deepspeedai/DeepSpeed

Pull Requests Commits

Update .github/workflows/nv-inference.yml

mrwyattii committed 1 year ago

Verified 2a267375

Update .github/workflows/cpu-inference.yml

mrwyattii committed 1 year ago

Verified b758fc1e

Update nv-inference.yml

mrwyattii committed 1 year ago

Verified 0eb72732

Update .github/workflows/cpu-inference.yml

mrwyattii committed 1 year ago

Verified 89e285e1

Update cpu-inference and nv-inference ttriggers

loadams committed 1 year ago

439585e7

Update .github/workflows/cpu-inference.yml

loadams committed 1 year ago

Verified 2aab5767

Change PR/schedule triggers for CPU-inference

loadams committed 1 year ago

e1167627

Add paths to run SD tests (#4919)

loadams committed 1 year ago

Verified 5457363c

[Fix] Fix cpu inference UT failure (#4430)

delock committed 1 year ago

Verified d8d865f4

Fix SD workflow to work with latest diffusers version (#4918)

lekurile committed 1 year ago

Verified 75db3d7d

Unit tests for MiCS (#4792)

zarzen committed 1 year ago

Verified d7b764e3

Add Habana Labs HPU accelerator support (#4912)

deepcharm committed 1 year ago

Verified 838c69be

Add WarmupCosineLR to Read the Docs (#4916)

dwyatte committed 1 year ago

Verified 59c5f37e

Nvme offload checkpoint (#4707)

eisene committed 1 year ago

Verified d058d4b3

fix num_kv_heads sharding in uneven autoTP for Falcon-40b (#4712)

Yejing-Lai committed 1 year ago

Verified 1787673e

[NPU]Add ZeRO-Infinity feature for NPU (#4809)

misstek committed 1 year ago

Verified b596963b

Release overlap_comm & contiguous_gradients restrictions for ZeRO 1 (#4887)

li-plus committed 1 year ago

Verified af033831

enable starcode((kv_head=1)) autotp (#4896)

Yejing-Lai committed 1 year ago

Verified 85132adc

[XPU] XPU accelerator support for Intel GPU device (#4547)

delock committed 1 year ago

Verified f4f31317

add sharded loading for safetensors in AutoTP (#4854)

sywangyi committed 1 year ago

Verified c8c57b8c

Support cpu tensors without direct device invocation (#3842)

abhilash1910 committed 1 year ago

Verified c84c28d2

Add support of Microsoft Phi-2 model to DeepSpeed-FastGen (#4812)

arashb committed 1 year ago

Verified 83427253

zero_to_fp32.py: Handle a case where shape doesn't have numel attr (#4842)

nelyahu committed 1 year ago

Verified 691458f8

Pipeline: Add support to eval micro bs configuration (#4859)

nelyahu committed 1 year ago

Verified ac84cf3f

Partition parameters: Minor refactoring of use_secondary_tensor condition (#4868)

deepcharm committed 1 year ago

Verified 81cc3207

[NPU] Fix npu offload bug (#4883)

CurryRice233 committed 1 year ago

Verified d873ce61

Fix f-string messages (#4865)

li-plus committed 1 year ago

Verified 9ec55bd9

optimize grad_norm calculation in stage3.py (#4436)

mmhab committed 1 year ago

Verified ea0d8114

Remove hooks on gradient accumulation on engine/optimizer destroy (#4858)

chiragjn committed 1 year ago

Verified 40342055

deepspeed engine: fp16 support validation on init (#4843)

nelyahu committed 1 year ago

Verified 3e94f8c7

Older