microsoft/DeepSpeed

Pull Requests Commits

Merge bloom+zero_inference

tjruwase committed 3 years ago

52d49a2e

Merge branch 'olruwase/parameter_offload' of github.com:microsoft/DeepSpeed into olruwase/parameter_offload

tjruwase committed 3 years ago

a1757547

Release swap buffers for persisted params

tjruwase committed 3 years ago

59dc8f14

Add Inference support for running the BigScience-BLOOM Architecture

Reza Yazdani committed 3 years ago

7cf7c89f

Comms Benchmarks (#2040)

Quentin-Anthony committed 3 years ago

Verified 9b70ce56

Fix missing import in replace_module.py (#2050)

aphedges committed 3 years ago

Verified 76ea0534

correct partition_id in fp32 param -> fp16 param for MoE+z2 (#2058)

siddharth9820 committed 3 years ago

Verified 38a00bee

Fix for AMD unit tests (#2047)

mrwyattii committed 3 years ago

Verified 7bae53d1

Fix the half-precision version of CPU-Adam (#2032)

RezaYazdaniAminabadi committed 3 years ago

Verified a04480e1

Add compression papers (#2042)

conglongli committed 3 years ago

Verified ff87c4e1

fixed print statement (#2038)

mrwyattii committed 3 years ago

Verified 52181779

Split parameter offload from z3 (#2009)

tjruwase committed 3 years ago

Verified 678c3fe3

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified 8993ec43

Retain available params until last use (#2016)

tjruwase committed 3 years ago

Verified 2a1a4096

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified ecc34b17

Fix inference unit test import error catching (#2024)

mrwyattii committed 3 years ago

Verified ec1ec204

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified afb6f6cb

fix import errors (#2026)

KMFODA committed 3 years ago

Verified 735406e5

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified 92b90f30

Use partition size (#2011)

tjruwase committed 3 years ago

Verified d86a2de9

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified 60111c5c

DeepSpeed Monitor Module (Master) (#2013)

Quentin-Anthony committed 3 years ago

Verified c87f6ee2

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified e86c41ff

DataLoader Length Fix (#1718)

Sanger2000 committed 3 years ago

Verified ae198e20

Fix transformer API for training-evaluation pipeline (#2018)

RezaYazdaniAminabadi committed 3 years ago

Verified 5dce73fa

DeepSpeed examples refresh (#2021)

jeffra committed 3 years ago

Verified 7c3344e2

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified 55739669

[inference] test suite for ds-kernels (bert, roberta, gpt2, gpt-neo, gpt-j) (#1992)

jeffra committed 3 years ago

Verified b666d5cd

Merge branch 'master' into olruwase/parameter_offload

tjruwase committed 3 years ago

Verified 2a1b2451

[CI] force upgrade HF dependencies & output py env (#2015)

jeffra committed 3 years ago

Verified e6f444ae

Older