microsoft/DeepSpeed

Pull Requests Commits

Merge branch 'master' into loadams/update-transformers

loadams committed 2 years ago

Verified afc52e8e

Switch paths-ignore to single quotes, update paths-ignore on nv-pre-compile-ops (#4805)

loadams committed 2 years ago

Verified 3324efd9

Add support of Falcon models (7b, 40b, 180b) to DeepSpeed-FastGen (#4790)

arashb committed 2 years ago

Verified a7900bcc

Unpin even latest

loadams committed 2 years ago

f2fdb3b8

Add more to torch version

loadams committed 2 years ago

dd1d4ccb

[CI] Use latest transformers

loadams committed 2 years ago

7b7c8760

Fix for stage3 when setting different communication data type (#4540)

BacharL committed 2 years ago

Verified b1868169

[NPU] Add HcclBackend for 1-bit adam, 1-bit lamb, 0/1 adam (#4733)

RUAN-ZX committed 2 years ago

Verified 2ce6bf8c

improve the way to determine whether a variable is None (#4782)

RUAN-ZX committed 2 years ago

Verified 7b818ee9

[BUG] partition_balanced return wrong result. (#4312)

zjjMaiMai committed 2 years ago

Verified 2bdf061f

BF16_Optimizer: add support for bf16 grad acc (#4713)

nelyahu committed 2 years ago

Verified ce607080

fix autoTP issue for mpt (trust_remote_code=True) (#4787)

sywangyi committed 2 years ago

Verified 29f840fd

Fix Hybrid Engine metrics printing (#4789)

lekurile committed 2 years ago

Verified 5f41bd06

Run workflows when they are edited (#4779)

loadams committed 2 years ago

Verified 93a81b53

Update dockerfile with updated versions (#4780)

loadams committed 2 years ago

Verified 15ed83a9

Resolve any '..' in the file paths using os.path.abspath() (#4709)

rraminen committed 2 years ago

Verified 9dfb06de

Fix DS Stable Diffusion for latest diffusers version (#4770)

lekurile committed 2 years ago

Verified f57fc4c9

bump version to 0.12.5

mrwyattii committed 2 years ago

Verified 4997b0fd

Add safetensors support (#4659)

Jih-neng Lin committed 2 years ago

Verified 7122362f

fix: to solve #4726 (#4727)

RUAN-ZX committed 2 years ago

Verified 02288bc1

Fix logger formatting for partitioning flags (#4728)

OAfzal committed 2 years ago

Verified 8640b8e5

fix mics run with offload++ (#4749)

GuanhuaWang committed 2 years ago

Verified ba6bfd0b

Re-organize `.gitignore` file to be parsed properly (#4740)

aphedges committed 2 years ago

Verified 3418b869

Add nv-sd badge to README (#4747)

loadams committed 2 years ago

Verified 8db9379f

Install `deepspeed-kernels` only on Linux (#4739)

aphedges committed 2 years ago

Verified 969b8cf5

use all_gather_into_tensor instead of all_gather (#4705)

taozhiwei committed 2 years ago

Verified fd0a52c1

[docs] Intel inference blog (#4734)

jeffra committed 2 years ago

Verified 6b8103b4

Communication Optimization for Large-Scale Training (#4695)

RezaYazdaniAminabadi committed 2 years ago

Verified 2afa1c7f

Update flops profiler to recurse (#4374)

loadams committed 2 years ago

Verified 61391229

Remove unneeded dict reinit (fix for #4565) (#4702)

eisene committed 2 years ago

Verified a603a213

Older