microsoft/DeepSpeed

Pull Requests Commits

Merge branch 'master' into lekurile/kernel_hip_amd

Lev Kurilenko committed 1 year ago

84f9b59a

Stop tracking backward chain of broadcast in initialization (#5075)

tohtana committed 1 year ago

Verified 5a721de3

Fix verification for ZeRO3 leaf module (#5074)

tohtana committed 1 year ago

Verified f02d7bda

Remove old vector add code

Lev Kurilenko committed 1 year ago

010d2632

Further refactor deepspeed.moe.utils + deepspeed.moe.layer type hints (#5060)

ringohoffman committed 1 year ago

Verified 9922270f

Remove hip_ops folder

Lev Kurilenko committed 1 year ago

d7d309b3

Remove old Makefile

Lev Kurilenko committed 1 year ago

cddb8236

Lev Kurilenko committed 1 year ago

25e662b5

[doc/1-line change] default stage3_param_persistence_threshold is wrong in the doc (#5073)

ByronHsu committed 1 year ago

Verified 3e6d6069

Make batch size documentation clearer (#5072)

segyges committed 1 year ago

Verified dde64b00

[Zero++ qgZ] Fall back to reduce_scatter if `tensor.numel() % (2 * global_world_size) != 0` (#5056)

ByronHsu committed 1 year ago

Verified 592325ab

Add checker code - all activations passing

lekurile committed 1 year ago

da610166

adding hccl to init_distributed function description (#5034)

nelyahu committed 1 year ago

Verified 2eafe41b

Update import for changes to latest diffusers (#5065)

mrwyattii committed 1 year ago

Verified a049370c

load linear layer weight with given dtype (#4044)

polisettyvarma committed 1 year ago

Verified 567f97b2

Optimize grad_norm calculations by reducing device/host dependency (#4974)

nelyahu committed 1 year ago

Verified 61daaa1e

Delay reduce-scatter for ZeRO3 leaf modules (#5008)

tohtana committed 1 year ago

Verified 19e0dc39

Update host to device mem copy

lekurile committed 1 year ago

5593c13c

[NPU] Change log level to debug (#5051)

CurryRice233 committed 1 year ago

Verified 6de31de7

Fix broken model names in inference CI (#5053)

mrwyattii committed 1 year ago

Verified 449f9ad0

Fix compilation errors

lekurile committed 1 year ago

74ffe741

HIP-ify headers

lekurile committed 1 year ago

439e2dcf

Initial HIP work

lekurile committed 1 year ago

422ebd15

[doc] update inference related docs from `mp_size` to `tensor_parallel` for TP (#5048)

yundai424 committed 1 year ago

Verified 76ec8b49

MoE type hints (#5043)

ringohoffman committed 1 year ago

Verified 971d82b5

[NPU] Add NPU to support hybrid engine (#4831)

CurryRice233 committed 1 year ago

Verified 88cca60a

Fix nv-torch-latest-cpu CI (#5045)

mrwyattii committed 1 year ago

Verified 93e9537d

launcher_helper: enable fds passing (#5042)

YizhouZ committed 1 year ago

Verified 8f627700

update inference pages to point to FastGen (#5029)

mrwyattii committed 1 year ago

Verified 24f20ef0

Fix nv-accelerate and nv-torch-latest-v100. (#5035)

loadams committed 1 year ago

Verified 546d08ee

Older