Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Nov 7, 2024

• DeepSpeed-Domino (#929)

  Show description for faa0420
  

  6 peopleauthored
  faa0420

  Copy full SHA for faa0420

Commits on Oct 30, 2024

• Remove the fixed `eot_token` mechanism for SFT (#927)

  Show description for eefb0ef
  

  3 peopleauthored
  eefb0ef

  Copy full SHA for eefb0ef

Commits on Oct 29, 2024

• Εnable reward model offloading option (#930)

  Show description for aa4459f
  

  3 peopleauthored
  aa4459f

  Copy full SHA for aa4459f

• Replace deprecated transformers.deepspeed module (#872)

  Show description for cab3361
  

  HollowMan6

  and

  loadams

  authored
  cab3361

  Copy full SHA for cab3361

• Fixed mistake in readme (#933)

  Show description for 5a61193
  

  SCheekati

  and

  tjruwase

  authored
  5a61193

  Copy full SHA for 5a61193

• [cifar ds training]: Set cuda device during initialization of distributed backend. (#931)

  Show description for 130fb58
  jagadish-amd

  authored
  130fb58

  Copy full SHA for 130fb58

Commits on Sep 17, 2024

• Enable overlap_comm for better performance (#846)

  Show description for f73a6ed
  

  li-plus

  and

  tjruwase

  authored
  f73a6ed

  Copy full SHA for f73a6ed

• DeepNVMe ZeRO-inf Tutorial (#921)

  Show description for 90c2a9f
  

  jomayeri

  and

  loadams

  authored
  90c2a9f

  Copy full SHA for 90c2a9f

Commits on Sep 10, 2024

• Fix labels & eos_token for SFT (#819)

  li-plus

  authored
  a256c04

  Copy full SHA for a256c04

Commits on Sep 3, 2024

• Update requirements for opencv-python CVE (#925)

  Show description for c961379
  loadams

  authored
  c961379

  Copy full SHA for c961379

• extend max_prompt_length and input text for 128k evaluation (#891)

  Show description for 1293d45
  

  HeyangQin

  and

  loadams

  authored
  1293d45

  Copy full SHA for 1293d45

Commits on Aug 24, 2024

• DeepNVMe README.md add xref (#919)

  stas00

  authored
  957ae31

  Copy full SHA for 957ae31

Commits on Aug 21, 2024

• DeepNVMe example scripts (#914)

  Show description for 0d40b31
  

  tjruwase

  and

  Ubuntu

  authored
  0d40b31

  Copy full SHA for 0d40b31

• Add openai client to deepspeedometer (#913)

  Show description for 9563904
  

  delock

  and

  tjruwase

  authored
  9563904

  Copy full SHA for 9563904

Commits on Aug 16, 2024

• Update README.md (#916)

  keshavkowshik

  authored
  8d91a5a

  Copy full SHA for 8d91a5a

Commits on Aug 14, 2024

• Enable cpu/xpu support for the benchmarking suite (#905)

  Show description for b04fedd
  louie-tsai

  authored
  b04fedd

  Copy full SHA for b04fedd

Commits on Jun 26, 2024

• Refactored LLM benchmark code (#899)

  Show description for bbab278
  mrwyattii

  authored
  bbab278

  Copy full SHA for bbab278

Commits on Jun 6, 2024

• add client-only mode to mii benchmark (#900)

  Show description for 75df1d7
  

  delock

  and

  lekurile

  authored
  75df1d7

  Copy full SHA for 75df1d7

Commits on May 1, 2024

• fix bug with queue.empty not being reliable (#898)

  Show description for cce6223
  mrwyattii

  authored
  cce6223

  Copy full SHA for cce6223

Commits on Apr 30, 2024

• Update tokens_per_sec calculation to work w/ stream and non-stream cases (#897)

  Show description for fdb8ee2
  lekurile

  authored
  fdb8ee2

  Copy full SHA for fdb8ee2

Commits on Apr 29, 2024

• updating tokens per second to include the token count of generated tokens. (#895)

  Show description for 1be0fc7
  

  guptha23

  and

  Chandra Sekhar Gupta Aravapalli

  authored
  1be0fc7

  Copy full SHA for 1be0fc7

Commits on Apr 9, 2024

• Update throughput-latency plot script (#881)

  Show description for fab5d06
  lekurile

  authored
  fab5d06

  Copy full SHA for fab5d06

Commits on Mar 29, 2024

• Fix response check in call_aml function (#882)

  HeyangQin

  authored
  df7119e

  Copy full SHA for df7119e

Commits on Mar 27, 2024

• dynamic setting of requst num and formatting (#880)

  mrwyattii

  authored
  02fc578

  Copy full SHA for 02fc578

Commits on Mar 19, 2024

• change kwargs for AML call to match vllm kwargs (#876)

  mrwyattii

  authored
  279a8fe

  Copy full SHA for 279a8fe

• Improve robustness of infernece AML benchmark (#875)

  HeyangQin

  authored
  18200d5

  Copy full SHA for 18200d5

Commits on Mar 18, 2024

• Update README.md

  awan-10

  authored
  c3ffec2

  Copy full SHA for c3ffec2

Commits on Mar 14, 2024

• Fix AML benchmark E2E measurment (#874)

  Show description for b0a5533
  mrwyattii

  authored
  b0a5533

  Copy full SHA for b0a5533

Commits on Mar 8, 2024

• Xiaoxia/fp v1 (#871)

  Show description for f415ec8
  

  xiaoxiawu-microsoft

  and

  Ubuntu

  authored
  f415ec8

  Copy full SHA for f415ec8

Commits on Mar 6, 2024

• Update Inference Benchmarking Scripts - Support AML (#868)

  Show description for 6e9ada6
  lekurile

  authored
  6e9ada6

  Copy full SHA for 6e9ada6

• Remove AML key from args dict when saving results (#870)

  Show description for b7ec5c3
  lekurile

  authored
  b7ec5c3

  Copy full SHA for b7ec5c3

Commits on Mar 1, 2024

• catch AML error response, add aml script (#869)

  mrwyattii

  authored
  ffb8a4b

  Copy full SHA for ffb8a4b

Commits on Feb 29, 2024

• Extend FastGen benchmark to use AML endpoints (#865)

  Show description for 8182a8b
  

  mrwyattii

  and

  lekurile

  authored
  8182a8b

  Copy full SHA for 8182a8b

Commits on Feb 26, 2024

• <fill-mask>Modify codes so that different accelerators can be called according to specific device conditions (#863)

  Show description for 6540db6
  

  foin6

  and

  tjruwase

  authored
  6540db6

  Copy full SHA for 6540db6

Commits on Feb 22, 2024

• Fix path in human-eval example README (#862)

  Show description for 0ac02da
  lekurile

  authored
  0ac02da

  Copy full SHA for 0ac02da

Pagination

Previous

Next