pytorch/xla

Pull Requests Commits

Fix up pin update (#8935)

tengyifei committed 213 days ago

Verified 980ead54

Include torchax in torch_xla (#8895)

tengyifei committed 214 days ago

Verified bb5c6a0b

pin update (#8908)

lsy323 committed 214 days ago

Verified 6400e169

[ragged-paged-attn] Unify kv strided load to one. (#8929)

bythew3i committed 215 days ago

Verified c044c691

Add heuristic default block sizes for different cases in ragged attention kernel (#8922)

yaochengji committed 215 days ago

Verified 6c3f2313

Clean up deprecated APIs (#8927)

zpcore committed 215 days ago

Verified f1fe8bc0

Use f32 scratch for output so we only need to transfer output with desired dtype back to HBM. (#8924)

vanbasten23 committed 215 days ago

Verified f0881b5a

[torchax] Support linalg.det and linalg.lu_solve (#8872)

dvhg committed 216 days ago

Verified 2d6f57a3

Fix cuda dependency not found (#8903)

zpcore committed 216 days ago

Verified dbb03aa1

Lower `isneginf()`. (#8912)

ysiraichi committed 216 days ago

Verified cba9ff9f

Fix fp8 test (#8909)

lsy323 committed 217 days ago

Verified 985fa188

Add Pallas tutorials to the Pallas guide at torch_xla. (#8904)

vanbasten23 committed 217 days ago

Verified 450b4dce

point build to rc3 (#8907)

zpcore committed 217 days ago

Verified 2999e72b

Use shard_as in scan to ensure that inputs and their gradients have the same sharding (#8879)

tengyifei committed 220 days ago

Verified 6d88c089

[Build] Pin cmake to build PyTorch (#8902)

lsy323 committed 220 days ago

Verified fe3bb7f6

Remove dynamic grid (#8896)

bythew3i committed 221 days ago

Verified b24e6e95

Update cudnn dependency for newer cuda versions (#8894)

zpcore committed 221 days ago

Verified 92873377

8740 add single processing to getting started instructions (#8875)

pgmoka committed 221 days ago

Verified e341ff09

[ragged-paged-attn] Combine k_pages and v_pages on num_kv_head (#8892)

bythew3i committed 222 days ago

Verified 7a3c0518

Encapsulate Mesh invariants (#8882)

rpsilva-aws committed 222 days ago

Verified 96ad8f5f

Avoid unnecessary copy in TensorSource (#8849)

lsy323 committed 222 days ago

Verified 8dc5b496

update Dockerfile config for CI (#8886)

zpcore committed 222 days ago

Verified 3066b034

Increase CPU node count in tpu-ci (#8885)

tengyifei committed 222 days ago

Verified 6a7ce475

Cache HLO in xb.call_jax and support non-tensor args (#8878)

tengyifei committed 224 days ago

Verified a3ef52e5

Add lowering for `bitwise_right_shift`. (#8866)

ysiraichi committed 224 days ago

Verified 2aa11ccc

Correct type conversions in OpBuilder and add corresponding test. (#8873)

iwknow committed 224 days ago

Verified f2bdecfa

8727 create a site map or centralize links in readme (#8843)

pgmoka committed 224 days ago

Verified 3808df01

Add lowering for `bitwise_left_shift`. (#8865)

ysiraichi committed 224 days ago

Verified 46a7990f

Update README.md with torch_xla-2.8.0 wheels (#8870)

bhavya01 committed 227 days ago

Verified 8030f059

Support collective matmul optimization in mp (#8855)

yaochengji committed 227 days ago

Verified 2ad5b5d3

Newer Older