024 medium

#7317 opened 2024-06-19 09:27 by data-scientist-arshi

[fori_loop|while_loop] add test case for Linear model with changeable weight

#7302 opened 2024-06-17 21:55 by ManfeiBai

Modularize zero step function

#7265 opened 2024-06-13 00:17 by karakusc

[ZeRO-1] Only warn on rank-0

#7252 opened 2024-06-12 02:06 by hgt312