pytorch/xla

Pull Requests Commits

Update gpu.md to run ResNet example on GPU (#7222)

ahoblitz committed 1 year ago

Verified 71c25e65

Add build trigger for 2.4.0-rc1 release (#7236)

bhavya01 committed 1 year ago

Verified 7bedfe6e

force to string output (#7239)

zpcore committed 1 year ago

Verified 02559649

Disable test_mp_early_exit.py temporarily (#7229)

ManfeiBai committed 1 year ago

Verified 258f9a71

Add python3.11 doc for XLA v2.3 (#7166)

Mon-ius committed 1 year ago

Verified cd8ac11b

Fix __getitem__ (#7237)

qihqi committed 1 year ago

Verified ad126b8b

[SPMD] Support reduce-scatter in manual sharding (#7231)

alanwaketan committed 1 year ago

Verified 70d2e9ea

Support int4 weight in quantized matmul/linear (#7235)

lsy323 committed 1 year ago

Verified ac371fb8

[ZeRO-1] Sync features from r2.1_aws_neuron branch (#7132)

hgt312 committed 1 year ago

Verified 551a76cd

temporarily skip the pjrt test (#7230)

zpcore committed 1 year ago

Verified d0fb59e7

Use std::optional (#7228)

cyyever committed 1 year ago

Verified 2c71e9b3

torchbench CI (#7162)

zpcore committed 1 year ago

Verified 291764f6

Fix wrong env flag for using auto sharding in documentation (#7225)

lausannel committed 1 year ago

Verified 28c5e141

Remove stack dump for compiler failure which can happen often (#7207)

jeffhataws committed 1 year ago

Verified 5338a3b0

Copy examples dir for TPU CI (#7199)

JackCaoG committed 1 year ago

Verified 1ac8968d

Add stub plugins for XPU and Neuron (#7209)

will-cromar committed 1 year ago

Verified f9851b36

Remove unused variables from run_tests.sh (#7217)

bhavya01 committed 1 year ago

Verified de283c98

init glossary (#7214)

ahoblitz committed 1 year ago

Verified 66c7ee8b

Training example for llama3; misc changes to make the training work. (#7194)

qihqi committed 1 year ago

Verified f4a612cb

Add int8 per channel weight-only quantized matmul (#7201)

lsy323 committed 1 year ago

Verified 56ddd5de

bhavya01 committed 1 year ago

Verified 4a30ea7d

ask user to install tf-cpu instead for the profiler (#7212)

JackCaoG committed 1 year ago

Verified 26327a2f

let 3.8 compatiable (#7210)

zpcore committed 1 year ago

Verified 8748a2fb

Allow user to configure distributed runtime service. (#7204)

vanbasten23 committed 1 year ago

Verified 0e1f765c

Add options in benchmark script to keep model on CUDA (#7193)

vanbasten23 committed 1 year ago

Verified f95454ac

Let grad ckpt apply opt-barrier to all params and buffers (#7206)

alanwaketan committed 1 year ago

Verified aec27305

Make clip_grad_norm_ follow the input's dtype (#7205)

alanwaketan committed 1 year ago

Verified f5c00624

openxla pin update to 06/05 (#7200)

zpcore committed 1 year ago

Verified 676853f2

[Pallas] Make _histogram output int32 (#7202)

alanwaketan committed 1 year ago

Verified 218f4e54

Update images to support dark mode (#7195)

loganthomas committed 1 year ago

Verified cdb4aec6

Newer Older